Home

Network Troubleshooting

1. 310 ATM 1 1 1 1 53 bytes e a gt lt gt gt 4 1 A2 P11 26 L2 PDU 1 A2 P10 Z5 L2 PDU 1 2 P9 24 L2 PDU 1 2 Z3 L2 PDU 1 2 P7 22 L2_PDU 1 A2 6 21 L2 PDU 1 A2 P5 F1 L2 PDU 1 A2 B1 L2_PDU 1 A2 G1 L2 PDU 1 2 2 M2 L2 PDU 13 14 1 2 P1 M1 L2 PDU nibbles 1 A2 C1 L2 PDU Trailer 125 us 1 A2 Alignment bytes path status PO P11 Path overhead identification B1 Bit interleaved parity BIP 8 Padding bit count PLCP path user channel SIP Layer 1 control information Z1 Z6 reserved for future use Alignment bytes A1 and A2 PLCP Path Status Byte Incoming signal Outgoing 4 bits bit 3 bits or LSS code LSS code connected connected Far End Block Error endi Link Status Signal TX link up connected rx link dn rX link up PLCP Out of Frame TX link up PLCP Loss of Frame TX link dn Path Overhead POH Identifiers Coding of the Link Status Field LSS Code 155 Name Status 000 connected Received Link Connected 011 rx link dn Received Link Down No Input or Forced Down 110 rx link up Received Link Up P8 00100000 P7 00011100 P6 00011001 P5 00010101 P4 00010000 P3 00001101 P2 00001000 P1 00000100 PO 00000001 Figure 10
2. 370 ATM Timer Default S S On First On Second impi Number Value tart top Expiration Expiration mplementation T301 gt 3 ALERT CONNECT Clear call M Q 2931 received received if symmetrical only connections supported T302 10 15s SETUP ACK SENDING COMPLETE if information in M Q 2931 sent indication complete clear call if overlap sending only Otherwise CALL and receiving PROCEEDING supported T303 4s SETUP CONNECT Repeat SETUP Abort call set up M sent CALL PROCEEDING restart T303 ALERT RELEASE COMPLETE received T304 30s SETUP ACK CALL PROCEEDING Clear call M Q 2931 received ALERT only restart when CONNECT received INFO received T308 30s RELEASE RELEASE Repeat sent COMPLETE or RELEASE RELEASE received restart T308 T309 10s SAAL aborted SAAL active again Clear call M delete VCIs and call references T310 30 120s CALL ALERT CONNECT send RELEASE M PROCEEDING or RELEASE received received T313 4s CONNECT CONNECT ACK send RELEASE M sent received T316 2 min RESTART RESTART Repeat Repeat M sent ACK received RESTART RESTART several times several times T317 Implemen RESTART call references Report error M tation received deleted specific but T316 T322 4s STATUS STATUS RELEASE Repeat Repeat M ENQUIRY or RELEASE STATUS ENQUIRY STATUS ENQUIRY sen COMPLETE received T398 4s DROP DROP PARTY Send Timer is not restarted M ATM Forum P
3. 2 0 e d S f 4 1 Public NNI Figure 10 42 ATM signaling in public and private ATM networks ATM NETWORK NG TROUBLESHOOTING LocAL AREA NETWORKS 1 0 360 Connection setup can take place either the reserved signaling virtual chan nel VPI VCI 0 5 or over VCs selected by means of meta signaling Any signal ing VCs can be set up without an existing AAL connection because the simple meta signaling commands can be transferred in one cell Once the signaling virtual channel has been determined the AAL for the signaling virtual channel SAAL is established At this point the signaling protocol becomes active In practice however the signaling virtual channel VPI VCI 0 5 is generally used and meta signaling is not necessary 10 1 9 1 The 0 2931 Message Format Q 2931 defines a total of 15 UNI signaling messages which can be classified in groups for connection setup connection clear down and other messages For all ordinary signaling processes the following ten message types are available Connection Setup ALERTING CALL PROCEEDING CONNECT CONNECT ACKNOWLEDGE SETUP Connection Clear Down RELEASE RELEASE COMPLETE Other NOTIFY STATUS STATUS ENQUIRY For interoperability with narrow band ISDN three more messages have been defined in addition to those listed previously
4. N soss 5955 5955 N ce La 50 N 1 N nas b WV b Jesn Jesn pejueuo uoneuuoo 55215 ssejuonoeuuoo 5 5521 Figure 10 36 Structure of AAL3 4 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 353 very short data packets more efficiently reducing the cell tax inefficient use of ATM cell payloads inherent in AAL3 4 and AALS ITU T Recommendation 1 366 1 describes the Segmentation and Reassembly Service Specific Conver gence Sublayer for AAL type 2 This allows the encapsulation of higher layer variable length data packets 1 to 65 535 bytes over AAL2 mini cells in almost exactly the same way that the SAR function of AAL5 allows the encapsulation of data packets over ATM cells see AAL type 5 in the following discussion Even the AAL2 SSTED PDU trailer structure is similar to that of the AAL5 CPCS PDU trailer it is 8 bytes long and provides a length field and CRC 32 field in the same positions thus allowing the re use of hardware designs for processing the SAR function Third generation wireless services make use of this AAL2 SAR process for carrying data services 10 1 8 4 AAL Type 3 4 AAL Type 3 4 ITU T Recommendation 1 363 3 specifies connection oriented and non connection oriented transportation of data packets in ATM networks Both point to point and point to multipoint connecti
5. Figure 10 57 Format of LANE data packets TROUBLESHOOTING LocAL AREA NETWORKS 1 0 382 16 Mbit s are calculated based on a Token Holding Timer value of 9 1 ms However LANE data packets contain no SD FCS ED or FS fields and no interframe spacing gap so that the resulting lengths are 4 544 and 18 190 bytes 10 1 10 4 Multiprotocol over ATM MPOA ATM networks can use LAN Emulation to emulate Ethernet and Token Ring topologies Such emulated ELANs can then form the basis of subnetworks virtual LANs that include network nodes of all three network topologies Ethernet Token Ring and ATM Any ATM network can contain several ELANs but LAN Emulation alone does not make it possible to connect ELANs to one another except through the use of conventional routers The MPOA specifica tion does away with this restriction MPOA or Multiprotocol over ATM inte grates LANE v2 with the Next Hop Resolution Protocol NHRP and the Multicast Address Resolution Protocol MARS and permits inter ELAN communication without the use of additional routers By separating the route selection and data forwarding functions MPOA is able to provide efficient routing even in large complex networks Every MPOA network consists of MPOA servers MPS and MPOA clients MPC MPOA servers are responsible for route selection while MPOA clients perform the actual forwarding of data from their LE Client interfaces requesting the appropriate ro
6. 1 End of a PDU i PT 1 48 bytes SDU ATM Layer m Cell information field 5 bytes 48 bytes Figure 10 37 Structure of AAL5 Convergence Sublayer CPCS and a Service Specific Convergence Sublayer SSCS The processes within the AAL Type 5 sublayers are significantly simpler than in AAL3 4 There is no mechanism for cell multiplexing for example cells that belong to AAL5 CS PDU are transmitted in one sequential cell stream The variable length data packets 1 to 65 535 bytes of the application building on 5 are first padded to an integer multiple of 48 bytes this ensures that no partially filled cells are transmitted after segmentation and a trailer is added TROUBLESHOOTING LocAL AREA NETWORKS 1 0 355 no header The resulting CS PDU is then broken into 48 byte segments that fit directly into the data field of ATM cells The PT field of the ATM header is used to identify whether further segments follow or whether the data field contains the end of the CS PDU Trailer based PDUs are efficient to process because the large amounts of data often involved can be further processed without having to move them in memory based implementations The next section will describe the SAAL which is also trailer based PDU building on AALS processing of the higher sublayers of the SAAL is thus simplified Additionally AAL5 allows processing on 32 bit boundaries
7. 2 AIS 18 R 34 6 5 3 PSN 19 R 35 EB8 4 N 20 R 36 R 5 21 R 37 R 6 MBS 22 R 38 R 7 NMB EDC 23 R 39 R 8 EDC B1 24 R 40 R 9 EDC B2 25 R 41 R 10 EDC B3 26 R 42 R 11 EDC B4 27 R 43 R 12 EDC B5 28 R 44 R 13 EDC B6 29 R 45 R 14 EDC B7 30 RDI 1 46 STET 15 EDC B8 31 47 10 16 R 32 EB2 48 Figure 10 20 PL OAM cell format F1 cells TROUBLESHOOTING LocAL AREA NETWORKS 1 0 327 mission path OAM cells have a special header that makes them easily reco gnizable Figure 10 19 Note that the lower limit of 27 exists to allow easier interworking between cell based interfaces running at 155 Mbit s and 622 Mbit s and the corresponding SDH SONET interfaces STM 1 OC 3 and STM 4c OC 12 because the overhead in SDH SONET SOH TOH POH occu pies exactly 1 27 of the bandwidth that is 9 1 columns out of 270 Figure 10 20 illustrates the structure of PL OAM 1 and 8 cells The following fields are reserved for the F1 and F3 flows OAM F1 Cell Fields PSN PL OAM Sequential Number 8 bits modulo 256 NIC Number of Included Cells maximum value 512 MBS Monitoring Block Size maximum value 64 NMB EDC Number of Monitored Blocks recommended value 8 EDC Error Detection Code BIP 8 value calculated from the cells of the MBS NMB EB Number of Monitored Blocks at the Far End recommended value 8 REI Remote Error Indication number of bit parity errors
8. Severely Errored Second 9 10 Cell Loss Ratio 0 8 0 6 0 4 0 2 0 0 Figure 10 65 Determining CDV using a protocol analyzer TROUBLESHOOTING LocAL AREA NETWORKS 1 0 405 Cell Misinsertion Rate The cell misinsertion rate is defined as the number of defective cells cells containing a wrong VPI VCI due to non corrected header errors transmitted within a time interval divided by this interval Cell Transfer Delay CTD The cell transfer delay is defined as the time t t between two corresponding cell transmission reception events CRE t and CRE t where t gt t Cell Delay Variation CDV Two types of variations in the cell transfer delay are defined one point cell delay variation which examines cells arriving at one measurement point and two point cell delay variation which examines cell delay variation at measurement point two relative to measurement point one One Point Cell Delay Variation The one point cell delay variation y for cell k at measurement point MP is defined as the difference between the reference arrival time c of the cell and Reference clock Measurement point MP Cell 0 t 0 a C T 1 Ds T a3 Clockitestan a C T Y Y as T Y 5 ak C Y Y yk Actual arriving time of cell Reference arrival time of cell One point cell delay va
9. semi permanent connection is set up through network management Meta signaling User to network signaling Network to network signaling Setting up permanent or semi permanent VCCs is a good idea especially when a relatively low number of network nodes need to be interconnected over stati cally defined connections or when dynamic signaling processes are undesired for security reasons In ATM networks with a large number of nodes connec TROUBLESHOOTING LocAL AREA NETWORKS 1 0 338 tions are set up dynamically over individual transmission sections using signal ing protocols UNI lt gt NNI_ Signaling from an end system into ATM network that consists of several networked switches user to network NNI lt gt NNI Signaling within an ATM network that consists of several net worked switches network to network Meta signaling is not used in practice because the use of the reserved signaling virtual channel VCI 5 has been found to be sufficient even in large ATM networks Virtual Path Connections VPCs Virtual path connections are a hierarchical level above virtual channel connec tions In other words a virtual path can contain several virtual channels note however that in terms of the layered model the VC is above the VP in the stack VPCs have the same properties as VCCs Path connections can likewise be set up manually as permanent virtual paths or on request by means of signaling proc
10. Connection Setup Additional Messages for N ISDN Interworking SETUP ACKNOWLEDGE PROGRESSING Other Additional Messages for N ISDN Interworking INFORMATION Furthermore the messages RESTART and RESTART ACKNOWLEDGE are de fined for the purpose of requesting a new connection setup attempt in the event TROUBLESHOOTING LocAL AREA NETWORKS 1 0 361 that a connection enters an undefined state for example These messages may be used only with the global call reference 0 Every signaling message consists of the following five sections called information elements Protocol discriminator Call reference Message type Message length Message specific information elements The first four information elements are mandatory and must be present in every Q 2931 message The usage of other information elements is dependent on the message type The diagram in Figure 10 43 illustrates the structure of a UNI signaling message Bit 8 7 6 5 4 3 2 1 Byte Protocol Discriminator 0000 Call Reference Length Flag Call Reference Call Reference Message Type Message Length ON Oo Information Elements Figure 10 43 Q 2931 message format Protocol Discriminator The protocol discriminator is the first element in every Q 2931 message It identifies the protocol used and is coded as 00001001 decimal 9 for Q 2931 The ATM Forum signal
11. 0010 Performance Management 0000 Forward Monitoring 0001 Backward Reporting 0010 Monitoring Reporting 1000 Activation Deactivation 0000 Performance Monitoring 0001 Continuity Check 0101 APS Automatic 0000 Group Protection Protection Switching 0001 Individual Protection Error type location Not used optional optional 6A hex Function specific fields of AIS RDI Error Management Cells TSTP MCSN BEDCon optiona Notused TRCC BLER TRCC 8bits 16bits _ 1665 16bits 32 bits 29 octets 16 bits 8bits 16 bits Function specific fields of Performance Management Cells BEDC TRCC BLER and TRCC Used only in Backward Reporting cells MCSN TUC TSTP and TUC Used for both types of PL OAM cells Used only in Forward Monitoring cells Figure 10 30 F4 and F5 OAM cell format TROUBLESHOOTING LocAL AREA NETWORKS 1 0 343 Fault management OAM cells are used to detect and localize communication faults and report them to the stations concerned Performance management OAM cells detect parameters such as the cell block ratio cell loss ratio or the cell misinsertion rate and thus provide information about the performance of a connection APS coordination protocol OAM cells are used to manage ATM protection switching Activation deactivation cells are used to start and stop OAM fun
12. ATM over 100 Copper Cable Cat 5 Unshielded Twisted Pair When Category 5 UTP cabling is used to transmit data at up to 155 Mbit s the segment length must not exceed 150 m As for Cat 3 wiring the standard connector is the 8 pin RJ 45 plug With higher quality receivers segments can be up to 350 m long The electrical signal is NRZ encoded ATM over 150 Shielded Twisted Pair STP The use of 150 shielded twisted pair also permits segment lengths of up to 350 m at 155 Mbit s Nine pin D sub connectors or IBM MIC connectors as used in Token Ring networks are recommended Here too the bit stream is NRZ en coded 10 1 6 4 Cell Based Physical Layer When ATM cells are transmitted over data lines as a plain bit stream using neither PDH nor SDH SONET framing this is called cell based physical layer Interface specifications exist for cell based physical layer at 51 84 Mbit s with optional partial transfer rates of 25 92 Mbit s and 12 96 Mbit s 155 Mbit s and TROUBLESHOOTING LocAL AREA NETWORKS 1 0 322 622 Mbit s The cells are transmitted in a continuous stream of ordinary ATM data cells OAM cells and idle unassigned cells Every 27th cell at most may be physical layer cell which is either an idle cell inserted when no ATM user data cells are queued for transmission or a PL OAM cell The latter type is used to carry out the operation monitoring functions that are otherwise accomplished by SDH SONET headers At
13. Coaxial Cable SDH based ATM transmission at 155 Mbit s over 75 Ohm coaxial cable is used almost exclusively in European wide area networks this interface is rarely used in North America where it is called EC 3 EC stands for Electrical Carrier TROUBLESHOOTING LocAL AREA NETWORKS 320 am 10 Transmission parameters for ATM plastic fiber interfaces POF HPCF Units Maximum spectral width FWHM 40 40 nm Numerical aperture transmitter 0 2 to 0 3 0 2 to 0 3 Mean optical power 8 to 2 20 to 14 dBm Wavelength 640 to 660 640 to 660 nm Minimum extinction rate 10 10 dB Maximum rise and fall time 10 90 4 5 4 5 ns Maximum overshoot 25 25 96 Maximum systematic jitter 1 6 1 6 ns Maximum random interface jitter 0 6 0 6 ns Reception parameters for ATM plastic fiber interfaces POF HPCF Units Minimum sensitivity 25 26 5 dBm Minimum overload 2 14 Maximum rise and fall time 10 90 5 0 6 0 ns Maximum systematic jitter 2 0 2 0 ns Minimum eye diagram aperture 1 23 1 23 ns time interval reserved for clock regeneration after electrical optical conversion Maximum random interface jitter 0 6 0 6 ns Figure 10 16 Optical transmission and reception parameters for POF and HPCF interfaces ATM over 75 Ohm Coaxial Cable The physical medium and the encoding technique are the same as those speci fied in ITU T Recommendation G 703 for the 140 Mbit s E4 interface Two 75 coaxial cables are used
14. SONET is the North American equivalent of SDH and can be considered to be identical in most respects for equivalent interface rates The cell stream is encapsulated with byte alignment in VC x or in concatenated VC xc containers synchronous payload envelopes or SPEs in SONET Because the container SPE payload is not an integer multiple of 53 bytes a cell can be split across two containers SPEs Before the ATM cells are inserted in the containers SPEs for transport the user data field of the cell is scrambled in order to facilitate delineation of the individual cell at the receiving station The scrambling method used is self synchronizing scrambling SSS with the generator polynomial 2 1 If the cell transfer rate is different from the user data bandwidth of the ATM cell 53 bytes Header 5 bytes User data 48 bytes VC 4 4c container VC 4 4c container 1040 bytes Figure 10 11 Transportation of ATM cells in the STM 4c transport module TROUBLESHOOTING LocAL AREA NETWORKS 315 ATM 10 SDH SONET containers SPEs empty idle unassigned cells are inserted ATM cells that
15. Troubleshooting by Othmar Kyas ATM An Agilent Technologies Publication Agilent Technologies To err is human but to really foul things up requires computer ANONYMOUS 10 1 ATM Specification and Implementation Since the early 1980s data communications have been divided into two separate areas with little in common local area and wide area networking For technical reasons data transport methods have been fundamentally different in these two areas In wide area networks data communication has been connection ori ented before the first bit of user data is transmitted a signaling process takes place in which a dedicated connection to a given remote station is set up Local area networks by contrast have used connectionless broadcast transmis sions every data packet is sent out over a medium shared among all stations without waiting for acknowledgment It is the receiver s job to detect packets in the data stream that have its address as the destination and process them Asynchronous Transfer Mode ATM was originally conceived for wide area data communication but was soon adapted for local area networks ATM creates a unified system that does away with the historical distinction between local and wide area data communication techniques In both ATM LANs and WANs data is transported by means of switches according to principles that had been customary only in wide area communications suc
16. a Routed non ISO 03 00000010800 Non ISO PDU CPIILNG data packet g Prene 03 0080 2 0001 0000 Ethernet CRC PAD UU CPI LNG CRC 3 Token Ring 03 0080 210003 0000 ET CRC PAD UU CPI LNG CRC 3 F AA AA 03 0080C2 0004 0000 FDDI CRC PAD UU CPI LNG CRC OUI Organizational Unique Identifier PID Protocol Identifier UU isss Transparent User to User Info CPI Common Part Identification CRC Cyclic Redundancy Check Figure 10 52 ATM LAN LLC encapsulation RFC 2684 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 376 The first method transports all LAN protocols by encapsulating the Logical Link Control LLC data packets in AAL5 CPCS PDUs The entire data stream is transmitted within one VC All protocols based on Ethernet Token Ring FDDI or Distributed Queue Dual Bus DQDB IEEE 802 6 MAN can be transported over ATM networks in this way see Figure 10 52 LLC encapsulation is used mainly in networks that only support PVCs and which are unable to manage constantly changing VCs in use In networks using SVCs which typically have no trouble dynamically managing a large number of VCs VC based multiplexing can be used This method avoids transporting the LLC header by setting up a separate virtual channel for each protocol This makes transportation significantly more efficient overall which is w
17. 10 2 6 Problems with Routers Routers are internetwork components that connect network segments at OSI Layer 3 and are therefore able to link networks of different topologies For this reason there are no troubleshooting issues for routers that apply specifi cally to ATM networks Please refer to the troubleshooting section on routers in Chapter 7 418 10 2 7 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Cause 6 Cause 7 Cause 8 Cause 9 Cause 10 Cause 11 Cause 12 Cause 13 Symptom Cause 1 Cause 2 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Cause 6 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 Symptoms Causes ATM No Connection over a PVC Problems with ATM interface card or driver No PVC set selected VCI is incorrect Hardware or software problems on the switch Misconfigured ATM port bit rate scrambling interface type frame type PLCP SDH SONET SONET No Connection over SVC UNI Signaling Problems Problems with ATM interface card or driver ATMARP server misconfigured clients are not set up with their correct ATM and IP address on the ATMARP server The address of the ATMARP server is not configured correctly on the client system ILMI is not active on the client or the server The ILMI software versions on client and server are incomp
18. Cause 7 Cause 8 Cause 9 Symptom Cause 1 Cause 2 Cause 3 Cause 4 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 The primary LANE service failed and the backup LANE service was not activated No Connection over PNNI Network Signaling problems SVC between the end systems involved Wrong route selection due to incorrect ATM addressing of the end systems Topology information on the switch port is incomplete or out dated Misconfigured peer group leader PGL not active or no designated parent LGN Hello protocol is not active on the PNNI The PNNI Routing Control Channel SVCC RCC is inactive Misconfigured PNNI port parameters cell rate cell transfer delay bit rate Uplinks to neighboring peer groups are inactive or not defined PNNI addresses prefixes or summary addresses are incorrect Loss of ATM Connections Violation of the traffic contract traffic shaping activated Cell streams with different priorities are transmitted at high load and cells with low priority are discarded Clocking and synchronization problems due to configuration er rors on the ATM port Problems on the physical layer cabling connectors ATM port Gathering Information Common Errors The first step in any troubleshooting process is to gather information In diag nosing ATM problems comprehensive information about the context of the problem provides a detailed description of the symptoms and clues to possible cau
19. STATUS ENQUIRY or STATUS is received with a call reference that does not refer to an active call or to a call in progress the receiver shall initiate clearing by sending a RELEASE COMPLETE message with cause No 81 Invalid call reference value specifying the call reference of the message received and shall remain in the null state b When a RELEASE COMPLETE message is received with a call reference that does not refer to an active call or to a call in progress no action should be taken TROUBLESHOOTING LocAL AREA NETWORKS 1 0 411 When a SETUP message is received with call reference that does not refer to an active call or to a call in progress and with a call reference flag incorrectly set to 1 the message shall be ignored d When a SETUP message is received with a call reference that does refer to an active call or to a call in progress the SETUP message shall be ignored e When any message except RESTART RESTART ACKNOWLEDGE or STATUS is received with the global call reference no action should be taken on this message but a STATUS message shall be returned using the global call reference with a call state indicating the current state associated with the global call reference and cause No 81 Invalid call reference f When a STATUS message is received specifying a call reference that is not recognized as relating to an active call or to a call in progress it shall be cleared with cause 101 Message not
20. The individual fields of performance management cells are the following Monitoring cell sequence number MCSN 8 bits Total user cell count for the CLP user cell flow TUC 16 bits This is the number of cells transmitted with cell loss priority of 0 or 1 up to the time of OAM cell insertion Total user cell number for the CLP user cell flow TUC 16 bits Identical to TUC but counting only cells with CLP 0 e Block error detection code for the CLP user cell flow BEDC 16 bits This field is used only in forward monitoring cells It contains a BIP 16 checksum of the information fields of the cells transmitted since the last forward monitoring cell Time Stamp TSTP 32 bits This is the time at which the OAM cell was inserted in the cell stream Currently the time stamp is optional its use has yet to be fully defined Total received cell count for the CLP user cell flow TRCC 16 bits This field is used only in backward reporting OAM cells It contains the number of cells received before the corresponding forward monitoring cell Total received cell count for the CLP user cell flow TRCC 16 bits This field is used only in backward reporting OAM cells It contains the number of cells with CLP 0 received before the corresponding forward monitoring cell Block error result BLER 8 bits This field is used only in backward report ing OAM cells It contains the number of incorrect pa
21. The LE Server either forwards the LE ARP request directly to the desired station or provides the sender with the desired ATM address in the form of an LE ARP response Several data VCCs may be set up between two LE Clients if both of them support this The application layer may specify QoS parameters for the connections such as ATM Traffic Descriptor Alternative ATM Traffic Descriptor only in UNI 4 0 Minimum Acceptable Traffic Descrip 1 LECS Connect Phase 2 Configuration Phase LAN Emulation Configuration n 4 8 Workstation Server LECS Bridge Configuration Direct VCC Configuration Direct VCC 3 Join Phase 4 Registration Phase LAN Emulation LAN Server LAN Emulation Emulation Client Control Direct VCC Control Direct VCC Client LEC LEC Control Distribute VCC FU LAN 5 BUS Connect Phase Broadcast and Unknown r 805 Multicast Send VCC Multicast Send VCC Multicast Forward VCC Data Direct VCC Figure 10 56 LANE functional processes TROUBLESHOOTING LocAL AREA NETWORKS 1 0 381 only UNI 4 0 Broadband Bearer Capability QoS parameters End to End Transmit Delay only in UNI 4 0 etc If no specific QoS parameters are re quested the default QoS parameters for LANE data VCCs must be used UBR or ABR QoS class 0 Because LE Clients can send data packets to
22. one for transmission and one for reception The signals are coded using Coded Mark Inversion CMI at a voltage level of 0 5 volts ATM over 100 Copper Cable Cat 3 Unshielded Twisted Pair Using a special coding technique called CAP 64 which obtains high data rates at low frequency bandwidths ATM in SDH SONET framing can be transported at a speed of 155 Mbit s even over low quality UTP 3 data cable which is very common in North America The maximum segment length is 100 m In addition to the STM 1 OC 3 rate of 155 Mbit s the ATM Forum has also specified slower rates for this cable type 51 84 Mbit s 25 92 Mbit s and 12 96 Mbit s The coding of the corresponding bit stream is CAP 16 for 51 84 Mbit s CAP 4 for 25 92 Mbit s and CAP 2 for 12 96 Mbit s CAP stands for Carrierless Amplitude Modulation Phase Modulation and is an extremely efficient method for achiev ing high data rates in spite of low available frequency bandwidth The encoding process divides the symbol stream to be transmitted into n data paths where n is the symbol period Of the resulting symbol streams one is sent through an in phase filter and the others through phase shift filters The output signal of the in phase filter is added to the inverted output signal of the phase shift filter and then sent to the twisted pair through a low pass filter In this way the informa TROUBLESHOOTING LocAL AREA NETWORKS 1 0 321 tion is coded the form of phase sh
23. with defined QoS parameters and service classes without performance param eters QoS classes with performance parameters must specify at least two such parameters If a QoS class contains two parameters for cell loss ratio then one value applies to cells with a cell loss priority CLP of 1 and the other to cells with CLP 0 Performance parameters that can be specified in a QoS class include Maximum Cell Transfer Delay Cell Delay Variation e Cell Loss Ratio for cells with CLP 0 Cell Loss Ratio for cells with CLP 1 Quality of service classes and traffic parameters are defined for each of the six service classes defined in the ATM Forum s Traffic Management specification CBR Constant Bit Rate e rt VBR Real Time Variable Bit Rate e nrt VBR Non Real Time Variable Bit Rate UBR Unspecified Bit Rate ABR Available Bit Rate GFR Guaranteed Frame Rate service class can be requested without QoS parameters such as in a request for a connection with the best possible network service for example Such a request in which no QoS parameters are specified may be made when no explicit network performance guarantee is required The load level and error frequency in ATM networks have a direct influence on the QoS parameters ATM Payload Types PT Ordinary user data cells can be distinguished from special purpose non user cells by means of the payload type field The PT values 0 1 2 and 3 identify user data ce
24. 0 is known as the global call reference and refers to all connec tions within a signaling virtual channel Message Type This field indicates the message type All message types except SETUP ACKNOWLEDGE and INFORMATION are also supported by the corresponding ATM Forum specification UNI 4 0 see Figure 10 46 TROUBLESHOOTING LocAL AREA NETwoRKS 1 0 363 Call Reference Format Bit 8 7 6 5 4 3 2 1 Byte 0 0 0 0 Length of the Call Reference 1 field in bytes Flag 2 Call Reference 3 4 Flag 1 Message sent by the station called 0 Message sent by the calling station The Global Call Reference Bit 8 7 6 5 4 3 2 1 Byte 0 0 0 0 0 0 1 1 1 0 1 0 0 0 0 0 0 0 2 Call Reference 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 4 Figure 10 45 Call reference and global call reference The Message Length Field The message length field indicates the length of the signaling message in bytes not counting the protocol discriminator call reference message type and mes sage length fields The length field itself can be 1 or 2 bytes long In the first byte only the first 7 bits can be used Bit 8 indicates whether the length field includes a second byte If bit 8 is set to 0 the length field is continued in the following byte otherwise the length field is a single byte If the message contains no other information elements after the length field the message length is coded as
25. 0000000 P2 7 11 13 57 9639613 0 5 5 CLP High PTI SDU1 HEC Good AAL 5 Type EOM Len 12 CRC32 Good SAAL STAT 5 0000001 N MR 0000100 N R P1 8 11 13 57 9639613 0 5 5 CLP High PTI SDU1 HEC Good 111 25 TunezEnM ILen zi CDo32z26aad la r Ready in 0 01 14 20 EQ 0 0 LN 0 0 Figure 10 71 UNI signaling decoded by a protocol analyzer for LANE analysis 10 2 3 3 PNNI Signaling PNNI is the signaling protocol used to set up SVC connections between two NNIs that is switch to switch signaling It actually consists of two protocols the topology protocol which distributes information about the network topology to the network nodes and the signaling protocol which is basically an extension of the ITU T Recommendation Q 2931 UNI protocol Diagnosis of PNNI problems begins with an examination of the UNI signaling functions at the end systems in the problem domain Are the interfaces of the ATM nodes and switches active Are the signaling versions of all systems compatible UNI 3 1 4 0 Are ILMI and the SSCOP layer active Ifthe end systems begin signaling processes that cannot be finished successfully the next step is to analyze the PNNI SVC routing First read out the PNNI routes determined by the UNI send port of the switch to the destination node Then check whether an active route to the destination was found This is done using the vendor s system management interface to read out the PN
26. 30s RELEASE RELEASE COMPLETE Repeat Set VC to M sent or RELEASE RELEASE maintenance state received restart T308 T309 10s SAAL SAAL active again Clear call aborted delete VCls and call references T310 10s CALL ALERT CONNECT Clear call M PROCEEDING or RELEASE received received T316 2 min RESTART RESTART Repeat Repeat M sent ACK RESTART RESTART received several times several times T317 Implemen RESTART call references Report error M tation received deleted specific but T316 T322 4s STATUS STATUS RELEASE Repeat Repeat M ENQUIRY or RELEASE STATUS ENQUIRY STATUS ENQUIRY sent COMPLETE received T398 4s DROP DROP PARTY ACK Send Timer is not restarted M ATM Forum PARTY or DROP PARTY ACK only sent RELEASE received or RELEASE T399 14s ADD PARTY ADD PARTY ACK Delete party Timer is not restarted M ATM Forum sent ADD PARTY REJECT only or RELEASE received Figure 10 50 Timers for UNI signaling processes Q 2931 and ATM Forum network system TROUBLESHOOTING LocAL AREA NETWORKS 1 0 372 Unknown and International number and the numbering plans 164 the ISDN numbering system and ISO NSAP When E 164 numbers are used the number type is International number ISO NSAP numbers are designated as Unknown 8 1 does not support the following Q 2931 signaling messages ALERTING PROGRESS SETUP ACKNOWLEDGE INFORMATION NOTIFY Due to the lack of support for the
27. 6 Lg oe 2 3 uc egg gy Onn p B N uc K1 Transport frame K1 ZO w vw plexer plexer 1 x K3 demuti M LI AM iemuti K3 s crannets Chanera I WIN 7 Packet Transfer Mode Asynchronous Time Division multiplexing Packet switched Variable packet length Packets buffered if necessary H zm ee ee eS Sa Bj EB o A Packet cell B 2 anne annel 1 gt Channel 1 5 T2 T1 i Channel 2 Bl EN Lll l Channel 2 x Aa M Channel 3 d M Channel 3 am Channel Channel 4 Channel 1 Le Channel 4 Channel 4 LEM Asynchronous Transfer Mode ATM Synchronous Time Division multiplexing Packet switched Fixed cell length Cells buffered if necessary Cells of equal length au gt PEN Bag Ne m C man pO ES A 53 byte cells B d ERN Channel 1 Channel 1 ELS Fa T2 Ti x 4 Channel 2 B Channel 2 Y j E B Channel 3 Channel 3 E E Channel Channel 4 Channel 1 ue M NM Channel 4 Channel 4 Le Figure 10 2 Principles of ATM TROUBLESHOOTING LocAL AREA NETWORKS 1 0 303 length of 53 bytes Five of the 53 bytes form the cell header which contains the Virtual Chann
28. 7 The 053 PLCP frame TROUBLESHOOTING LocAL AREA NETWORKS 1 0 311 frame mapping of ATM cells is thus two stage process that is complicated and also inefficient because there is additional overhead associ ated with the PLCP frame structure the cell payloads are not normally scrambled The more efficient direct mapping of cells to the DS3 frame some times referred to as HEC mapping makes use of the same cell delineation process used DS1 El and mapping cell payload scrambling again uses the self synchronizing scrambling with the generator polynomial 1 ATM over E4 139 264 Mbit s Similar to E3 ATM cell transport in the E4 transmission frame ITU T Recom mendation G 804 does not use the E4 frame format described in G 751 but a modified frame described in G 832 The ATM cells are inserted byte aligned in the E4 frame payload field of the 2 160 byte G 832 E4 frame All other process ing such as the scrambling of the ATM cell payload and the cell rate adaptation is performed as in the mapping of ATM cells to the E3 format 240 bytes 9 rows ATM cell Header Overhead byte 53 bytes Figure 10 8 ATM cell adaptation to the G 832 E4
29. BD ros23849089480 Summary M Detailed Hex amp Time BE _ Be Type Hello Type Hello Type Hello Type Hello Type DBSurnmary Type DBSummary Type DBSummary 165 NodalPTSESum Type DBSummary IGS NodalPTSESum Type DBSurnmary Type PTSEReq 165 ReqPTSEHead 03 26 6033198 1 03 26 6041566 03 26 6049842 03 27 6022495 1 03 27 6054062 1 03 27 6061848 03 27 6069008 03 27 6074503 1 03 27 6574641 on ooooo 45 P1 Captured on 04 16 99 at 14 03 23 8490834 Length 53 IPNNT Packet Type Hello 1 Packet Length 100 Bytes Protocol Version 1 Newest Version 1 Oldest Version 1 0x00 Flags 0x0000 Mada Ae EAA AIAAN NADER NAANA INATA and ANTIN AA ce Ready ZUtk Figure 10 72 Decoding PNNI signaling with a protocol analyzer TROUBLESHOOTING LocAL AREA NETWORKS 1 0 417 leader status is active parent logical group node LGN is configured then the following checks can be performed successively Is the Hello protocol active on the PNNI Is the PNNI Routing Control Channel SVCC RCC active Are PNNI port parameters set correctly cell rate cell delay cell rate Are the peer group uplinks configured and active Are PNNI addresses prefixes and the PNNI short form add
30. Cause Figure 10 47a Format and coding of information elements TROUBLESHOOTING LocAL AREA NETWORKS 1 0 366 Information Field Coding Byte 2 Compatibility Indicator Extension indicator bit8 Set to 1 reserved for future use Coding standard bits 7 6 00 ITU T coding 01 ISO IEC standard 10 National standard 11 Network specific standard private public Flag bit 5 0 Ignore instruction indicator 1 Obey instruction indicator Reserved field bit 4 Set to 0 reserved for future use Action indicator bits 3 2 1 000 Clear Call 001 Discard info element and proceed 010 Discard info element proceed and report status 101 Discard info element and ignore 110 Discard info element and report status Figure 10 47b Format and coding of information elements the setup request to the receiving station If the station called is able to receive the request the network sends the caller an ALERTING message If the station called accepts the call the network sends the caller a CONNECT message The caller may optionally respond with a CONNECT ACKNOWLEDGE This com pletes the connection setup and the call is then in the active state 10 1 9 3 Connection Setup at the Station Called UNI After the receiving station has been notified of an incoming call by a SETUP message from the network it performs a compatibility check This test compares the address information t
31. Configuration Discovery Target resolution Connection management Data communication MPOA obtains all its configuration information from the ELANs LE Configura tion Servers Additionally MPOA components can also be directly configured by means of the MPOA MIB All MPCs and MPSs automatically detect each other s existence by using an extended LE ARP protocol This protocol transports not only the ATM addresses of the individual stations but also information about the MPOA type MPC or MPS The target resolution process determines the route to the destination and is carried out using modified NHRP Resolution Request messages The connection management process is responsible for set ting up and operating the virtual control and data connections and the data transport process consists of transmitting the user data over the selected routes The various MPOA processes are controlled by means of eight MPOA and two NHRP control messages MPOA Resolution Request MPOA Resolution Reply MPOA Cache Imposition Request MPOA Cache Imposition Reply MPOA Egress Cache Purge Request MPOA Egress Cache Purge Reply MPOA Keep Alive MPOA Trigger NHRP Purge Request NHRP Purge Reply Route Selection in MPOA Networks In selecting routes for data packet transport MPOA makes a distinction between default routes and shortcuts Data packets initially enter the MPOA network through an MPC Because the MPC gener
32. Link Interface ODD and Data Link Provider Interface DLPI The LAN data packets themselves IEEE 802 3 Ethernet or IEEE 803 5 Token Ring are transported over AAL5 in LAN emulation frames No special LANE frame type is defined for FDDI either Ethernet or Token Ring frames can be used Using the Token Ring frame format yields better results because the MAC address format is the same as in FDDI LANE also provides quality of service functions for communication between LANE stations connected by ATM The LE Client software performs all the necessary control functions as well as the actual data communication over the ATM interface providing a standard LAN MAC interface NDIS ODI or DLPI to higher layer applications LE Clients consist of the following components System hardware PC workstation router etc Standard LAN software MAC address protocol stack drivers etc LE Client software ATM interface with ATM address TROUBLESHOOTING LocAL AREA NETWORKS 1 0 379 The LE Server module controls the emulated LAN ELAN It includes functions such as registration of LE Clients and MAC to ATM address resolution for all registered stations Every LE Client participating in an ELAN reports its LAN MAC address the corresponding ATM address and any necessary route infor mation to the LE Server When a LAN data packet needs to be sent the ATM address of the destination is first sought in the address table of the LE
33. NOTIFY message the information element Notification Indicator is also not supported This information element is used inthe Q 2931 NOTIFY message to obtain information about the connection state or more specifically the messages CALL PROCEEDING CONNECT ACKNOWL EDGE RELEASE and SETUP For point to multipoint signaling UNI 3 1 pro vides following message types not defined in Q 2931 ADD PARTY ADD PARTY ACKNOWLEDGE ADD PARTY REJECT DROP PARTY ACKNOWLEDGE UNI 3 1 supports only the standard code set 0 while Q 2931 also supports code sets 5 6 and 7 Furthermore the definitions of certain UNI 3 1 information elements are slightly modified The maximum length of the Traffic Descriptor information element in the SETUP message for example is increased to 30 bytes as opposed to 20 in Q 2931 the length of the Transit Network Selection element is limited to 8 bytes A complete list of the differences between UNI 3 1 and Q 2931 signaling is found in Appendix E of the ATM Forum UNI specifica tion The timers defined in UNI 3 1 for B ISDN signaling are the same as those in Q 2931 with the exception of T301 T302 and T304 which are not supported by UNI 3 1 The ATM Forum defines additional timers for point to multipoint processes T398 and T399 UNI 3 1 and T331 UNI 4 0 Figures 10 49 and 10 50 list the UNI protocol timers ATM Forum Signaling UNI 3 0 The most important difference between the signaling specifications
34. PDUs are encapsulated in DXI frames for transmission Mode 1b also supports AAL3 4 with a maximum SDU length of 9 224 bytes The LS DTE Service Data Unit SDU DTE SDU i y 1 H A H ATMDXIDataLinikPDU T ATMDXI __ al Data Link D m CPCS Z AAL5 CPCS PDU 5 CPCS a Trailer gt 2 9 SAR PDU o j 1 o i 5 d AALS SAR i 5 SAR i Cell Uo T H payload H ATM layer ei 5 payload Figure 10 17 Encapsulation of cells DXI PDUs Modes 1a and 1b TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 325 AAL3 4 PDUs are encapsulated by DTE but all segmentation reassem bly AAL3 4 SAR must be performed by the DCE Figure 10 17 In Mode 2 network interfaces support up to 16 777 215 virtual connections for AAL5 and AAL3 4 SDUs can be up to 65 535 bytes long The DXI frame address DFA is carried in the DXI header It is used to convey the VPI and VCI information between DTEs Data Terminal Equipment and DCEs Data Communication Equipment The DFA is 10 bits long in Modes la and 1b and 24 bits long in Mode 2 Figure 10 18 Note that no ATM cells are involved at this interface only AAL frames Fl
35. STP Cell stream FDDI multimode 100 single mode TAXI obsolete Figure 10 3 ATM transmission interfaces TROUBLESHOOTING LocAL AREA NETWORKS 1 0 307 transport containers these frames with their relatively simple structures multiplexing hierarchies or spatial limitations are actually superfluous Cell based physical layers are therefore currently only used in a few applications developed especially for this transport method such as video 10 1 6 1 ATM Data Rates At present a large number of transmission interfaces are defined both for LAN and WAN use spanning a wide range of bandwidths and transport media Data rates range from 1 544 Mbit s to 9 9 Gbit s or even higher The differences between the LAN and WAN specifications are primarily associated with the transport medium Whereas single mode fiber optic and coaxial cable are the primary media for WAN interfaces multimode fiber plastic optical fiber POF and copper twisted pair are becoming increasingly common in local area net working Figure 10 3 lists the currently available transmission interfaces 10 1 6 2 PDH Networks ATM over 051 1 544 Mbit s The mapping of ATM cells to DS1 line framing is specified in ITU T Recommen dation G 804 Under this standard ATM cells can be transported in a 24 frame multiframe Extended Superframe or ESF with the cells occupying bits 2 to 193 The individual ATM cells are byte aligned with the DS1
36. Server If it is not found there address resolution must be performed by BUS broadcasts The LE Configuration Server manages the assignments of LE Clients to various ELANS by maintaining the configuration information of the ELANs in a configu ration database An LE Client can belong to several ELANs simultaneously The BUS module retransmits all the LE Clients broadcast and multicast data packets These include All data packets with broadcast or multicast addresses Data packets sent to a MAC address for which the LE Client does not know the corresponding ATM address and which could not be resolved by the LE Server The source routing mechanism s explorer data packets used to determine optimum routes Data packets received by the BUS are retransmitted in sequence to the appropri ate group of destination LE Clients This is necessary in order to avoid overlap ping of AAL5 data packets from different senders Functional Processes in LAN Emulation Process control between the individual LE Servers takes place by means of control VCCs and data VCCs Control virtual channels are used to connect LE Clients with LE Servers and with LE Configuration Servers Communication between LE Clients as well as between BUSs and LE Clients takes place over data virtual channels data VCCs In contrast to LANE Specification 1 0 LANE v2 also supports LCC Multiplexing This means that the data streams of several ELANS and of several different pr
37. The POLL timer monitors the maximum time interval between the transmis sions of successive POLL PDUs while SD or SDP PDUs are being transmitted Error Type Error Code PDU or event that triggered the error SD PDU BGN PDU BGAK PDU BGREJ PDU END PDU ENDAK PDU POLL PDU STAT PDU Receipt of unsolicited or inappropriate PDU USTAT PDU RS RSAK PDU ER ERAK VT CC gt Timer NO RESPONSE expired SD or POLL N S error STAT N PS error STAT N R or list elements error USTAT N R or list elements error PDU length violation SD PDUs must be retransmitted Lack of credit Credit obtained Unsuccessful retransmission Other list elements error type SD loss Credit condition lt ajel Irl olnm olol Figure 10 40 SSCOP error codes TROUBLESHOOTING LocAL AREA NETWORKS 1 0 358 When SD or SDP PDUs are being transmitted the KEEP ALIVE Timer controls the interval between successive POLL PDUs The NO RESPONSE timer specifies the maximum delay between two PDUs of the type POLL or STAT The value of the NO RESPONSE timer must be greater than both that of the KEEP ALIVE timer and that of the POLL timer Furthermore the value of the NO RESPONSE timer must be more than twice the signal delay over the connection concerned The CONNECTION CONTROL timer determines the maxim
38. communication situations in ATM networks the ATM layer provides for different connection types with different characteristics Both virtual channel and virtual path connections can be struc tured as point to point or point to multipoint The bandwidth allocated to a connection can also be asymmetric that is the bandwidth for transmission can be lower than for reception or vice versa Furthermore a number of Quality of Service QoS parameters can be negotiated for each connection Virtual Channel Connections VCCs ATM virtual channels represent the lowest level in the structural hierarchy of ATM data streams All virtual channel connections have the following four properties The ATM switch assigns each virtual channel connection QoS parameters that define properties such as cell loss ratio or cell delay Virtual channel connections can be either dynamically switched SVC or semi permanent PVC The sequential order of cells in a virtual channel is preserved during trans portation through the ATM network the virtual channel is connection oriented For each virtual channel connection traffic parameters such as the maxi mum bandwidth available for the connection the result is the traffic con tract Cells sent to the network by the user are monitored to ensure con formance with the traffic contract A VCC can be set up by using four different signaling methods No signaling for example
39. compatible with call state Alterna tively any other action specific to the implementation that attempts to recover from this mismatch can be taken 5 Ifa STATUS or a STATUS ENQUIRY message is received with a call reference that does not refer to an active call or to a call in progress a STATUS ENQUIRY message shall be sent to check the correctness of the call state h When a RESTART message is received specifying the global call reference with a call reference flag incorrectly set to 1 or when a RESTART AC KNOWLEDGE message is received specifying the global call reference with a call reference flag incorrectly set to 0 no action should be taken on this message but a STATUS message shall be returned with a call state indicating the current state associated with the global call reference and cause No 81 Invalid call reference Message type or message sequence errors Whenever an unexpected message is received except RELEASE RELEASE COMPLETE or when an unrecognized message is received no state change shall occur and a STATUS message shall be returned with one of the following causes No 97 Message type non existent or not implemented b No 101 Message not compatible with call state Two exceptions to this procedure exist however The first is when the network or the user receives an unexpected RELEASE message in response to a SETUP message In this case a no STATUS or STATUS ENQUIRY message is sent Whenever the n
40. exceed the available bandwidth are discarded with low priority cells being discarded preferentially The resulting bit rate of the ATM cell stream is synchronous with that of the SDH container or SONET SPE Note that while Figure 10 10 shows cell mapping to an SDH STM 1 transport module cell mapping to a SONET STS 3c transport system is essentially identical Similarly the mapping of cells to a SONET STS 12c transport system is identical to the mapping of cells to the SDH STM 4c transport module shown in Figure 10 11 Byte Function Coding STM 1 STS 3c Section Transport Overhead A1 A2 Frame alignment C1 STM 1 identifier B1 Regenerator section error monitoring BIP 8 B2 Multiplexer section error monitoring BIP 24 H1 H2 AU 4 pointer path AlS A11 1s H3 Action pointer K2 bits 6 8 Multiplexer section AIS section FERF 111 110 Z2 bits 18 24 Multiplexer section error reporting FEBE S B2 error count VC 4 Path overhead J1 Path ID verification BIP 8 B3 Path error monitoring ATM cell C2 Path signal level B3 error count G1 bits 1 4 Path error reporting REI 1 G1 bits 5 Path RDI FFS Cell delineation supervision FFS FFS Header error performance monitoring FFS 1 Only the codes that are relevant for the monitoring function are listed 2 The use of B1 for regenerator section error monitoring is optional 3 The code for ATM cells is C2 13hex 4 The use of H1 and H2 for path AIS is provisional 5 The use of Z
41. frame but not frame aligned This means that an ATM cell can be split across two DS1 frames 193 bits 125 us Header Header Header Cell adaptation field 24 bytes channels 1 24 ATM Header Cell Transports OAM F3 information 53 bytes Reports frame and synchronization errors Transports RDI and LCD Figure 10 4 Direct cell adaptation to the DS1 frame format G 804 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 308 By inserting idle or unassigned cells when no valid ATM user data cells are available the cell transfer rate is adapted to the user data bandwidth of DS1 Scrambling of the ATM cell s data field is optional in contrast to the 1 ATM mapping The self synchronizing scrambling method is used with the generator polynomial x 1 ATM over E1 2 048 Mbit s Like 051 mapping the mapping of ATM cells to El frames is also described in ITU T Recommendation G 804 The ATM cells can be transported in bits 9 128 and 137 256 which correspond to timeslots 1 through 15 and 17 through 31 The cell rate is adapted to the El frame payload bandwidth of 1 920 Mbit s by means of idle cells when no ATM cells are queued for insertion Once again the ATM cells byte aligned with the E1 frame but not frame aligned The 48 data bytes of each ATM cell are scrambled before transport using self synchronizing scrambling SSS
42. frames in the LANs at data rates of 155 Mbit s and 622 Mbit s the ATM Forum has also specified multimode fiber in place of the single mode fiber prescribed in the ITU T recommendations The maximum range of 2 km is lower than the range for single mode fiber but this is seldom a problem in the LAN field especially in indoor cabling structures with typical segment lengths of 100 to 1 000 m The cost of the communication infrastructure can be significantly lowered however by avoiding expensive single mode fiber and the high quality laser sources it requires In addition to ATM cell transmission over single mode SDH SONET WAN interfaces multi mode fiber also makes it possible to implement the STM 4c OC 12 622 Mbit s STM 1 0C 3 155 Mbit s and STM 0 STS 1 51 84 Mbit s interfaces in LANs The 5 1 0 3 multimode fiber optic interface uses a wavelength of 1 300 nm over a 62 5 125 um multimode fiber with a modal bandwidth of 500 MHz km 50 125 um is also permissible The data stream is 8B 10B encoded for transmis sion so that the physical medium s 194 4 Mbaud yields an effective data rate of 155 52 Mbit s Either SC or BFOC 2 5 868 connectors may be used This specification originally defined for UNI 3 0 has been rarely implemented and is described here for the sake of completeness because it is still contained in the UNI 3 1 specification Figure 10 14 lists the parameters for the multimode fiber optic interface The multimode
43. in which one or more mandatory information elements has invalid contents a RELEASE COMPLETE message is returned with cause No 100 Invalid information element contents Unrecognized Information Element If a message is received that contains one or more unknown information ele ments action is taken on the message and those information elements that are recognized and have valid contents If the received message is not RELEASE or RELEASE COMPLETE a STATUS message is returned containing one cause information element The information element contains cause No 99 Informa tion element non existent or not implemented and the diagnostic field if present contains the information element identifier of each unrecognized infor mation element When a RELEASE message is received that has one or more unrecognized information elements a RELEASE COMPLETE message with cause No 99 Information element non existent or not implemented is returned The cause information element diagnostic field if present contains the information ele ment identifier of each unrecognized information element A RELEASE COM PLETE message with unknown information elements is ignored completely If a message contains one or more information elements with contents that are in part invalid then action is taken on those information elements that appear correctly A STATUS message is also sent with cause No 100 Invalid informa tion element contents and t
44. least one PL OAM cell is required for every 513 cells ATM Cell Streams over V 35 EIA TIA 449 530 HSSI and E1 The ATM Forum has specified a cell based transmission convergence sublayer based on ITU T Recommendation 1 432 for clear channel interfaces This term designates all interfaces that are capable of transporting any data stream with out imposing bit stream encoding and framing restrictions Examples include V 35 EIA TIA 449 530 EIA TIA 612 613 High Speed Serial Interface or HSSI and unframed E1 Any other clear channel interface can also be used however The cells are transferred in a continuous stream of ordinary ATM cells OAM cells and idle unassigned cells No F1 or F2 OAM functions are specified for monitoring the network F3 OAM functions can be optionally implemented using special physical layer OAM cells to monitor processes at the level of transmis sion paths In this case the following parameters can be analyzed The number of included cells NIC per OAM cell 128 The number of cells for which transmission error values are calculated Monitoring Block Size or 5 16 The number of blocks monitored per OAM cell 8 The number of monitored blocks received by the remote station 8 The cell rate is decoupled from an interface s data rate by inserting idle unassigned cells The remote station is synchronized with the individual ATM cells using the Header Error Control HEC mechanism described in the next
45. low bandwidth delay sensitive and digital data communication often higher bandwidth and almost always delay insensi tive A cell length of 53 bytes is short enough to be well suited to the transmis sion of delay sensitive low bandwidth digital signals in ATM s higher megabit s and gigabit s throughput ranges The small cell size also permits quick and exact bandwidth allocations as well as parallel cell processing at high speed in ATM switches 10 1 5 The ATM Layer Model The transmission procedure commonly referred to as the ATM protocol actually consists of a number of protocol layers that build on one another and map the various network services to ATM cells Taken as a whole this set of protocol layers is called the B ISDN protocol reference model This name is a reminder of the origins of the ATM specification which has its roots in the ITU T s Broad band ISDN project for a universal wide area network Since ATM technology in the data communications field has spread beyond the boundaries of conven tional telecommunications ATM has become accepted as the universal name for the B ISDN layers as a whole although in a strict sense ATM refers only to cell based data transport that is the ATM layer within the B ISDN model TROUBLESHOOTING LocAL AREA NETWORKS 1 0 304 The logical B ISDN network architecture was designed for four independent levels of communication after the ISO OSI reference model ITU T Recommen dation
46. low priority In case of network conges tion these cells will then be among the first to be discarded If cell tagging is not supported then cells that do not conform to the traffic contract are discarded immediately Traffic Shaping The traffic contract negotiated during the connection setup includes a Connec tion Traffic Descriptor that defines the parameters for permissible traffic including peak cell rate duration of the peak cell rate maximum burst size MBS etc A transmitting station can then force its communication to conform to the negotiated traffic contract by means of the optional traffic shaping function Another user strategy would be to send all queued cells as they occur and simply tolerate tagging of high priority cells if supported and discarding of low priority cells TROUBLESHOOTING LocAL AREA NETWORKS 1 0 341 10 1 7 4 Monitoring in ATM Networks OAM Flows Of the five ATM network operations and maintenance information flows F1 through F5 F4 and 5 are situated in the ATM layer F4 and F5 OAM information is gathered and transmitted by means of special OAM cells The F4 information flow is used for segment or end to end management at the virtual path level A segment refers to one section of a connection such as the link between two ATM switches more than one switch may exist within a section end to end refers to the entire communication path between the two endpoints for the complete virtual
47. mpoa 0087 000 July 1997 af mpoa 0092 000 July 1998 af mpoa 0114 000 May 1999 af mpoa 0092 000 July 1998 388 Technical Working Group Physical Layer TROUBLESHOOTING LocAL AREA NETWORKS Approved Specifications Utopia Mid Range Physical Layer Specification for Category 3 UTP E3 UNI Physical Interface Specification for 25 6 Mb s over Twisted Pair A Cell Based Transmission Conver gence Sublayer for Clear Channel Interfaces 622 08 Mbps Physical Layer 155 52 Mbps Physical Layer Specifi cation for Category 3 UTP See also UNI 3 1 af uni 0010 002 120 Ohm Addendum to ATM PMD Interface Spec for 155 Mbps over TP DS3 Physical Layer Interface Spec 155 Mbps over MMF Short Wave Length Lasers Addendum to UNI 3 1 WIRE PMD to TC layers E 1 Physical Layer Interface Specification 155 Mbps over Plastic Optical Fiber POF Version 1 0 155 Mb s Plastic Optical Fiber and Hard Polymer Clad Fiber PMD Specification Version 1 1 Inverse ATM Mux Version 1 0 Inverse Multiplexing for ATM IMA Specification Version 1 1 Physical Layer High Density Glass Optical Fiber Annex 622 and 2488 Mbit s Cell Based Physical Layer ATM on Fractional E1 T1 2 4 Gbps Physical Layer Specification Physical Layer Control Utopia 3 Physical Layer Interface Multiplexed Status Mode MSM3 am 10 Specification af phy 0017 000 af phy 0018 000 af phy 0034 000 af phy 0040 000 af phy 0043 000 af phy 004
48. number of cells lost divided by the total number of cells transmitted Lost cells and transmitted cells in severely errored cell blocks are excluded from the calculation of cell loss ratio There are three different cell loss ratio measurements a Cell loss ratio for cells with high priority cell loss priority bit 0 CLR If N O is the number of cells with CLP 0 and 0 is the number of lost cells plus the number of tagged cells then CLR is defined as N 0 N 0 b Cell loss ratio for the entire cell stream CLR If N O 1 is the number of all cells transmitted and N 0 1 is the number of lost cells CLR is defined as 1 0 1 c Cellloss ratio for cells with low priority CLR If N 1 is the number of cells with CLP 1 and the number of lost cells CLR is defined as N 1 N 1 Internet Advisor ATM Run Time Cell Loss Delay Statistics e File Run View GoTo Setup Window Help min Es lt gt HE Time Cell Loss Cell Delay CBR 1 UBR DBR CLP 0 1 QoS Contract Statistics Ceu Loss Ratio 0 00000E 000 GellMisinsert Rate eps 0 00000E 000 S E C B Ratio 0 00000E 000 _____ 00 i Cell Delay Variation 00 01 37 20 2 640006 006 sax eee 10 09006 006 Cell Statistics 20500 0 20500 41000 61500 nanoseconds 0 41000
49. over UTP Hardware or software problems on the switch High signal transit delay due to long transmission path e ILMI not active on the client or on the ATM switch ncompatible ILMI software versions on client and server ncorrect port configuration bit rate scrambling interface type frame type PLCP G 804 SDH SONET ncorrect router configuration port inactive wrong operating mode protocol not active ncorrect router filters Insufficient buffering in the switch LANE IP addresses and subnet masks are incorrect interfaces belong to different subnets LANE IP interfaces on the LE clients are not active or not functioning LANE LANE software on the client or switch is not active LANE LANE ARP entries are incorrect MAC ATM address resolution is not working LANE LE Clients are not registered on the same LE Server BUS LANE LE Clients trying to communicate do not belong to the same ELAN LANE The primary LANE service failed and the backup LANE service was not activated LANE The traffic contracts of the LE Clients are incompatible Figure 10 73a The most frequent causes of trouble in ATM networks TROUBLESHOOTING LocAL AREA NETWORKS 1 0 422 LANE The LE server LES VCC is inactive the ATM address of the LES is incorrect LANE The BUS VCC is inactive or the ATM address of the BUS is incorrect Loose or defective connectors on interface card
50. parameters If the network is able to provide the service requested and if the specified ATM virtual path and virtual channel are avail able the network answers with a CALL PROCEEDING message and forwards TROUBLESHOOTING LocAL AREA NETWORKS 1 0 365 Bit 8 7 6 5 4 3 2 1 Byte Information Element Identifier 1 1 Coding IE Instruction Field 2 Ext Standard Flag Res Action Indicator 3 Information Element Length 4 Information Element Contents 5 Information Element Identifiers Bits 87654321 01110000 Called Party Number 01110001 Called Party Sub Address 01111000 Transit Network Selection 01111001 Restart Indicator 01111100 Narrow Band Low Layer Compatibility 01111101 Narrow Band High Layer Compatibility 01100000 Broadband Locking Shift 01100001 Broadband Non Locking Shift 01100010 Broadband Sending Complete 01100011 Broadband Repeat Indicator 01101100 Calling Party Number 01101101 Calling Party Sub Address 01011000 ATM Adaption Layer Parameters 01011001 ATM Traffic Descriptor 01011010 Connection Identifier 01011011 OAM Traffic Descriptor 01011100 Quality of Service Parameter 01011110 Broadband Bearer Capability 01011111 Broadband Low Layer Information B LLI 01011101 Broadband High Layer Information B HLI 01000010 End to End Transit Delay 00100111 Notification Indicator 00010100 CallState 00011110 Progress Indicator 00000100 Narrowband Bearer Capability 00001000
51. received before T322 expires the STATUS ENQUIRY can be repeated one or more times depending on the implementation If the timer expires after the last attempt the connection is cleared down with cause 41 Temporary failure Procedure on Receipt of a STATUS Message When STATUS message is received that indicates that the peer station is in an incompatible state for call handling the connection can be cleared down with cause 101 Message not compatible with call state or if so implemented an attempt may be made to correct the fault The decision as to whether the two stations call states are incompatible with one another is left to the given implementation except in the following three cases a If a STATUS message is received signaling that the peer station is in a state other than null and the station receiving the STATUS message is in the null state itself then the receiver responds with a RELEASE message and cause 101 Message not compatible with call state b If a STATUS message is received signaling that the peer station is in a state other than null and the station receiving the STATUS message is in the Release Request state the receiver shall not respond c If a STATUS message is received signaling that the peer station is in the null state and the station receiving the STATUS message is not in the null state then the receiver of the STATUS message shall change to the null state If a STATUS message is rec
52. section Cells may be transmitted either in scrambled or in unscrambled form 10 1 6 5 ATM Cell Streams over FDDI Infrastructures TAXI TAXI is a special variant of cell based ATM transmission is TAXI The TAXI interface was defined to support the use of existing FDDI infrastructures to transport ATM cells In this way FDDI rings can be converted into ATM networks while conserving most of the existing FDDI hardware The name TAXI origi nated with the first commercially available chipset for FDDI based ATM The fiber optic media and signal characteristics specified for TAXI communication are exactly the same as those defined in the FDDI standard ISO 9314 3 The ATM cells are 4B 5B encoded and transmitted without any additional framing In 4B 5B encoding 4 bits of data are transmitted as symbols of 5 bits This is due to the requirement that no more than three 1 bits occur in a row The bit sequence 1111 is coded as 11101 for example Of the 32 possible 5 bit symbols only 16 are TROUBLESHOOTING LocAL AREA NETWORKS 1 0 323 used to transmit data byte is represented by a symbol pair Some of the remaining 16 symbols are used as line state or control symbols in FDDI and some others are unusable as they have long runs of 15 or Os The symbol pair sequence JK for example announces the beginning of an FDDI frame the I or Idle symbol is used as a continuous padding bit stream for clock synchroniza tion The beginning of an ATM cell i
53. the same destination by different paths via the BUS or over different data VCCs for example the Flush Message protocol is used to preserve the order of the LAN data packets during transmis sion over the ELAN The Flush Message protocol in effect deletes the old transmission path when communication is moved to a new path The LE Client sends an LE Flush message over the given VCC to ensure that all data packets sent over that virtual channel have arrived at the receiving station and that no more packets are forthcoming in the opposite direction On receiving the Flush response the LE Client knows that no more data will be sent to it over the given VCC and can begin using the new VCC All LANE components must support one of the following maximum sizes for AAL5 5005 1 516 1 580 4 544 9 234 or 18 190 bytes for non multiplexed data packets and 1 528 1 592 4 556 9 246 or 18 202 bytes for multiplex LLC data The SDU length of 1 516 is due to the fact that LANE data packets contain the 2 byte LAN emulation header LEH but not the 4 byte checksum For Token Ring maximum packet sizes of 4 450 bytes for 4 Mbit s and 18 200 bytes for DA SA Ethernet payload FC DA SA RIF Token Ring payload LANE header LANE payload AAL5 payload AALS trailer ATM ATM ATM header ATM payload header ATM payload header ATM payload
54. with the generator polynomial x 1 This per mits fast cell delineation allowing the receiving station to recover quickly from a loss of cell delineation due to physical layer bit errors for example Timeslot 0 Timeslot 16 reserved for signaling 256 bits 125 us _ ls Header Header Header Header ATM cell adaptation field 30 bytes timeslots 1 15 and 17 31 ATM Header cell 53 bytes Transports OAM F3 information Reports frame and synchronization errors Transports RDI and LCD Figure 10 5 ATM cell adaptation to the E1 frame format ATM over E3 34 368 Mbit s The mapping of cells to the E3 frame format is described in ITU T Recommenda tion G 804 The E3 frame format used is not the E3 frame format described in TROUBLESHOOTING LocAL AREA NETWORKS 1 0 309 4 751 however but a modified frame format described 8 832 The mapping of ATM cells to the older G 751 frame structure is difficult to accomplish the cells would have to be nibble aligned because each G 751 subframe is an integer multiple of 4 rather than 8 bits The newer G 832 frame consists of 537 bytes of which 7 bytes are used for overhead information see Figure 10 6 The remain ing 530 user data bytes correspond exactly to the length of ten ATM cells so that these can be byte and frame aligned though the latter is not required The ATM c
55. 2 Mbit s 34 360 Mbit s 97 728 Mbit s and 139 264 Mbit s Cell transport over the T3 line type 44 736 Mbit s common in North America was once defined only using the T3 PLCP frame format Bellcore TR TSV 000772 TR TSV 000773 originally developed for metropolitan area networks However a more efficient direct mapping was introduced a few years ago and is now the preferred mapping even though the PLCP mapping is still very common in older equipment In the cell based physical layer the cells are not inserted in an additional transport frame but simply converted bit for bit into the given communication medium s electrical or optical signals and transmitted A different scrambling arrangement is performed from that used in other interfaces in order to provide sufficient clock transitions so that the clock timing can be recovered from the incoming bit stream this distributed sample scrambler DSS uses 31 order polynomial x x 1 and is described in 1 432 1 One advantage of a cell TROUBLESHOOTING LocAL AREA NETWORKS 1 0 306 based physical layer is the efficient utilization of bandwidth Because the ATM cell is not inserted in another frame structure with its own overhead the ratio of total overhead to user data is maintained at about 1 9 5 bytes of header 48 bytes of data The disadvantage is that existing transport infrastructures can no longer be used Another problem with direct cell transfer in wide area n
56. 2 for multiplexer section error monitoring is provisional Figure 10 12 SDH SONET overhead bytes in ATM cell transport TROUBLESHOOTING LocAL AREA NETWORKS 1 0 316 Alarm and Monitoring Signals Two types of monitoring signal are defined for ATM cell transport over SDH SONET networks the Alarm Indication Signal AIS and the Remote Defect Indicator RDI formerly known as FERF for Far End Receive Failure signal An AIS is sent downstream to report an error An RDI signal is sent upstream to report a receiving or transmission error Both kinds of monitoring signal can be implemented using the Section Transport Overhead SOH TOH bytes of the STM 1 STS 3c frame with the Path Overhead POH of the VC4 container SPE Figure 10 12 shows the use of the corresponding SDH overhead bytes for ATM cell transport the equivalent SONET overhead bytes are similar ATM over Single Mode Fiber ITU T Recommendation G 957 defines six different single mode fiber optic types that can be used to transport SDH SONET in three different communication scenarios in house links medium range WAN links and long range WAN links In house links can attain a maximum range of 2 km The optical transmitters may consist of light emitting diodes LEDs or multilongitudinal mode MLM lasers with a wavelength of 1 310 nm The permissible loss is between 0 and 7 dB Medium range WAN links span distances of up to 15 km The optical transmit ters may be e
57. 6 000 af phy 0047 000 af phy 0053 000 af phy 0054 000 af phy 0062 000 af phy 0063 000 af phy 0064 000 af phy 0079 000 af phy 0079 001 af phy 0086 000 af phy 0086 001 af phy 0110 000 af phy 0128 000 af phy 0130 000 af phy 0133 000 af phy 0134 000 af phy 0136 000 af phy 0142 000 Approved Date Mar 1994 Sep 1994 Aug 1995 Nov 1995 Jan 1996 Jan 1996 Nov 1995 Jan 1996 Mar 1996 July 1996 July 1996 Sep 1996 May 1997 Jan 1999 July 1997 Mar 1999 Feb 1999 July 1999 Oct 1999 Oct 1999 Oct 1999 Nov 1999 Mar 2000 TROUBLESHOOTING LocAL AREA NETWORKS am 10 389 Technical Working Group Service Aspects and Applications Management Voice amp Teleph ony over ATM Approved Specifications Frame Based ATM Interface Level 3 UTOPIA Level 4 PNNI ABR Addendum PNNI v1 0 Errata and PICs PNNI 1 0 Addendum soft PVC MIB Audio Visual Multimedia Services Video on Demand v1 0 Audio Visual Multimedia Services Video on Demand v1 1 ATM Names Service FUNI 2 0 Native ATM Services DLPI Addendum Version 1 0 H 323 Media Transport over ATM Traffic Management 4 1 DBCES Dynamic Bandwidth Utiliza tion in 64 KBPS Time Slot Trunking Over ATM Using CES ATM Trunking Using AAL1 for Narrow Band Services v1 0 ATM Trunking Using AAL2 for Narrowband Services Low Speed Circuit Emulation Service ICS for ATM Trunking U
58. AAL Type 2 349 AAL Type 3 4 353 AAL Type 5 353 AAL2 error messages 351 Alarm Indication Signal AIS 316 Analyzing physical layer OAM information flows 397 Asynchronous Transfer Mode ATM 297 301 ATM Adaptation Layer AAL 346 ATM addressing 367 ATM and DXI interfaces 323 ATM cell 328 cell streams at 25 6 Mbit s 323 ATM cell streams over FDDI infrastructures TAXI 322 ATM cell streams over V 35 EIA TIA 449 530 HSSI 322 ATM cell types 335 ATM data rates 307 ATM forum signaling UNI 3 0 372 ATM forum signaling UNI 4 0 373 ATM forum UNI signaling UNI 3 0 3 1 4 0 369 ATM in heterogeneous LAN environments 299 ATM in homogeneous private networks 298 ATM in PDH networks 307 ATM in public Wide Area Networks 301 ATM in SDH and SONET networks 314 ATM interface cards 417 ATM interworking 375 ATM layer 328 ATM layer model 303 ATM layer OAM functions 341 ATM over 100 Ohm copper cable 320 321 ATM over 150 Ohm Shielded Twisted Pair STP 321 ATM over 6 312 Mbit s and 97 728 Mbit s 311 ATM over 75 Ohm coaxial cable 319 ATM over 053 44 736 Mbit s 309 ATM over E1 2 048 Mbit s 308 ATM over E3 34 368 Mbit s 308 TROUBLESHOOTING LocAL AREA NETwoRKS 1 0 2 ATM over 4 139 264 Mbit s 311 ATM over multimode fiber 318 ATM over Plastic Optical Fiber 318 ATM over single mode fiber 316 ATM performance parameters 401 ATM physical layer 305 ATM routing label field VPI VCI 329 ATM si
59. AL2 guarantees the QoS param eters for each connection using the QoS mechanisms of the underlying ATM layer AAL2 itself only specifies the format of the short AAL2 SDUS mini cells optimized for real time applications and their transport within ATM cells The bi directional AAL2 connections can be set up either as PVCs or as SVOs but provide only non guaranteed transport services Corrupt or lost CPS SDUS are neither corrected nor repeated The data to be transported is first filled into Common Part Sublayer CPS packets which consist of 3 header bytes and 1 to 45 or 64 bytes of user data AAL2 connection setup the assignment of Channel IDentifiers CIDs and the negotiation of CPS service parameters takes place on AAL2 channel 1 using the ANP protocol AAL2 Negotiation Procedure TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 350 AAL SDU Service Specific Convergence SSCS PDU SSCS PDU 9909 PDU Sublayer header ayload trailer SSCS if present pay if present SSCS PDU CPS SDU CPS packet CPS packet header payload CPS PH CPS PP Common Part Sublayer CPS packet CPS Start Field CPS payload STF CPS PDU gt ry ATM SDU ATM Layer ATM cell Cell payload header ATM cell gt Figure 10 34 Structure of AAL2 The CPS packets mini cells are then inserted in CPS PDUs which consist of 1 head
60. AN emulation service behaves the same as a conventional MAC LAN driver The ATM Forum s LAN Emulation Version 2 is defined in the two documents LANE LUNI and LANE LNNI and is based on five ATM service modules that build on AALS LAN Emulation Client LE Client or LEC LAN Emulation Server LE Server or LES LAN Emulation Configuration Server LE Configuration Server or LECS Broadcast and Unknown Server BUS Selected Multicast Server SMS LUNI describes the processes at the interface to the LE Client while LNNI specifies the protocols for networking LE Client LE Server BUS and SMS components The use of LNNI components permits the coordination of several LE Servers LE Configuration Servers and BUS systems so that redundant LANE TROUBLESHOOTING LocAL AREA NETWORKS 1 0 378 structures can be built LE Clients can also be grouped in subnetworks regard less of their location or the LE Server or BUS used which makes it possible to build virtual Emulated LANs ELANSs LNNI Protocol LUNI Protocol LEC 1 LEC 2 LEC 3 LEC 4 Figure 10 55 LAN emulation LUNI and LNNI With regard to the protocol stacks of end systems LANE emulates the most widely used protocol driver specifications Network Driver Interface Specifica tion NDIS Open Data
61. ARTY ACK or DROP PARTY ACK only sen RELEASE received or RELEASE COMPLETE T399 14s ADD PARTY ADD PARTY Delete party Timer is not restarted M ATM Forum UNI 3 0 3 1 sen ADD PARTY REJECT only or RELEASE 34 124 s received UNI 4 0 T331 60s LEAF SETUP SETUP Repeat LEAF Delete connection M ATM REQUEST ADD PARTY SETUP REQUEST Forum 4 0 sent or LEAF SETUP and restart T331 only FAILURE received Figure 10 49 Timers for UNI signaling processes Q 2931 and ATM Forum user system TROUBLESHOOTING LocAL AREA NETWORKS 371 ATM Timer Default S S On First On Second Impl Number Value tart top Expiration Expiration implementation T301 23min ALERT CONNECT Clear call M Q 2931 received received if symmetrical only connections supported T302 10 15s SETUP SENDING COMPLETE if information in M Q 2931 sent indication complete clear call if overlap sending only restart when Otherwise CALL and receiving INFO sent PROCEEDING supported T303 4s SETUP CONNECT CALL Repeat SETUP Abort call set up M sent PROCEEDING ALERT restart T 303 SETUP ACK RELEASE COMPLETE received T304 20s SETUP ACK INFO sent or Clear call Q 2931 received CALL PROCEEDING only ALERT CONNECT received T306 20s RELEASE RELEASE Stop ringing M Q 2931 with Progress COMPLETE if inband alarms only Indicator 8 received supported sent T308
62. EA NETWORKS 1 0 347 stant bit rates AAL 2 for real time sensitive services with variable bit rates AAL Types 3 and 4 for connection oriented and non connection oriented transmission of non real time sensitive data Later AAL Type 5 was defined by the ATM Forum as a simplified version of AAL3 and was quickly adopted by the ITU T Early on the distinction between connection oriented and non connec tion oriented data communication in the AAL was found to be unnecessary and AAL Types 3 and 4 were accordingly merged into AAL Type 3 4 Note that AAL types are never mixed within a single virtual channel for reasons that will become obvious Service Class A Class B Class C Class D Parameter lime required not required Compensation q Bit rate constant variable Communication 3 mode connection oriented connectionless wireless connection connectionless Example emulation voice oriented data data communication communication communication AAL type AAL1 AAL2 AAL3 AAL5 AAL4 Figure 10 32 Service classes and AAL types 10 1 8 1 AAL Type 0 The AAL Type 0 indicates the absence of any AAL capability The application data is inserted directly in the payload fields of the ATM cells and transmitted Strictly speaking AALO is thus not an AAL type at all the communication mechanisms are already cell based and the adaptation layer functions don t exist 10 1 8 2 AAL Type 1 The
63. ICP 313 ISDN telephony numbering plan E 164 368 ISO NSAP addressing ISO 8348 AD2 368 L LAN Emulation LANE 377 408 LAN Emulation Client LE Client or LEC 377 LAN Emulation Configuration Server 377 LAN Emulation Server LE Server or LES 377 LANE 299 LLC encapsulation RFC 1483 375 Loopback cells 344 Loss of ATM connections 420 Loss of cell delineation rate 399 Loss of Continuity LOC 343 M MARS Multicast Address Resolution Protocol 382 Mean loss of delineation duration 399 Multi Longitudinal Mode MLM lasers 316 Multiprotocol over ATM MPOA 299 382 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 Index p4 ATM N Next Hop Resolution Protocol NHRP 385 Nibble aligned 309 NNI header 334 NNI signaling 373 No connection over a PVC 418 No connection over Emulated LAN ELAN 419 No connection over PNNI network 420 No connection over SVC UNI signaling problems 418 0 191 test cells 402 OAM cell format 341 OAM F1 F3 for cell based physical layer 326 OAM F1 F3 for PDH based ATM systems 326 OAM 1 F3 SDH SONET based ATM systems 325 OAM fault management AIS RDI 343 OAM performance management 344 OAM F1 cell fields 327 cell fields 327 One point cell delay variation 405 Payload Type field 330 Physical Layer Convergence Protocol PLCP 309 Physical Layer OAM cells 335 Physical Layer Operation and Maintenance PL OAM 306 Physical Layer Operation
64. ID All loopback cells other than end system to end system loopback cells are segment cells the point at which they loop back is determined by the Loopback Location ID field If a station receives a loopback OAM cell with a value other than 0 in the Loopback Indication field it must return the cell within one second Stations that transmit loopback cells must do so at such an interval that the management cell traffic is less than one percent of the capacity of each virtual channel or virtual path involved in the connection It should be obvious from the above that only the loopback OAM cells are looped related user cells are unaffected so loopback test can be performed safely at any time OAM Performance Management In addition to faults the performance of the individual VPs and VCs can also be monitored by the periodical insertion of special performance management OAM cells in the user cell stream Analysis of the special measurement data contained in these cells cell sequence number total user cell count time stamp cell loss TROUBLESHOOTING LocAL AREA NETWORKS 1 0 345 count yields direct information about the operating condition of the given ATM connection OAM performance management involves two distinct functions known as forward performance monitoring and backward reporting If both functions are activated then information is transmitted in both directions when OAM performance information about a given cell block is determined
65. M UNI signaling problems are the following basic checks Are the interfaces of the affected systems active Are the affected systems using compatible signaling variants for example UNI 3 1 4 0 Are ILMI if supported and the SSCOP layer active If no error is detected the next step is to try to set up an SVC connection while using the debug mode of the ATM nodes or a protocol analyzer to trace the signaling process Such an analysis of the signaling process usually leads to the cause of the problem Typical causes are invalid called party or calling party addresses invalid unknown or disordered mandatory information elements invalid call reference numbers or rejection of the connection setup by a RE LEASE message from the destination station The most important error states in UNI 3 1 4 0 Q 2931 signaling that can occur during the connection setup and clear down processes are described in the following sections Invalid protocol discriminator Messages with an invalid protocol discriminator are discarded Short messages Messages that are too short to contain a complete information element are discarded Invalid call reference format If bytes 1 and 5 through 8 of the call reference information element are not set to the value 0 or if the call reference length field contains a value other than 3 the message is discarded Invalid call reference a Whenever any message except SETUP RELEASE RELEASE COMPLETE
66. NI 4 0 signaling protocols gener ally are compatible with equipment based on ITU T recommendations 10 1 9 6 NNI Signaling For connection setup between two NNIs the ITU T developed the broadband protocol B ISUP ITU T Recommendations Q 2761 Q 2764 modeled after the narrow band ISDN protocol ISUP B ISUP contains many of the mechanisms and functions of the proven ISUP Only the characteristic parameters and processes of broadband networks with their virtual connections are added By the same token ISUP functions that are only practical in conventional connection ori ented networks are omitted While B ISUP is used for NNI signaling in wide area networks NNIs in private local ATM networks use the ATM Forum PNNI proto col Unlike B ISUP PNNI is able to select routes for the desired connections a function that has to be implemented completely independently by manufactur ers of B ISUP switching systems In local ATM networks the ATM network can be made operational immediately based on PNNI whereas large public ATM WAN networks with B ISUP signaling first require the definition and implemen tation of route selection and bandwidth monitoring functions Although de signed for private networks PNNI is also being used in some large public networks particularly in the United States B ISUP Signaling The B ISUP protocol is the NNI counterpart to the UNI protocol in ITU T Recommendation Q 2931 or ATM Forum UNI 4 0 and extends virtual UNI c
67. NI operations log TROUBLESHOOTING LocAL AREA NETWORKS 1 0 416 and examine the PNNI routes calculated by the sending node If the route leads to a wrong node a wrong ATM address is probably involved If the switch cannot find a route to the destination at all its topology information may be incorrect If the topology information does not contain the destination node this may be due either to physical problems in the network or to the switch s inability to update its topology database If the sending node does have an uninterrupted route to the destination system but the connectivity problem remains the trouble may be caused by traffic parameters that cannot be provided by one of the ATM switches If the topology information is wrong or incomplete the first troubleshooting step is to list all peer neighbors that can be reached directly from the UNI port and their PNNI operation states up down changing loading Then the topol ogy information PNNI Topology State Elements or PTSE of the peer nodes at the lowest hierarchical level are analyzed to determine whether they reach the peer group leader Conversely the PTSEs of the peer group leaders must also be examined to verify that they reach the peer nodes at the lowest level If all the PTSEs are in order and no configuration errors can be detected peer group Internet Advisor newtest2 dat Decode 5 x File Run View GoTo Setup Window Help xl
68. OOTING LocAL AREA NETWORKS 1 0 369 Connection Clear Down by the Network The network can initiate a connection clear down by sending a RELEASE command and starting its timer T308 The user responds with a RELEASE COMPLETE whereupon the network stops T308 clears down the virtual chan nels and releases the call reference The connection clear down is then complete If timer T308 expires before a RELEASE COMPLETE is received from the user the RELEASE command is repeated and T308 restarted If T308 expires a second time before a RELEASE COMPLETE is received the network marks the ATM virtual channel as out of order and clears the call reference 10 1 9 5 Forum UNI Signaling 3 0 3 1 4 0 In order to allow for manufacturers to develop ATM components before the definitive adoption of international standards by the ITU T the ATM Forum also developed specifications for ATM signaling in its UNI 3 0 document The signal ing specifications in UNI 3 0 were published a year before ITU T Recommenda tion Q 2931 and are largely incompatible with it because of the use of a different SSCOP layer as mentioned earlier UNI 3 1 however drafted after work on Q 2931 had been completed corrected this by specifying the ITU T Q 2110 SSCOP the higher layer signaling messages were largely unchanged and repre sented a significant step toward harmonization Next the ITU T introduced Q 2971 which added route initiated join point to multi poin
69. Type 1 ATM Adaptation Layer ITU T Recommendation I 363 1 serves to transport data streams with constant bit rates these include all the interfaces in the PDH hierarchy T1 1 T3 etc and provide them to the destination node in synchronization with the original transmission clock This requires that the am 10 TROUBLESHOOTING LocAL AREA NETWORKS 348 uonoejoud jequinu eouenbeg dNS play Jequinu eouenbeg 7 NS uoneoynuepi Je e qns eousBueAuo2 777 159 seq gr 8 Ploy SO Jo play oyu 1189 WLY WLY 234 gt 5 y 9 pue y Z 0 NS NLY prey 6 21 Nd HYS au JO uogeuuojul 2 y SI 9 U SO 941 WLY SLY ay urejuoo pue G NS 159 941 e Tif 7 peuounddsvs 2 uaaviang 1 75 ry Sq NVS seq Ly 14 7 eccccce peuogunadsvs 2 2 19 Bu Syd Ly qag u3AV nad HVS 2 218 Ja gt VAS nas Se uvs S y
70. Type field Figure 10 63 Format of the O 191 test cell Out of service test cells have standard ATM cell headers and can be sent using any VPI VCI label value VCI gt 31 In the payload they carry a 32 bit sequence number to permit detection of cell loss and cell misinsertion errors and a 32 bit time stamp to measure cell delay and cell delay variation This allows CDV and 2 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 402 point CDV measurements up to transmission speeds of 2 4 Gbit s The least significant bit of the time stamp has a granularity of 10 ns though for physical links slower than 2 4 Gbit s the time stamp is normally incremented from a higher order bit In the simplest case a protocol analyzer with one transmit and one receive port is sufficient to perform these out of service measurements Care must be taken with using loopbacks however because the traffic contracts of virtual circuits may be asymmetrically specified for the different directions A very useful extension to the use of ITU T Recommendation 0 191 test cells is available if the test cells can be shaped to simulate traffic that only just meets the traffic contract in terms of peak cell rate PCR cell delay variation toler ance CDVT sustainable cell rate SCR and maximum burst size MBS Shaped test cell traffic can be injected into a network at the UNI and on the far side UNI and measurements can be made of the delivered QoS in terms of cell
71. UBLESHOOTING LocAL AREA NETWORKS 386 am 10 The following tables list the limits for the maximum distances that can be spanned using various transmission media as defined in the appropriate ATM standards SDH SONET based ATM over twisted pair cabling Cable type Transmission speed Maximum distance UTP 5 12 96 Mbit s 400 m UTP 5 25 92 Mbit s 320 m UTP 5 51 85 Mbit s 160 UTP 5 155 Mbit s 150m UTP 3 155 Mbit s 100 Cell based ATM over twisted pair cabling Cable type Transmission speed Maximum distance UTP 3 100W 25 Mbit s 100 UTP 4 120W 25 Mbit s 100 STP 150W 25 Mbit s 100 SDH SONET based ATM over plastic fiber Cable type Transmission speed Maximum distance Plastic Optical Fiber POF Hard Polymer Clad Fiber HPCF 155 Mbit s 155 Mbit s 50 100 SDH SONET based ATM over single mode fiber Fiber type Wavelength Transmission Laser type speed Maximum Maximum attenuation distance SM 1 310 nm SM 1 310 1 550 nm SM 1 310 1 550 nm 9 9 Gbit s SLM MLM 9 9 Gbit s SLM MLM 9 9Gbit s SLM MLM 0 7 dB 2 km 0 12 dB 15 km 10 14 dB 40 km TROUBLESHOOTING LocAL AREA NETWORKS 387 SDH SONET based ATM over multimode fiber am 10 Fiber type Wavelength Transmission Laser type Maximum speed distance MM 1 300 nm 155 Mbit s SW LED MMF 2km MM 1 300 nm 622 Mbit s LED MMF 500
72. VPI Virtual Channel ID VCI 3 Virtual Channel ID VCI 4 Virtual Channel ID VCI Payload Type PT CLP 5 Header Error Control HEC CLP Cell Loss Priority Figure 10 22 UNI cell header The Generic Flow Control Field GFC The generic flow control field which consists of 4 bits is used to control local functions and to manage access and transmission rights in ATM networks Its contents are not forwarded beyond the range of the local ATM switch because this field is replaced with VPI data at the NNI Its significance is thus limited to the local ATM network segment In practice although specified by the ITU T in Recommendation 1 361 generic flow control has seldom if ever been imple mented and the ATM Forum in particular does not support this function The ATM Routing Label Field VPI VCI The UNI header contains a total of 24 bits for ATM layer routing purposes 8 bits for the Virtual Path Identifier VPI and 16 bits for the Virtual Channel Identifier TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 330 VCD The VPI VCI value is label that has only local significance that is between two ATM interfaces in a single transmission link This label has no end to end significance which is why the VPI VCI field is referred to as a label not an address There is such a thing as an ATM Address that is used in connection with signaling see the section on Connection Setup at the Caller s End and some publications do refe
73. W93S 5242 ANVd NOWWOD soss uaAviang 32N39H3ANO2 2141234 321 3 seyAq seq 5 Play uoneuuojul 92 WLY XX E a T di Joje n 5 nas Jepeeu nad vs ue e NLY ue 500 9199 seq SUGOL Haz 7 P 52 n e ouo Play GIN NS IS e e 191121 CX nad M HVS gt ue 500 5149 seq py SuqOL SHG y 5192 di Y Ploy uogeuuojul YYS GIN NS 1S XX 4 1 ms itr Add 59890 nas 5980 1 8 P avs s qz SES S9 seMqz q uoneoipui Y uoneounuepi play 3 play uoneuuojul 5980 ezig va Idd Jaquinu aouanbas CU Jeyruepi CC AS uoneoipul ezis sayng nas CS Buipu3 5040 CN 40jeoipur
74. X 200 Whereas the OSI model provides for seven layers however the B ISDN protocol reference model defines the Physical Layer the ATM Layer the ATM Adaptation Layer AAL and the application layer designated in the model as Higher Layer Protocols The physical layer consists of two sublayers the Transmission Convergence TC sublayer and the Physical Medium Dependent PMD sublayer The TC sublayer embeds the ATM layer cells in the transmission framework of the given transport medium If ATM cells are transported over a 34 Mbit s E3 link for example they must be fitted into the user data field of E3 frames The transport medium could also be a Synchronous Digital Hierarchy Synchronous Optical NETwork SDH SONET or a Plesiochronous Digital Hierarchy PDH frame such as DS1 El DS3 or E4 In the case of direct cell transfer over the physical medium with no intermediate transport frame cell based physical layer this sublayer is not necessary that is it is null The functions of the ATM layer are completely independent of the underlying physical layer Their chief purpose is to bring the data received from the higher order AAL to its destination The 53 byte cells comprise the information units of the ATM layer Each cell has an identification number in its header that assigns it to a certain connection The cells belonging to various connections are multi plexed into a cell stream in each direction unused b
75. ace SSFC at UNI B ISDN ATM Adaptation Layer Service Specific Coordination Function for Signaling at the Network Node Interface SSCF at NNI Broadband integrated services digital network B ISDN Usage of Cause and Location in B ISDN User Part and DSS 2 Broadband ISDN Interworking Between Signaling System No 7 Broadband ISDN User Part B ISUP and Digital Subscriber Signaling System No 2 055 2 Broadband integrated services digital network B ISDN Interworking Between Signaling System No 7 Broadband ISDN User Part B ISUP and Narrow band ISDN User Part N ISUP Broadband Integrated Services Digital Network B ISDN Signaling System No 7 B ISDN User Part B ISUP Supplementary Services Broadband Integrated Services Digital Network B ISDN Func tional Description of the B ISDN User Part B ISUP of Signaling System No 7 Broadband Integrated Services Digital Network B ISDN General Functions of Messages and Signals of the B ISDN User Part B ISUP of Signaling System No 7 Broadband Integrated Services Digital Network B ISDN Signaling System No 7 B ISDN User Part B ISUP Formats and Codes Broadband Integrated Services Digital Network B ISDN Signaling System No 7 B ISDN User Part B ISUP Basic Call Procedures TROUBLESHOOTING LocAL AREA NETWORKS 1 0 392 Q 2931 Broadband Integrated Services Digital Network B ISDN Digital Subscriber Signaling System No 2 DSS 2 User Net
76. acts This is done by verifying the compatibility of the traffic contracts for the LE Client interfaces If a protocol analyzer is available a trace of the unsuccessful connection setups is the fastest way to find the cause of the problem Internet Advisor ATM Stopped Decode 8 x Ele Run View GoTo Setup Window Help 18 x B 024 amp 21 9158335 14 E Time Piin Summar I Detailed Hex ASC Vs re Search Next Error Cell Time VPI VC Al rot ription P2 15 0 24 21 9158158 5 150 5 ATM CLP PTI SDUO HEC Good L 5 Type NotEOM P1 24 21 9158159 CLP Low PTI SDUO HEC Good AML 5 Type NotEOM P2 17 0 24 21 9158332 5 150 ATM CLP Low PTI SDU1 HEC Good AML 5 Type EOM Len 113 CRC32 Good RFC 1483 DSAP aa SSAP aa Ctrl UI SNAP OUI Bridged PID 802 3 Ethernet ETHER DEC 00 01 04 gt DEC 00 00 0F Proto D LAT Service Announcement Msg Incarnatio ETHER Header Destination DEC 00 00 0F 09 00 2B 00 00 0F Source DEC 00 01 04 a2 00 04 00 01 04 Protocol DEC LAT Service Announcement 0x0A PROS 0 Response ted Flag FALSE Server Circuit Timer 8 milliseconds Ready 38 6 00215 20 EQ 16 7 LN 167 Figure 10 70 LANE analysis by means of a protocol analyzer TROUBLESHOOTING LocAL AREA NETWORKS 1 0 410 10 2 3 2 UNI Signaling The first steps in troubleshooting AT
77. ag DXI header DTE SDU DXIFCS 1 2 0 lt lt 9232 2 1 Bytes Figure 10 18 DXI data packet for Modes 1a and 1b DTE AAL5 10 1 6 8 ATM Physical Layer Operations and Maintenance OAM 1 F3 A total of five information flows are defined for operations and maintenance of ATM networks called the OAM flows Each of these flows numbered F1 to F5 is responsible for the monitoring of a certain part of a connection Flows F1 to F3 are concerned with the physical layer of the B ISDN protocol model F4 and F5 with virtual paths VP and virtual channels VC in the ATM layer The OAM parameters are determined by different mechanisms depending on the network topology used for ATM cell transport SDH SONET PDH cell based physical layer A transmission path is defined as the path between two network compo nents that insert the user data into the transport medium and extract it again from the medium Examples of transmission paths include the link between a B NT2 and a switch VP cross connect or between a B NT2 and the connection end point A transmission path is composed of several digital sections each of which may lead through one or more regenerator sections OAM F1 SDH SONET Based Systems In SDH SONET based ATM networks the OAM flows F1 and F2 are transported in the SOH TOH of the transport frames and F3 in the POH of the given virtual container SPE Part of the F3 information can also be trans
78. all Os UNI Information Elements After the four mandatory information elements the various message types use specific information elements of varying lengths to fulfill their respective func TROUBLESHOOTING LocAL AREA NETwoRKS 1 0 364 Bit 8 7 6 5 4 3 2 1 Byte Message Type 1 1 Message Action Ext 0 0 0 Indicator 2 Bits 8765 4321 Q 2931 Message 0000 0000 Escape sequence for national message types Connection set up 0000 0001 ALERTING 0000 0111 CONNECT 0000 1111 CONNECT ACKNOWLEDGE 0000 0010 CALL PROCEEDING 0000 0011 PROGRESS 0000 0101 SETUP 0000 1101 SETUP ACKNOWLEDGE Connection clear down 0100 1110 RESTART ACKNOWLEDGE 0100 0110 RESTART 0101 1010 RELEASE COMPLETE 0100 1101 RELEASE Other Messages 0110 1110 NOTIFY 0111 1011 INFORMATION 0111 0101 STATUS ENQUIRY 0111 1101 STATUS Figure 10 46 Q 2931 B DSS2 signaling messages tions Each of these information elements consists of an information element identifier a length field a compatibility indicator and the actual information element contents Figure 10 47 illustrates the general format of information elements and lists the specific information elements defined 10 1 9 2 Connection Setup at the Caller s End UNI The caller initiates the connection setup by transmitting a SETUP message containing the desired ATM virtual path the ATM virtual channel and quality of service and traffic
79. ally does not know the ATM address that corresponds to the Layer 3 destination address of the packet it does not attempt to set up a direct VC to the destination By default the data is first TROUBLESHOOTING LocAL AREA NETWORKS 1 0 384 encapsulated LANE frame forwarded to the default MPS router by the MPC s LE Client unit This router forwards the data to the destination MPC according to the information available in its routing tables Then the MPC attempts to resolve the network address of the packets by sending an MPOA Resolution Request message to the MPS The MPS obtains the corresponding ATM address along with the information whether a direct connection shortcut is available for the given connection from its assigned Next Hop Server NHS If a shortcut is available the MPS sends an Imposition Request message to the destination MPC to ask whether it is able to accept the shortcut connection The destination MPC responds with an Imposition Reply message which the MPS forwards to the originating MPC in the form of an MPOA Resolution Reply The originating MPC can then set up the shortcut connection The shortcut route information is entered in the MPC s routing cache and all data packets for the given destination address are then sent directly over the shortcut Figure 10 58 shows a schematic illustration of the MPOA route selection process The default eee er
80. andwidth being filled with unassigned or idle cells These cell streams are hierarchically structured in virtual channels VC and virtual paths VP which correspond to one or more virtual connections A physical transport medium such as an optical fiber can transport a number of virtual connections Each cell on the medium can be unambiguously attributed to a certain connection by the VPI and VCI in its header comprising the identification number referred to earlier The AAL disassembles the higher layer data streams into 48 byte information segments for transport in ATM cells five of the cell s 53 bytes are used for header information so that 48 bytes of user data are transported in each cell At the receiving end the original data streams must be reassembled from the individual ATM cells The functions of the AAL are thus dependent to a great degree on the characteristics of the higher order applications For this reason the AAL functions are performed by two sublayers the Convergence Sublayer CS and the Segmentation and Reassembly SAR sublayer To limit the number of different AAL implementations four AAL types have been defined AALI AAL2 AAL3 4 and AALS Each of these AAL types is defined for a certain class of applications By far the most widely used AAL variant today is the simplest one to implement AAL5 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 305 The actual network services transported on the basis of th
81. anges to the PRESYNC state Under the assumption that a cell has been detected the PRESYNC state examines subsequent candidate cells If the HEC fields of the next DELTA candidate cells are also possible checksums then the receiver assumes that it has synchronized with the cell stream and switches to the SYNC state and cell synchronization or delineation is declared If less than DELTA con secutive cells meet the HEC test the receiving station reverts to the HUNT state If ALPHA consecutive cells fail the HEC test the receiver in SYNC state assumes it has lost its cell delineation and switches to HUNT state Loss of Cell Delineation LCD is then declared For SDH SONET physical layers the value of ALPHA and DELTA are 7 and 6 respectively for cell based physical layers these values are 7 and 8 respec tively see ITU T Recommendation 1 432 1 Data Field Scrambling The data field payload of the ATM cells is scrambled in order to optimize cell delineation by HEC patterns In the HUNT state potential HEC sequences are easier to detect when the data is scrambled In the PRESYNC and SYNC states an unscrambling function is activated for the 48 user data bytes of the cell but is not applied to the cell headers In all framing types except cell based physical layer the scrambling of the data field is always based on the self synchronizing scrambler SSS 2 1 scrambling is optional in some interface types Due to th
82. ant Bi level U CTD Maximum average none 400 ms U U U CTD 2 point Maximum none 3 ms U U U CDV difference between upper and lower 10 range of CTD 1055 3 107 105 U U probability CLR Maximum cell loss none none none 105 U probability CER Maximum cell error 4 10 default default default U probability CMR Maximum cell 1 day default default default U misinsertion rate SECBR Maximum SECB 104 default default default U probability U unspecified or unlimited Figure 10 68 ATM layer performance parameters 10 2 2 3 Symptoms and Causes The most frequent causes of cell loss are buffer overflows or faults in the physical layer that lead to non correctable errors Cell misinsertion is caused by multiple bit transmission errors in the header which can no longer be corrected as a result of physical layer problems or malfunctions in the switching fabric Cell errors usually indicate the occurrence of bit errors in the payload field bit errors in the header are reflected in cell loss figures In most cases these bit errors are caused by a higher degree of signal jitter than the ATM interface can tolerate Cell transfer delay is caused by ordinary electronic switching and signal propagation delays The cause of cell delay variations usually lies in the varying states of buffers that the cells must pass through on their way to the destination and in
83. ate network resources are reserved 10 1 1 ATM in Homogeneous Private Networks If a network is built using only ATM components then data communication can take place using classic ATM transmission techniques Because ATM is connec tion oriented the data virtual channel used to transport actual user data must be defined before the data transmission by means of a signaling or provisioned connection With signaling special communication protocols govern the nego tiation of the user data connection parameters bandwidth delay routing etc Data transmission takes place until the connection is terminated by an appro priate command on the signaling virtual channel Different signaling protocols are used at the interface between the end system and the ATM switch the user to network interface or UNI and at the interfaces between ATM switches network to network interfaces or NNI In a private network NNIs that is interfaces between two ATM switches use the Private Network to Network Interface PNNI protocol It makes no differ ence whether the private network is LAN or a WAN In public ATM WANs however the NNI protocol commonly used is the Broadband ISDN User Part B ISUP protocol The reason for the use of different protocols in the private and public spheres is that the ITU standardization body has left the internal operat ing aspects of public data networks in the control of national telecommunica tions companies For exampl
84. atible SSCOP layer not established Incompatible UNI signaling variants UNI 3 0 3 1 4 0 Wrong Called Party or Calling Party number Unknown or invalid information elements or mandatory informa tion elements in wrong order Incorrect call reference numbers Called party is not ready to accept call call setup attempt is rejected with RELEASE message Misconfigured ATM port bit rate scrambling interface type frame type PLCP G 804 SDH SONET Hardware or software problems on the switch High Cell Error Rate CER Problems on the physical layer cabling connectors ATM port Too many nodes along the transmission path of a VP or VC con nection High Cell Loss Ratio CLR Problems on the physical layer cabling connectors ATM port Too many nodes along the transmission path of a VP or VC con nection ATM switch overloaded Insufficient buffering in the switch High network load Limits of traffic contract are exceeded 419 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Cause 6 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Cause 6 Cause 7 Cause 8 Cause 9 Cause 10 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 High Cell Misinsertion Rate CMR Problems on the physical layer cabling conne
85. ation Network address for example IP address of the destination Network address of the sender ATM address of the sender The NHS analyzes the incoming address resolution request and verifies whether it is itself competent to resolve the given destination If not it forwards the request to another NHS Next Hop Servers may forward address resolution requests only to other NHSs that are not more than one hop away thus the name next hop server If the NHS is competent for the destination indicated in the request it resolves the address and sends a reply packet containing the Next Hop Network address and the ATM address of the destination to the MPS which forwards the information to the MPC The MPC can then set up a direct VCC or shortcut to the destination 10 1 11 Design Rules for ATM Networks The design of ATM networks is subject to far fewer restrictions than classical LAN technologies such as Ethernet FDDI or Token Ring This is due to the wide range of possible transmission media and to the fact that ATM was originally designed for use in wide area networks Distance limitations are therefore not a fundamental issue So long as the specifications for the various segments of an ATM network are respected distances of several thousand km can be bridged The transmission delay can be calculated using a value of 6 us per km and 2 us per switch The transmission delay for a path of 1 000 km and five switches is 6 01 ms TRO
86. ation of AAL Type 3 4 Like AAL3 4 it is suitable for connec tion oriented and non connection oriented data communications And like AAL3 4 15 also consists of a Segmentation and Reassembly SAR sublayer and a Convergence Sublayer CS with the CS subdivided into a Common Part TROUBLESHOOTING LocAL AREA NETWORKS ATM 1 354 Class C connectionless Class B connection oriented user data user data AAL i SDU 1 X 1 1 gt 1 1 cS SERVICE SPECIFIC e SSCS PDU SSCS PDU SSCS PDU CONVERGENCE SUBLAYER information field trailer SSCS SA if i TUN i 6685 i N SDU CS Common Part S TTE CONVERGENCE SUBLAYER T EN CPCS PDU CPCS CPCS PDU information field PAD trailer T 1 65 535 bytes 0 47 bytes 8 bytes ii SAR PDU 1 information field i L i PT 0 48 bytes SS SAR PDU SEGMENTATION information field 1 22 AND REASSEMBLY SUBLAYER PT 0 48 bytes 1 SAR uU men SAR PDU PT information field 1 Padding to fill the CPCS PDU to a PT 0 48 bytes multiple of 48 bytes Payload type field 0 Beginning or continuation of a SAR PDU
87. bsolescent Transparent connection in this context means that not only user data but also protocol information is extracted from the foreign network topology and interpreted In the case of Frame Relay for example the ATM network interprets and understands the Frame Relay Forward Explicit Congestion Notification FECN bit and the Discard Eligible DE bit and represents them using the corresponding ATM protocol parameters EFCI and CLP In connection with Frame Relay networks ATM emulates the Frame Relay UNI by means of the Frame Relay Service Function FRSF Because Frame Relay is also connection oriented it can be emulated relatively easily in ATM Connecting SMDS networks to ATM is some what more complicated because these networks like conventional LANs are based on the principle of connectionless communication As in the case of LAN emulation the connectionless transmission principle of SMDS must be trans lated to ATM s connection oriented structure This is done by means of the two protocols Connectionless Network Interface Protocol CLNIP and Connectionless Network Access Protocol CLNAP Another way to transport data over an ATM Network is to use the ATM Circuit Emulation Service CES With CES the ATM network becomes a conduit for circuits such as 1 T1 E3 T3 etc so any service that can be transported over these circuits is by default also transported over ATM Permanent virtual circuits are frequently used for CES an
88. ctions such as error or performance management The fourth OAM cell type system management cells has no defined purpose in the specifications its usage is left to individual manufacturers implementations OAM Fault Management AIS RDI In analogy to SDH SONET fault management in the ATM layer involves two kinds of alarm signals Alarm Indication Signal AIS and Remote Defect Indica tion RDI The VP AIS or VC AIS is sent by the virtual path or virtual channel node that detects the fault downstream to all network nodes directly affected The VP VC AIS OAM cell is transmitted periodically approximately 2 second intervals until the fault is eliminated Immediately after the AIS has been sent a VP RDI or VC AIS OAM cell is sent upstream to the endpoint of the virtual path or virtual channel connection concerned This OAM cell is likewise repeated periodically as long as the AIS condition persists VP AIS and VP RDI messages are always transmitted by means of cells with VCI 4 while VC AIS and VC RDI messages are transported in cells with PT 101 fault at the virtual path level inevitably affects the virtual channels contained within that virtual path This faulty virtual path may terminate at a switch before the endpoint of these virtual channels which will continue in one or more new virtual paths determined by switching It is therefore necessary that the virtual channels affected continue to be notified of the fault condition This
89. ctors jitter ATM port Too many nodes along the transmission path of a VP or VC con nection High network load ATM switch malfunction High Mean Cell Transfer Delay MCTD High signal delay due to long transmission path Too many nodes along the transmission path of a VP or VC con nection ATM switch overloaded Insufficient buffering in the switch High network load Limits of traffic contract are exceeded High Cell Delay Variation CDV Too many nodes along the transmission path of a VP or VC con nection ATM switch overloaded Insufficient buffering in the switch High network load Limits of traffic contract are exceeded No Connection over Emulated LAN ELAN Problems with the connected traditional LANs Ethernet FDDI Token Ring IP interfaces on the LAN emulation clients are not active or not functioning IP addresses and subnet masks are incorrect interfaces belong to different subnets LANE software on the client is not active The LE Clients trying to communicate do not belong to the same ELAN The LE Clients are not registered on the same LE Server BUS The VCC and ATM address of the LANE server LE Server are incorrect The VCC and ATM address of the BUS are incorrect LANE ARP entries are incorrect MAC ATM address resolution is not working The traffic contracts of the LE Clients are incompatible 420 Cause 11 Symptom Cause 1 Cause 2 Cause 3 Cause 4 Cause 5 Cause 6
90. d special interface equipment is needed to implement CES More details are given later in the section dealing with the ATM Adaptation Layer 10 1 8 As mentioned previously the B ISUP communications protocol developed by the ITU T is used at the NNI between two ATM switches in a public network The Broadband Inter Carrier Interface B ICI is an enhanced network to network interface protocol developed by the ATM Forum for connecting the networks of different carriers communications service providers and is used mainly in North America A number of telecommunications providers use PNNI between their public ATM switches however for the sake of greater simplicity in opera tion 10 1 4 Asynchronous Transfer Mode ATM ATM is defined as a packet switched connection oriented data communication technique based on asynchronous time division multiplexing and data packets of a fixed length ATM data packets are called cells because of their constant TROUBLESHOOTING LocAL AREA NETwoRKS 302 ATM Synchronous Transfer Mode STM Synchronous Time Division multiplexing Circuit switched Physical connection established on call set up 5
91. dd further links 0010 Group start up 0011 Failure protocol error 0100 Failure insufficient links 0101 Failure unsupported value of M 0110 Failure incompatible group symmetry 0111 Failure symmetry not supported Bits 3 2 Group symmetry mode 00 Symmetrical configuration and operation 01 Symmetrical configuration and asymmetrical operation optional 10 Asymmetrical configuration and operation optional 11 Reserved Bits 1 0 IMA Frame length 00 M 32 01 M 64 10 M 128 11 M 256 Byte 14 Transmit timing information Bits 7 6 not used set to 0 Bit 5 Transmit clock mode 0 ITC mode 1 CTC mode Bits 4 0 Tx LID of the timing reference 0 to 31 Byte 15 Tx test control Bits 7 6 not used set to 0 Bit 5 Test link command 0 inactive 1 active Bits 4 0 Tx LID of test link 0 to 31 Byte 16 Tx test pattern Bits 7 0 Tx test pattern values 0 255 Byte 17 Rx test pattern Bits 7 0 Rx test pattern values 0 255 Byte 18 Link 0 information Bits 7 5 Tx status Bits 4 2 Rx status Bits 1 0 Remote Defect Indicator Bytes 19 49 Link 1 31 info status Byte 50 not used set to 6A hex ITU 1 432 Byte 51 End to end channel proprietary channel set to 0 if not used Bytes 52 53 CRC 10 ITU 1 432 Figure 10 9 ICP cell format TROUBLESHOOTING LocAL AREA NETWORKS 1 0 313 Inverse Multiplexing for ATM IMA ATM inverse multiplexing is the process of dividing an ATM cell stream into several component st
92. e Line Status 32 File Run View GoTo Setup Window Help mn uo ee ew Re EO P1 Current Line Status LN P2 m Select Signal AIS Signal AIS Iv History Frame 5 Frame 5 Cell Sync Yellow Cell Sync Yellow Far End c i Far End Count P2 Last Occurred P2 171 10 15 07 009089615 172 10 15 07 003089615 0 112 10 15 05 006145520 207 10 15 05 006145520 106 10 15 05 006145520 0 10 14 34 001472047 9 10 15 05 006145520 Figure 10 59 F1 F3 OAM flow statistics with the Agilent Technologies Advisor ATM Last Occurred P1 Signal Count P1 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 395 Analysis of the integrated network diagnostic functions and the OAM flows is usually sufficient to isolate the systems that are affected by the problem The troubleshooting process then entails basic functional tests of these components including loopback tests on the ATM interfaces firmware tests is the firmware active and hardware self tests Cabling and connectors are checked by running bit error ratio tests and performing OTDR and LED laser power spectrum measurements OTDRs or optical time domain reflectometers are physical layer test instruments for fiber connections They send defined light pulses over the fiber and measure amplitude and response time of the reflected return signal The test results includ
93. e B ISUP does not specify how routing information is propagated or how topology detection mechanisms should be implemented In private networks however it is desirable for all data communication mecha nisms to be precisely defined so that communication works automatically with out custom additions For this reason PNNI also specifies appropriate routing mechanisms in addition to the signaling processes Because PNNI has the basic capabilities necessary for use in public WANs and because many ATM system manufacturers provide an implementation of PNNI but not B ISUP PNNI is sometimes also used in public networks especially in North America In Europe and Asia however almost all public networks use B ISUP At the interface between an ATM switch and an ATM end system that is the UNI connection setup is governed by the ITU T Recommendation Q 2931 and Q 2971 for point to multi point connections or by one of the ATM Forum s UNI signaling protocols UNI 3 0 3 1 or 4 0 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 299 When an ATM end system is installed becomes active it must first register with its assigned ATM switch Information exchanged in this process includes the network address of the switch the user address of the ATM end system and the service characteristics of the ATM end system All of this information is stored in the ATM switch in a defined format as the Management Information Base MIB In order for this regist
94. e relatively poor overall communication characteristics of the SSS process cell based physical layer uses Distributed Sample Scrambling DSS with the genera TROUBLESHOOTING LocAL AREA NETWORKS 1 0 333 tor polynomial 221 228 1 Note that because the scrambling takes place within the transmission convergence sublayer of the physical layer for non cell based interfaces all cells at the interface are scrambled using the SSS process inde pendent of the header routing label value VPI VCI and including any idle unassigned and OAM cells etc Reserved Header Values Certain header values are reserved for cells with special operation management or signaling functions in the ATM network Such cells include broadcasts meta signaling resource management cells and in cell based physical layer PL OAM cells All special cells have a VCI value in the range 0 through 31 the ITU T reserves to itself all values of VCI below 16 and makes available all values between 16 and 31 for proprietary purposes the ATM Forum has used several values in this range for its purposes for example VCI 16 is used for ILMI VCI 17 is used for LANE VCI 18 is used for PNNI With the exception of these defined values all values VCI 32 and above can be used for user traffic by the ATM layer Figure 10 25 lists the reserved header byte values ITU T Recommendation 1 361 Furthermore the ITU T recommendations and ATM Forum specifications define certa
95. e appropriate AALs The most common applications are Ethernet Token Ring using LAN emulation Classical IP over ATM MPEG Video Moving Pictures Experts Group Frame Relay Leased Line Data Links Voice Telephony 10 1 6 The ATM Physical Layer Three methods can be used to transport ATM cells cell adaptation to the frame structure of the transport medium such as PDH SDH SONET or DXD trans port in PLCP frames or cell based physical layer In order to continue using existing data communications infrastructures tech niques have been developed to transport ATM cells in the container frames of the most common communications interfaces Cell transport in SDH containers or SONET Synchronous Payload Envelopes SPEs is the most widely used transport method for ATM cells today Adaptation methods have also been developed to transport ATM cells in wide area networks over the older but still widely used 1 E3 4 T1 and T3 interfaces In this technique the cells are inserted directly into the user data field of the given transport frame and the cell rate is adapted to the throughput capacity of PDH by inserting fillers called idle cells In North America unassigned cells are often used for most purposes idle and unassigned cells can be considered identical Adaptation to the existing PDH frame is defined in ITU T Recommendation G 804 for the following transfer rates 1 544 Mbit s 2 048 Mbit s 6 31
96. e the fiber length and all attenuation components splices connectors fiber loss along the segment If the components pass the basic functional tests OAM flows F1 F3 in the SDH SONET layer are examined If no fault is found here the ATM layer and the application protocols above it must be analyzed This begins with checking whether the required PVCs and SVOCs are active and working and whether the ATM addresses are correct To determine whether the traffic contract parameters for the connections or appli cations in question are being met characteristic ATM layer traffic parameters are measured including User cell rate Cell loss Cell delay Number of cell sync losses Number of cells with corrected headers ATM payload bit error ratio Finally if the ATM layer seems to work correctly the application layer protocols such as Classical IP over ATM IP encapsulation LAN emulation or PNNI must be examined Each step in the ATM troubleshooting process outlined previously is discussed in detail in the following sections 10 2 1 Troubleshooting the Physical Layer Once the error domain has been located the troubleshooting process can start If connections are interrupted or network nodes are down the first step usually consists of basic functional tests of the component systems Do the activity LEDs of the interface in the problem domain indicate normal working order Most ATM interfaces indicate normal sending a
97. ed Default route MPC 1 MPC 2 Shortcut LEC LEC MPC MPS MPC CT DA Reus _ Reques Imposition Request imposition Reply Resolution Reply Shortcut Figure 10 58 Route selection and shortcuts in MPOA networks TROUBLESHOOTING LocAL AREA NETWORKS 1 0 385 route between the MPOA clients 1 MPC2 leads first to the LE Client interfaces of the three MPOA components connected to one another over the three ELANs Once the shortcut has been obtained from the Next Hop Server direct connection can be set up between the LE Client interfaces of the two MPOA clients The Next Hop Resolution Protocol NHRP Route optimization in MPOA networks is performed by means of the Non Broadcast Multiple Access NBMA and NHRP because broadcast based address resolution mechanisms are completely unsuitable for ATM networks The pur pose of NHRP is to find the ATM address that corresponds to a given network address such as an IP address so that a direct connection shortcut can be set up between two communicating stations NHRP is based on a client server model in which Next Hop Clients NHC send address resolution requests to a NHS In MPOA the MPOA servers play the part of NHCs and initiate Next Hop Resolution requests to the NHS as required by MPCs Every NHR request contains the following inform
98. efault 5 7 00000000 00000000 00000000 00010a0c Meta signalling 7 0000 yyyy0000 00000000 00010a0c General Broadcast signalling default 5 00000000 00000000 00000000 00100aac General Broadcast signalling 0000 yyyy0000 00000000 00100aac Point to point signalling default 5 00000000 00000000 00000000 01010aac Point to point signalling 0000 yyyy0000 00000000 01010aac Invalid Pattern xxxx0000 00000000 00000000 0000xxx1 Segment OAM F4 flow cell 7 0000aaaa aaaa0000 00000000 00110a0a End to End OAM F4 flow cell 7 0000aaaa aaaa0000 00000000 01000a0a Tuus a indicates that the bit is available for use by the appropriate ATM layer function 2 X indicates don t care bits 3 indicates any VPI value other than 00000000 4 c indicates that the originating signalling entity shall set the CLP bit to 0 The network may change the value of the CLP bit S Reserved for user signalling with the local exchange 8 ct Reserved for signalling with other signalling entities e g other users or remote networks 7 The transmitting ATM entity shall set bit 2 of octet 4 to zero The receiving ATM entity shall ignore bit 2 of octet 4 Figure 10 26 Header bytes reserved by the ATM Forum The NNI Header Unlike the UNI header the NNI header provides 28 bits for addressing 12 bits for the VPI and 16 bits for the VCI This permits the definition of a greater number of virtual paths at the network node in
99. eived that signals a compatible call state but contains cause 96 977 99 100 or 101 the response is left to the given implementation If no particular reaction is specified the connection in question should be cleared down with the cause indicated in the STATUS message received TROUBLESHOOTING LocAL AREA NETWORKS 1 0 415 Internet Advisor Stopped Decode 5 Run View GoTo Setup Window Help 18 x Bv se REE Z 1125527838355 14 gt Rect Time Piin Gunmen I Detailed EBCDIC Filter Search Nest Error cell Time ot cription E P2 1 11 13 56 2783835 0 5 5 ATM CLP High PTI SDU1 HEC Good ALL 5 Type EOM Len 8 CRC32 Good SAAL BGN 50 001 N MR 0000100 Pl 2 11 13 56 2783835 0 5 5 ATM CLP High PTI SDU1 HEC Good 1 5 Type EOM Len 8 CRC32 Good SAAL BGN 50 001 N MR 0000100 21 3 11 13 57 1128672 085 5 CLP High PTI SDU1 HEC Good 1 5 Type EOM Len 8 CRC32 Good SAAL BGAK N MR 0000100 P2 4 11 13 57 1128673 0 5 5 CLP High PTI SDU1 HEC Good ALL 5 Type EOM Len 8 CRC32 Good SAAL BGAK 0000100 P2 5 11 13 57 8880282 0 5 5 CLP High PTI SDU1 HEC Good ALL 5 Type EOM Len 8 CRC32 Good SAAL POLL 5 0000001 5 0000000 P1 6 11 13 57 8880282 025 5 CLP High PTI SDU1 HEC Good AAL 5 Type EOM Len 8 CRC32 Good SAAL POLL 5 0000001 N S
100. el Identifier VCI and the Virtual Path Identifier VPI Note that ATM is asynchronous because cells are sent asynchronously not because the octets bytes are asynchronous as in RS 232C All network nodes are con nected by one or more ATM switches that forward the cells toward their destination point The fixed cell size makes it possible for the ATM switches to forward multiple cells simultaneously with high efficiency so that a very high throughput is attained in comparison to conventional routers The network nodes do not share a communication medium as LAN nodes do but simply give up their cells to the local ATM switch without having to deal with media access algorithms Before an actual data transmission begins however the transmis sion path for the user data cells is defined and set up by a signaling or provision ing procedure A traffic contract guarantees that the traffic parameters granted such as the maximum transfer delay bandwidth or cell loss ratio will be provided for the duration of the connection This feature makes it possible to transmit many high quality multimedia communication services over ATM net works If the requested traffic parameters cannot be met by the network at call setup time the call is rejected by the network so that the quality of service of other network users is not compromised The fixed cell length of 53 bytes results from a necessary compromise between the requirements of analog voice
101. ell rate is adapted to the E3 user data rate by the insertion of idle cells when no ATM cells are queued for transport in North America unassigned cells are often used for this purpose both idle and unassigned cells are discarded at the input to switches etc The 48 byte data field of the ATM cell including idle unassigned cells is scrambled using self synchronizing scrambling with the generator polynomial x 1 59 bytes E Header A Header Header Header 2 o Header Header Head Header Y ATM cell Header 53 bytes Overhead byte Figure 10 6 Cell adaptation to the frame format ATM over DS3 44 736 Mbit s ATM cells can be transported over DS3 links using either the Physical Layer Convergence Protocol PLCP mapping found in older equipment or the direct cell mapping to the DS3 frame format preferred today A DS3 PLCP frame consists of twelve rows of 57 bytes each The last row contains an additional trailer of twelve or thirteen nibbles half bytes to fill out the user data field of a DS3 multiframe The 053 PLCP frame has a transmission time of 125 us corresponding to a rate of 44 21 Mbit s and thus fits exactly in the user data field of a DS3 multiframe Figure 10 7 TROUBLESHOOTING LocAL AREA NETWORKS
102. en ordinary user data and various special types such as traffic or network management informa tion for example The default value of the PT field is 000 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 331 Cell Loss Priority CLP The cell loss priority bit can be used to assign cells a relative priority If the CLP bit is set to 1 the cell has low priority 0 indicates normal higher priority If the capacity of a connection is exceeded or if other transmission problems arise cells with low priority are discarded first Header Error Control HEC Before an ATM cell is transmitted a cyclic redundancy check CRC of the entire cell header is calculated and placed in the 8 bit header error control field At the receiving end of each link the header is checked to see if the HEC value is correct it would normally only be correct if there were no bit errors anywhere in the header If a single bit error has occurred somewhere in the header including the HEC itself because of physical layer bit errors for example there is sufficient redundancy in the HEC to allow that bit error to be corrected Similarly two bit errors will always be detectable However the detection of three or more bit errors in the header may not be detectable For example multiple bit errors may change the header such that it becomes a header with a different corrupt VPI VCI value and for which the corrupt HEC is correct If the new VPI VCI value happe
103. er Specification Type 5 AALI 432 1 B ISDN User Network Interface Physical Layer General Specification B ISDN Operation and Maintenance Principles and Functions B ISDN ATM Layer Cell Transfer Performance formerly I 35B Traffic Control and Congestion Control in B ISDN General Aspects of Quality of Service and Network Performance in Digital Networks Including ISDNs Frame Relaying Bearer Service Interworking B ISDN ATM Adaptation Layer Sublayers Frame Relaying Service Specific Convergence Sublayer FR SSCS Congestion Management for the ISDN Frame Relaying Bearer Ser vice Frame Relaying Bearer Service Network to Network Interface Requirements ISDN Frame Mode Bearer Services ISDN Frame Relaying Bearer Service Broadband Connection Oriented Bearer Service 391 812 Q 2010 Q 2100 Q 2110 Q 2120 Q 2130 Q 2140 Q 2610 Q 2650 Q 2660 Q 2730 Q 2761 Q 2762 Q 2763 Q 2764 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 Broadband Connectionless Data Bearer Service Broadband Integrated Services Digital Network Overview Signaling Capability Set 1 Release 1 B ISDN Signaling ATM Adaptation Layer SAAL Overview Description B ISDN ATM Adaptation Layer Service Specific Connection Ori ented Protocol SSCOP B ISDN Meta Signaling Protocol B ISDN Signaling ATM Adaptation Layer Service Specific Coordi nation Function for Support of Signaling at the User Network Interf
104. er byte and 47 payload bytes The CPS PDUs in turn are transported in the payload fields of ATM cells The start field STF comprises an offset value pointing to the start of the next CPS packet to aid CPS packet delineation recovery in the event that an ATM cell has been lost Figure 10 34 shows a diagram of the structure of AAL2 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 351 AAL2 Error Messages The following error messages in Figure 10 35 are passed to the layer manage ment in case of transmission errors in AAL2 Error number Description 0 The start field in the CPS PDU indicates a parity error The entire PDU is discarded 1 The sequence number in the start field is invalid If the offset is 0 the entire PDU is discarded Otherwise processing continues at the byte indicated by the offset 2 The number of bytes of an overlapping CPS packet does not match the parameters in the start field If the offset is less than 47 proces sing continues at the byte indicated by the offset 3 The offset value is greater than 47 The entire PDU is discarded 4 The HEC checksum of a CPS packet indicates a transmission error in the header The corresponding part of the CPS PDU is discarded 5 The padding bits extend into the following CPS PDU This extension is ignored and not processed 6 A CPS packet fragment was received and had to be discarded before it could be reassembled 7 The HEC checksum
105. ers situated above it in the protocol hierar chy Q 2931 MTP 3 and B ISUP with a reliable transportation service because these signaling protocols have no error compensation mechanisms themselves Fault tolerance is provided by SAAL s SSCS sublayer using the Service Specific TROUBLESHOOTING LocAL AREA NETWORKS 1 0 356 Connection Oriented Protocol SSCOP which builds on CPCS SAR sublayers of AAL3 4 or AALS Because the CPCS sublayer of AAL3 4 and AAL5 is only able to perform unassured information transfer a substantial part of the SSCOP protocol is concerned with procedures to guarantee transmission of the SSCOP information field contents This is analogous to the function of the TCP layer used with IP for guaranteeing reliable transport of data across IP net works The given signaling layer s requests are translated to SSCOP by an appropriate Service Specific Coordination Function SSCF Although specified as an option for use in practice AAL3 4 is not used for signaling The SSCOP protocol defines 15 distinct PDUs that are used to implement different functions The PDUs Begin BGN and Begin Acknowledge BGAK are used to set up the SSCOP connection between two stations and reset the send and receive buffers of the receiving station Assured data communication can then take place in two ways the data packets can be sequentially numbered SD PDUs and each data packet may contain an individual request for confirmati
106. esses or network management functions One aspect of VPC worth noting is that while the VPI of a VPC changes at every switching node in the network the VCI values of all VCs within the VP are preserved end to end 10 1 7 3 Quality of Service QoS Parameters In ATM networks services with widely differing communication requirements are transported concurrently Real time applications with varying bit rates are much more exacting with regard to average cell delay for example than simple data transfers at a constant bit rate For this reason the ATM layer assigns each connection QoS parameters that specify certain characteristics of the connec tion when the connection is set up ITU T Recommendation 1 356 defines seven cell transfer performance parameters Cell Error Ratio Severely Errored Cell Block Ratio Cell Loss Ratio Cell Misinsertion Rate the proportion of cells with a valid but incorrect header Cell Transfer Delay Mean Cell Transfer Delay Cell Delay Variation TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 339 On connection setup any user can request traffic contract that specifies particular QoS class for each direction of communication Once the traffic contract has been negotiated the ATM network or rather the ATM switches along the transfer path guarantee the QoS parameters granted for as long as the user respects the traffic contract A distinction is made between service classes
107. etwork receives an unexpected RELEASE message the network shall release the virtual channel clear the network connection and the call to the remote user indicating the cause received in the RELEASE message sent by the user if no cause was included cause No 31 Normal unspecified Further more the network shall return a RELEASE COMPLETE message to the user TROUBLESHOOTING LocAL AREA NETWORKS 1 0 412 release the call reference stop all timers and enter the null state Whenever the user receives an unexpected RELEASE message the user shall release the virtual channel return a RELEASE COMPLETE message to the network release the call reference stop all timers and enter the null state The second exception is when the network or the user receives an unexpected RELEASE COMPLETE message Whenever the network receives an unexpected RELEASE COMPLETE message the network shall disconnect and release the virtual channel clear the network connection and the call to the remote user indicating the cause given by the user or if no cause was included cause No 111 Protocol error unspecified Furthermore the network shall release the call reference stop all timers and enter the null state Whenever the user receives an unexpected RELEASE COMPLETE message the user shall discon nect and release the virtual channel release the call reference stop all timers and enter the null state Information Element Sequence Infor
108. etworks is the need to transport monitoring and management information In both classical PDH and modern SDH SONET networks the frames or containers in which data are transported also contain the monitoring and error handling information required by the switching equipment For this reason special operation control cells have been defined for the physical layer called Physical Layer Operation and Maintenance PL OAM cells These cells transport infor mation required for the monitoring and management of the cell stream Many of the existing infrastructures in wide area networking are unable to support ATM cell based physical layers however The use of this technique is therefore limited to local area networks In practice transportation of ATM cells in SDH SONET frames has also become accepted in LANs even though as ATM Transport frame Communication medium Data rates Mbit s SDH SONET Single mode fiber 155 622 2460 9960 SDH SONET Multimode fiber 155 622 SDH SONET 155 SDH 750 coaxial cable 155 SDH SONET UTP3 UTP5 150Q STP 12 96 25 92 51 84 155 750 coaxial cable 2 048 34 36 44 73 139 26 PDH 100Q TP 1 544 2 048 DNI V 35 EIA TIA 449 HSSI up to 52 Cell stream Single mode fiber 155 622 Cell stream Multimode fiber 155 622 Cell stream V 35 1 449 HSSI up to 52 Cell stream UTP3 1200 Cat 4 25 6 150Q
109. fiber MMF interfaces are specified for 622 Mbit s interfaces using LEDs or short wave SW lasers Both types of MMF interfaces support the following multimode fiber types 62 5 125 um IEC 793 2 Type Alb 50 125 um IEC 793 2 Type Ala The data stream is NRZ encoded LED based MMFs can be used for segments of 300 to 500 m SW MMFs have a range of about 300 m Figure 10 15 lists the transmission and reception parameters for both MMF types at 622 Mbit s ATM over Plastic Optical Fiber As an alternative to both single mode and multimode fiber optic media the ATM Forum has also specified plastic optical fiber as a transport medium for ATM networks For STM 1 OC 3 interfaces 155 Mbit s and segments of up to 100 m Plastic Optical Fiber POF can be used and segments of up to 100 m can use Hard Polymer Clad Fiber HPCF The POF medium is 1 000 um multimode TROUBLESHOOTING LocAL AREA NETwoRKS 1 319 LED based MMF parameters Transmitter characteristics 62 5 um MMF 50 um MMF Units Wavelength 1270 to 1380 1270 to 1380 nm Maximum spectral width 200 200 nm Mean optical power 20 to 14 24 to 14 dBm Minimum extinction rate 10 10 dB Maximum rise and fall time 10 90 1 25 1 25 ns Maximum systematic interface peak to peak jitter 0 4 0 4 ns Maximum random interface peak to peak jitter 0 15 0 15 ns Maximum overshoot 25 25 Receive characteristics Minimum sensitivity 26 26 dBm Minimum overload 14 14 dBm Maximum ri
110. frame format ATM over 6 312 Mbit s and 97 728 Mbit s For the sake of thoroughness we may mention at this point that ATM cell transport is also specified for the PDH bit rates 6 312 and 97 728 Mbit s These interfaces are only of regional importance however because they are almost never used outside Japan TROUBLESHOOTING LocAL AREA NETWORKS 312 ATM Bytes 1 5 header GFC 0 VPI 0 VCI 0 PTI 101 CLP 1 HEC Byte 6 IMA label for compatibility with UNI 3 1 OAM cell format OAM type field Byte 7 Cell ID Bit 7 IMA OAM Cell Types 0 idle cell 1 ICP cell Link ID Bits 6 5 not used set to 0 Bits 4 0 logical ID for lines 0 to 31 Byte 8 IMA frame sequence number 0 255 cyclical Byte 9 ICP cell offset range 0 to 1 indicates ICP cell position in the IMA frame Byte 10 Link stuff indication Bits 7 3 not used set to 0 Bits 2 0 111 No imminent stuff event 100 Stuff event in 4 ICP cell locations optional 011 Stuff event in 3 ICP cell locations optional 010 Stuff event in 2 ICP cell locations optional 001 Stuff event at the next ICP cell location mandatory 000 this is one of the two ICP cells compromising the stuff event mandatory Byte 11 Status and control change indication Bits 7 0 Status change indication 0 to 255 Byte 12 IMA ID Byte 13 Group status and control Bits 7 4 Group state 0000 Group is active and ready to add further links 0001 Group is inactive and not ready to a
111. gnaling 359 ATM symptoms and causes 418 B ISUP Signaling 373 Broadcast and Unknown Server BUS 377 Carrierless Amplitude Modulation Phase Modulation 320 Cell Delay Variation CDV 405 Cell Error Ratio 402 Cell Loss Priority CLP 331 340 Cell Loss Ratio CLR 404 Cell Misinsertion Rate CMR 405 Cell Transfer Delay CTD 405 Cell based physical layer 305 321 Classical IP and ARP over ATM RFC 2225 376 Connectionless Network Access Protocol CLNAP 301 Connectionless Network Interface Protocol CLNIP 301 Continuity Checks CC 343 Corrected header ratio 398 D Data field scrambling 332 Demux error ratio 399 Design rules for ATM networks 385 Discarded cell ratio 399 Distributed Sample Scrambling DSS 332 E Emulated LANs ELANs 299 Extended Superframe 307 F F4 and F5 OAM cell format 342 Far End Receive Failure FERF 316 TROUBLESHOOTING LocAL AREA NETwoRKS 1 0 Index p3 6 832 E4 frame format 311 Generic Flow Control Field GFC 329 H Header Error Control HEC 331 HEC cell synchronization algorithm 331 High Cell Delay Variation CDV 419 High Cell Error Rate CER 418 High Cell Loss Ratio CLR 418 High Cell Misinsertion Rate CMR 419 High Mean Cell Transfer Delay MCTD 419 HUNT state 332 ICP cell format 312 Idle cells 305 335 IMA Inverse Multiplexing for ATM Control Protoco 313 IMA Control Protocol cells 336 Inverse Multiplexing for ATM Control Protocol
112. h as telephony Every ATM end system is connected to a dedicated switch port Every data packet or cell that is addressed to a given station is delivered by an ATM switch to a corre sponding switch port Consequently all data packets no longer travel over the same shared broadcast medium Pick up and delivery of cells data packets is managed entirely by an ATM switch The key advantage of this technique is that every station connected to an ATM switch is guaranteed a certain bandwidth on its port regardless of how many other nodes are connected This is possible because the switch s internal data throughput is many times higher than the bandwidth of any switch port In conventional LANs the average bandwidth available to any node is inversely proportional to the total number of active nodes Furthermore the switching principle is combined in ATM networks with connection oriented data transmission The traffic parameters necessary for a given service can be negotiated during the signaling or provisioning procedure TROUBLESHOOTING LocAL AREA NETWORKS 1 0 298 required for every connection For example an end system may request transmission path to a destination with a certain bandwidth and a certain maximum transmission delay If the switches along the transmission route have the capacity necessary to grant the requested connection then these communi cation parameters are guaranteed for the duration of the connection and appro pri
113. he QoS parameters and the traffic parameters of the SETUP request with the local services available If the SETUP parameters requested are not locally available the receiving station responds with a RE LEASE COMPLETE message and Cause 88 Incompatible Destination If the SETUP parameters are compatible the receiving station can respond with a CALL PROCEEDING ALERTING or CONNECT message depending on the type of end system The network confirms a CONNECT message with a CONNECT ACKNOWLEDGE and the connection enters the active state Numbers The SETUP message identifies the station called by its number in the informa tion elements Called Party Subaddress and Called Party Number The net TROUBLESHOOTING LocAL AREA NETWORKS 1 0 367 CALLING STATION ATM NETWORK RECEIVING STATION VPCI VCI OK QoS OK User cell rate OK Network service OK Transit network known VPCI VCI OK QoS OK CALL PROCEEDING EDING User cell rate OK CALL PROCE Network service OK Caller number OK Figure 10 48 UNI connection setup work verifies that this is a valid number If not the network sends a RELEASE COMPLETE with one of the following messages 1 Unassigned unallocated number 2 No route to destination 3 Number changed 28 Invalid number format incomplete number ATM Addressing Six types of ATM addresses are specified in Q 2931 Unknown nternational numbe
114. he information element identifier of each invalid information element in the diagnostic field If address information fields are also corrupt then cause 43 Access information discarded is sent in place of cause 100 If an information element is recognized but should not be present in the given message it is treated as an unrecognized information element AAL Signaling Error If an AAL signaling error occurs all connections not yet started are initialized and a T309 timer is started for each active connection Then a restart ofthe AAL signaling layer is initiated If a connection s T309 expires before the signaling layer can be restarted that connection is deactivated with cause 27 Destina tion out of order and its call reference is deleted TROUBLESHOOTING LocAL AREA NETWORKS 1 0 414 Status Enquiry A STATUS ENQUIRY message can be sent to check the call state at a peer entity Furthermore whenever the SAAL indicates that a disruption has occurred at the data link layer a STATUS ENQUIRY message is sent to check for a correct call state at the peer entity When the STATUS ENQUIRY message is sent timer T322 is started in anticipation of an incoming STATUS message Only one unanswered STATUS ENQUIRY may be outstanding at any given time The receiver of a STATUS ENQUIRY message responds with a STATUS message indicating cause 30 Response to STATUS ENQUIRY and reporting the current call state If no STATUS response is
115. hence the 8 byte length of the trailer 2 x 32 bits there is really no need for more than 7 bytes in the trailer the CPCS UU and CPI fields 8 bits each are rarely used and could have shared a single byte 10 1 8 6 The Signaling ATM Adaptation Layer SAAL As in narrow band ISDN B ISDN signaling is also handled in signaling virtual channels that are separate from the user connections The Signaling AAL ITU T Recommendations Q 2100 Q 2144 SAAL is the AAL used for all ITU T signaling protocols and ATM Forum UNI signaling protocols versions 3 1 and 4 0 Version 3 0 used a prenormative version of the SAAL UNI 3 0 was pub lished about a year before the ITU T had completed Recommendation Q 2931 UNI 3 0 is now obsolete Figure 10 38 shows the protocol layer model for ITU T UNI and NNI signaling in ATM networks with the position of the SAAL in the protocol model UNI NNI UNI B ISUP B ISUP Q 2931 Q 2931 MTP 3 MTP 3 Q 2931 Q 2931 SAAL SAAL SAAL SAAL SAAL SAAL ATM ATM ATM ATM ATM ATM PHY PHY PHY PHY PHY PHY Figure 10 38 Protocol layer model for UNI and NNI signaling The SAALs for UNI and NNI signaling have several features in common but the NNI SAAL has a more complex structure due to the greater number of mecha nisms that need to be provided for MTP 3 and B ISUP The purpose of the SAAL is to provide the actual signaling lay
116. hy this method is preferable to LLC encapsulation wherever possible Figure 10 53 Routed data packet Network Layer packet PAD UU CPI LNG CRC PAD Ethemet 0000 Ethernet CRC PAD UU CPI LNG frame UU Transparent User to User Info CPI Common Part Identification CRC Cyclic Redundancy Check Figure 10 53 VC based multiplexing RFC 2684 10 1 10 2 LAN ATM Classical IP and ARP over RFC 2225 Classical IP and Address Resolution Protocol ARP over ATM go beyond the RFC 2684encapsulation technique to provide a complete implementation of the Internet protocol for ATM IP address resolution which is realized in Ethernet by ARP and Reverse Address Resolution Protocol RARP is handled by ATMARP and InATMARP functions The mapping tables for the ATMARP and nATMARP functions are stored in an ATMARP server which must be present in each logical IP subnet LIS The ARP client itself is responsible for registering its own IP ATM address information with the ATMARP server and for obtaining the IP ATM address of the desired destination system from the ARP server Entries in clients and servers ATMARP are subject to an aging process Client ATMARP entries are valid for a maximum of 15 minutes server ATMARP entries for at least 20 minutes ATMARP PDUS like the IP data packets themselves are transported in AAL5 CPCS PDUs according to the rules of LLC encapsulation see Fig
117. ifts Each phase now contains not just one data bit but an entire bit sequence In the case of CAP 16 a given signal level can represent any of 16 different values depending on its phase The information contained in a single signal amplitude thus corresponds to 4 bits If each amplitude phase state is to represent 6 bits a total of 64 different phase states are necessary The ATM cells are framed in STM 1 STS 3c containers SPEs in accordance with ITU T Recommendation G 707 ANSI T1 646 Section 7 4 At transfer rates of 51 84 Mbit s the ATM cells are transported in the payload field of the STM 0 STS 1 frame The entire payload field of the STM 0 STS 1 frame is filled with cells with the exception of columns 30 and 59 for a net bandwidth of 48 384 Mbit s The 25 92 Mbit s and 12 96 Mbit s data rates are obtained by reducing the STS 1 frame rate 51 84 Mbit s frame period of 125 us 25 92 Mbit s frame period of 250 us 12 96 Mbit s frame period of 500 us As in the case of 155 Mbit s network the maximum length of network segments at 51 84 Mbit s is 90 m plus 10 m of flexible patch cable If higher quality Cat 5 cable is used rather than Cat 3 up to 160 m can be spanned at 51 84 Mbit s The lower rates 25 92 Mbit s and 12 96 Mbit s permit segment lengths of 320 and 400 m with Cat 5 copper cabling UTP 3 cabling is used with 8 pin RJ 45 connectors IEC 603 7 that conform to the electrical specification ANSI TIA EIA 568 A
118. in UNI 3 0 and 3 1 is that the adaptation layer for signaling SAAL in UNI3 1 is compatible with the corresponding definition in Q 2931 This is not true of UNI 3 0 which uses the older SAAL specifications Q SAALI and Q SAAL2 Furthermore UNI 3 0 signaling supports neither meta signaling nor broadcast signaling TROUBLESHOOTING LocAL AREA NETWORKS 1 0 373 ATM Forum Signaling UNI 4 0 The current ATM Forum signaling specification UNI 4 0 closely approximates ITU T Recommendation Q 2931 with Q 2971 point to multipoint signaling ex tensions With the exception of the messages SETUP ACKNOWLEDGE and INFORMATION which in any case are only required in the case of transitions between narrow band and broadband ISDN networks UNI 4 0 supports all the Q 2931 and Q 2971 signaling messages In addition to the messages specific to ATM Forum UNI 3 1 4 0 also provides the messages PARTY ALERTING and LEAF SETUP REQUEST as well as the corresponding protocol states It is worth noting that UNI 4 0 was based primarily upon Q 2931 and Q 2971 rather than UNI 3 1 Furthermore as in much standardization work many of the same signaling experts were involved in the development of both ITU T recommenda tions and ATM Forum specifications Consequently with the exception of the SETUP ACKNOWLEDGE and INFORMATION messages previously mentioned UNI 4 0 can be considered to be a superset of ITU T UNI recommendations equipment conforming to the ATM Forum s U
119. in UNI cell header values as reserved These are listed in Figure 10 26 Reserved Header Bytes in UNI ATM cells ITU 1 361 1 432 Byte 1 Byte 2 Byte 3 Byte 4 Reserved for physical layer 2 PPPP 0000 0000 0000 0000 0000 0000 PPP1 Physical layer F1 OAM cell 0000 0000 0000 0000 0000 0000 0000 0011 Physical layer F3 OAM cell 0000 0000 0000 0000 0000 0000 0000 1001 IMA ICP cell 0000 0000 0000 0000 0000 0000 0000 1011 Idle cells 0000 0000 0000 0000 0000 0000 0000 0001 Unassigned cells AAAA 0000 0000 0000 0000 0000 0000 AAAO Bit available for use by the ATM layer Pus Bit available for use by the physical layer In cells with the VPI VCI value equal to 0 0 the four bits normally used to represent the PT and CLP fields are reinterpreted to distinguish between different types of unassigned and physical layer cells Cells identified by header information as physical layer cells are not passed to the ATM layer Reserved VPI and VCI Values in UNI ATM cells ITU 1 361 VPI Meta signaling channel 0000 0000 0000 0000 0000 0001 Broadcast signaling channel 0000 0000 0000 0000 0000 0010 Figure 10 25 Reserved header bytes and VPI VCI values TROUBLESHOOTING LocAL AREA NETWORKS 334 ATM lige Valuet234 Octet 1 Octet 2 Octet 3 Octet 4 Unassigned cell indication 00000000 00000000 00000000 0000xxx0 Meta signalling d
120. in each block AIS Alarm Indication Signal an AIS sent downstream to report an upstream error AIS L for SONET RDI Remote Defect Indicator sent upstream to report a reception failure downstream This occurs when frame synchronization or the data signal is lost for example CRC CRC 10 Checksum R Reserved set to 01101010 OAM F3 Cell Fields PSN PL OAM Sequential Number 8 bits modulo 256 NIC Number of Included Cells maximum value 512 MBS Monitoring Block Size maximum value 64 NMB EDC Number of Monitored Blocks EDC octets recommended value 8 EDC Error Detection Code BIP 8 Bit Interleaved Parity value calculated from the cells of the MBS TROUBLESHOOTING LocAL AREA NETWORKS 1 0 328 NMB EB Number of Monitored Blocks at the Far recommended value 8 REI Remote Error Indication number of bit parity errors in each block REI P for SONET AIS Alarm Indication Signal AIS P for SONET TP RDI Transmission Remote Defect Indicator formerly TP FERF RDI P for SONET CRC CRC 10 Checksum R Reserved set to 01101010 10 1 7 The ATM Layer The actual transport of ATM cells takes place at the ATM layer In order to provide for the varying connection quality requirements of different applica tions Quality of Service QoS parameters are first negotiated in the signaling process or provisioned by the network operator After a successful connection setup transmission begins with the ATM cell
121. ing variants UNI 3 x 4 0 share the same protocol discriminator value TROUBLESHOOTING LocAL AREA NETWORKS 1 0 362 Bit 8 7 6 5 4 3 2 1 0 0 0 0 1 0 0 1 Protocol discriminator value for Q 2931 messages Bit 8765 4321 0000 0000 Reserved to 0000 0111 0000 1000 Q 931 1 415 user network call control 0000 1001 Q 2931 user network call connection control 0001 0000 Other Layer 3 protocols to X 25 etc 0011 1111 0100 0000 National use to 0100 1111 0101 0000 Other Layer 3 protocols to X 25 etc 1111 1110 Figure 10 44 The protocol discriminator field Call Reference The call reference serves to associate Q 2931 messages with a given connection When a new connection is established all messages concerning that connection have the same call reference value When the connection has been cleared down the call reference is released and can be used again The same call reference can be used by two connections within an ATM virtual channel only if the respective connection setups take place in opposite directions The length of the call reference field is measured in bytes the default length is 3 bytes The call reference flag identifies the sending and receiving stations The station that indicates the connection always sets this flag to 0 in its messages while mes sages originating from the receiving station have the flag set to 1 The call reference value
122. inuity check functions are started and stopped by means of special activation and deactivation cells Figure 10 31 shows the format of these cells Message ID Direction of action Correlation tag PM block size PM block size B A Not used 6 bits 2 bits 8 bits 4 bits 4 bits 336 bits Figure 10 31 Function specific fields of activation deactivation cells The message ID field contains the various commands of the OAM activation deactivation cell Activate 000001 Activation confirmed 000010 Activation request denied 000011 Deactivate 000101 Deactivation confirmed 000110 Deactivation request denied 000111 The correlation tag serves to correlate commands with the corresponding re sponses The direction of action field specifies the direction of transmission of the activated OAM cells A B indicates the direction away from the activator or deactivator B A indicates transmission toward the activator The fields PM block size A B and PM block size B A specify the cell block length to be used for a performance measurement 10 1 8 The ATM Adaptation Layer AAL The AAL maps data structures of higher application layers to the cell structure of the ATM layer and provides appropriate control and management functions In order to meet the different requirements of various services four AAL types were originally defined AAL Type 1 the real time sensitive services with con TROUBLESHOOTING LocAL AR
123. ion SIGL 1 09 01 00 10 28 57 Tmr T308 Stopped Call Ref 0 1 Context 0x1f 1e40 SIGL 1 09 01 00 10 28 57 Tmr T310 Stopped Call Ref 0 1 Context 1 40 SIGL 1 09 01 00 10 28 57 Tmr T313 Stopped Call Ref 0 1 Context O 1 71e40 SIGL 1 09 01 00 10 28 57 Tmr 7322 Stopped Call Ref 0 1 Context 1 71240 03 01 00 10 28 57 Tmr T301 Stopped Call Ref 0 1 Context 1 71240 03 01 00 10 28 57 Tmr POLL Expired 09 01 00 10 28 57 Tmr POLL Started 09 01 00 10 28 57 Tmr NASP Stopped ILMI Link Establish 1 amd Fu SSCOP Signalling ILMI LANE Figure 10 62 Setting and releasing SVCs with the Agilent Technologies Advisor TROUBLESHOOTING LocAL AREA NETWORKS 1 0 401 arrive check on the switch whether the PVC is set up at all the values correctly configured and the internal path between the two switch ports is functional If you use SVCs you must make sure the signaling process that sets up your connections is working If pings do not go through between two SVC nodes the first step once again is to check the configuration of the switch and the end systems In the case of Classical IP over ATM are the systems registered on the server with their ATM and IP addresses Is the correct address of the ARP server registered on the clients If so at least the SVC setup between the clie
124. ions http www itu int itudoc itu t rec ATM Forum Specifications http www atmforum com atmforum specs The RFCs http www ietf org rfc html TROUBLESHOOTING LocAL AREA NETWORKS 1 0 394 10 2 Troubleshooting ATM In addition to protocol analyzers and cable testers for twisted pair and fiber optics troubleshooting in ATM networks involves the use of ATM switch and node management software that is able to track and display the various ATM Operations and Maintenance OAM information flows ATM contains a number of powerful OAM functions Because ATM is based on a switched architecture these integrated monitoring functions are very impor tant it is no longer possible to monitor the entire activity in a network from a single point as in traditional network technologies such as Ethernet Token Ring or FDDI Monitoring of a single ATM connection only yields information about the traffic between the two connection endpoints such as a computer system and an ATM switch port The first step in diagnosing problems in an ATM network is to monitor and analyze data obtained from operation logs and OAM statistics of the various ATM network nodes Although many of today s ATM components only support interpret or display a small proportion of the ATM diagnostics functions a protocol analyzer can be used to analyze all five OAM flows F1 F5 and determine whether they report a problem or not Agilent Advisor ATM Run Tim
125. is not done by propagating AIS or RDI on the F4 flows of the new virtual paths which would be misleading because the new virtual paths probably are not themselves faulty Instead the fault management system causes the fault condi tion to propagate upwards to the F5 virtual channel flow level at the endpoints of the faulty virtual path and VC OAM cells now carry RDI or AIS fault signals over all virtual channels that were contained in the original faulty virtual path This can obviously result in an explosion of fault indications of course but this is inevitable if full fault management is to be achieved Fault management uses two mechanisms to detect fault conditions continuity checks CC and loopback tests Continuity check cells can be inserted in the user data stream at regular intervals to provide continuous verification of the availability of a connection The ATM network nodes along the connection path can then monitor the presence of these cells If expected CC cells are not received loss of continuity LOC is signaled by AIS OAM cells The insertion of TROUBLESHOOTING LocAL AREA NETWORKS 1 0 344 CC cells is useful where user traffic is intermittent because the absence of user traffic does not necessarily mean that the virtual connection has been termi nated Without CC cells there would be no cells at all belonging to that virtual connection for perhaps long periods so if a fault had occurred during this period there wou
126. ither single longitudinal mode SLM or multilongitudinal mode MLM lasers with wavelengths of 1 310 or 1 550 nm Permissible loss is between 0 and 12 dB For long range WAN links of up to 40 km lasers with a wavelength of 1 310 nm can be used If high power SLM 500 uW or 3 dBm or MLM lasers are used at wavelengths of 1 550 nm the range can be up to 80 km Permissible loss is between 10 and 24 dB Parameter Medium range In house links 2 km Units WAN links 15 km Transmission characteristics Wavelength 1293 1334 1261 1360 Spectral width RMS width 4 14 5 MLM Laser35 LEDs nm Mean signal power 15 to 8 15 to 8 dBm Minimum extinction rate 8 2 8 2 dB Eye diagram See T1 646 See T1 646 Reception characteristics Minimum sensitivity 28 23 dBm Minimum overload 8 8 dBm Optical path power penalty 1 1 dB Figure 10 13 Single mode fiber optic parameters for 622 08 Mbit s interfaces ITU T Recommendation G 957 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 317 Using single mode fiber the attainable data rates 155 Mbit s with STM 1 0 3 622 Mbit s with STM 4 OC 12 2 4 Gbit s with STM 16 0C 48 and 9 9 Gbit s with STM 64 0C 192 In practice ATM is transported over single mode fiber optic links primarily in wide area networks Due to the steadily growing demands for range and capacity in local area networks however these Optical Parameters Uni
127. l the error condition is resolved VP AIS and VP RDI messages are always sent in cells with VCI 4 while VC AIS and VC RDI messages are sent in cells with 101 Two mechanisms are available to detect error conditions continuity checks CC and loopback tests Continuity checks continuously monitor the availabil ity of a connection To this end CC cells are periodically inserted into the user cell stream ATM network nodes along the connection path can then check for the presence of these CC cells When no more cells are received an AIS alarm for loss of continuity LOC is triggered If the ATM network components support OAM cell processing they can often locate the failure domain by analyzing the contents of the OAM flows If not the OAM flows must be captured and analyzed using a protocol analyzer 10 2 1 2 Verifying ATM Cell Transmission on the Physical Layer The analysis of ATM cell transmission parameters with the help of a protocol analyzer can also provide information about problems in the physical layer The traffic parameters to examine include corrected header ratio discarded cell ratio loss of cell delineation rate and the demux error ratio Corrected Header Ratio The corrected header ratio is the number of cells with errored but correctable headers divided by the total number of cells received This parameter is mainly TROUBLESHOOTING LocAL AREA NETWORKS 1 0 398 influenced by the bit error ratio of the
128. ld be no other mechanism to recognize this Loopback Cells Loopback cells are used to verify connections to specific sections of the ATM network There are five distinct kinds of loopback tests end to end access line inter domain network to endpoint and intra domain End to end loopback cells sent to one endpoint of a VP or VC connection are sent back to the originating endpoint The connection may cross several subnetworks or operator domains In this way the entire connection is tested from end system to end system By contrast access line loopback cells are returned by the first ATM network node that receives them This tests exactly one connection section or segment Inter domain loopback cells are reflected by the first net work node of a neighboring operator domain This makes it possible to test the connection to the neighboring network Network to endpoint loopback cells can be used by a network operator to test the connection out of the network to an endpoint in an adjacent network Finally intra domain loopback cells can be sent from a segment node to any other node in another segment within the same operator domain This tests the data flow across a certain sequence of segments within a domain Loopback cells are F4 or F5 OAM cells that contain the value 0100 in the function type field see Figure 10 30 The function specific fields are the Loopback Indication the Correlation Tag the Loopback Location ID and the Source
129. lls 4 5 and 6 indicate virtual channel segment OAM end to end OAM and resource management RM cells respectively Figure 10 28 lists the possible values of the payload type field TROUBLESHOOTING LocAL AREA NETWORKS 1 0 340 Payload type field Meaning 000 User data cell no congestion detected SDU type 0 001 User data cell no congestion detected SDU type 1 010 User data cell congestion detected SDU type 0 011 User data cell congestion detected SDU type 1 100 Segment OAM F5 cell 101 End to end OAM F5 cell 110 VC resource management 111 Reserved for future use Figure 10 28 Payload type field Cell Loss Priority and Selective Discarding of Cells To protect the network from users who produce unauthorized traffic loads or otherwise violate their traffic contract each station s data stream is monitored by the Usage Parameter Control UPC This entity analyzes and regulates the cell stream on each virtual path and virtual channel connection The UPC can act in three ways at the cell level to regulate the data stream each cell can be passed along tagged or discarded Cell passing is the normal transfer of all cells that conform to the traffic contract Cell tagging is performed on traffic that does not conform to the sustainable cell rate for one particular traffic contract type VBR 3 SBR3 When such cells are tagged their CLP value is changed by the UPC from 0 normal priority to 1
130. loss CDV etc With a pair of analyzers capable of simultaneously sending shaped traffic and analyzing received traffic bi directional measurements can be made simultaneously Asymmetrical traffic contracts can be tested without difficulty in this configuration because each analyzer can independently shape traffic according to the traffic contract for that direction The fact that it may be impossible to synchronize the clocks of the analyzers with respect to phase as well as frequency need not invalidate the most useful measurements because often the absolute delay is less important than CDV and cell loss For in service measurements however more sophisticated test equipment with at least two independent but time correlated receivers and transmitters is required Such a tester can be inserted into the lines of transmit and receive ports of the ATM element under test as shown in Figure 10 64 If supported by the network components in service tests can also be carried out with the help of special performance management OAM cells These cells are periodically inserted in the user cell streams of the connections to be monitored The measurement parameters contained in the payload of these cells sequence number user cell count time stamp cell loss counter provide information about the operational condition state of the ATM connection The ATM layer performance parameters are defined in the following sections Cell Error Ratio The cell erro
131. lue of 110 binary 6 VP VC Cells VP VC cells are common cells used for communication within ATM virtual channels VCs and virtual paths VPs VP VC cells can be of the following five types User data transport cells any VPI VCI gt 31 Metasignaling cells VPI 0 VCI 1 Broadcast signaling cells VCI 2 Point to point signaling cells VCI 5 ILMI cells for ATM network management VPI VCI 0 16 User data are assigned to a specific connection by their VPI VCI values and transport data for higher layer services in the 48 byte payload field Meta signaling cells were originally conceived to select and define signaling virtual channels Because meta signaling is not used in present day networks however these cells do not occur in actual practice Broadcast signaling cells are used to send signaling information to all network stations Network nodes that do not support broadcast signaling simply ignore all cells in the broadcast virtual channel VCI 2 VCI 5 cells are used for point to point signaling For this reason the virtual channel VCI 5 is also called the signaling virtual channel Finally ILMI cells are used for ATM network local management tasks at the UNI which include registration of new active stations with the ATM switch querying ATM MIBs or configuring network components TROUBLESHOOTING LocAL AREA NETWORKS 1 0 337 10 1 7 2 ATM Connections In order to manage the wide variety of
132. m MM 1 300 nm 622 Mbit s SW MMF 300 10 1 12 ATM Standards All the important standards for ATM technology are developed by the ITU T the ATM Forum and the IETF A selection of the most important ITU T ATM Forum and IETF Recommendations is listed here A complete list of ITU T and ATM Forum standards can be found in the Appendix or at the web sites listed below 10 1 12 1 Technical Working Group Control Signaling Data Exchange Interface Frame Based ATM LAN Emulation ATM Forum Standards Approved Specifications B ICI 2 0 integrated specification PNNI Version 1 0 Security Signaling Addendum Specification af bici 0013 003 af cs 0116 000 UNI Signaling 4 0 Security Addendum af cs 0117 000 Data Exchange Interface version 1 0 Frame Based ATM over Sonet SDH LANE Servers Management Spec v1 0 LANE v2 0 LUNI Interface LAN Emulation Client Management Specification Version 2 0 LAN Emulation over ATM Version 2 LNNI Specification Multi Protocol Over ATM Specification v1 0 Multi Protocol Over ATM Version 1 0 MIB Multi Protocol Over ATM Specification Version 1 1 Multi Protocol Over ATM Version 1 0 MIB af dxi 0014 000 Approved Date Dec 1995 May 1999 May 1999 Aug 1993 af fbatm 0151 000 July 2000 af lane 0057 000 Mar 1996 af lane 0084 000 July 1997 af lane 0093 000 af lane 0112 000 Oct 1998 Feb 1999 af
133. mation elements must be sent in the following order Protocol discriminator Call reference Message type Message length Other information elements Information elements of variable length can be sent in any order Duplicate Information Elements If an information element is repeated in a message in which repetition of the information element is not permitted only the contents of the information element appearing first shall be handled All subsequent repetitions of the information element are ignored Mandatory Information Element Missing When a message other than SETUP RELEASE or RELEASE COMPLETE is received that lacks one or more mandatory information elements no action shall be taken on the message and no state change should occur A STATUS message is then returned with cause No 96 Mandatory information element is missing When a SETUP message is received that lacks one or more mandatory informa tion elements a RELEASE COMPLETE message is returned with cause No 96 Mandatory information element is missing TROUBLESHOOTING LocAL AREA NETWORKS 1 0 413 Mandatory Information Element Content Error When a message other than SETUP RELEASE or RELEASE COMPLETE is received in which one or more mandatory information elements have invalid contents no state change occurs A STATUS message is returned with cause No 100 Invalid information element contents When a SETUP message is received
134. n existing B ISDN connection and is available for use Connection clear down is normally initiated by a RELEASE command except in the following three cases a If a SETUP message is received without a mandatory information element the receiver may refuse the connection setup by sending a RELEASE COM PLETE and Cause 96 Mandatory information element is missing b If a SETUP message is received with an incorrect mandatory information element the receiver may refuse the connection setup by sending a RE LEASE COMPLETE and message 100 Invalid information element con tents c After VPI VCI negotiation has failed the system that initiated the call may terminate the connection setup with a RELEASE COMPLETE Connection Clear Down by the User In all cases except the three mentioned previously sending a RELEASE com mand begins the connection clear down If the user initiates the connection clear down the user s timer T308 is started when the RELEASE message is sent The first time T308 expires the user repeats the RELEASE command and starts T308 again If no RELEASE COMPLETE is received from the network in answer to the second connection clear down attempt the user considers the virtual channel concerned as out of order and enters the null state If the user receives a RELEASE COMPLETE from the network before T308 expires it clears down the virtual channel releases the call reference and enters the null state TROUBLESH
135. n original specification developed by IBM and is derived from Token Ring technology The specified maximum segment length for all three cable types is 100 m 90 m plus 10 m for patch cables The cables must also conform to the values specified in EIA TIA 568 A or ISO IEC 11801 for attenuation and near end crosstalk NEXT The connectors are RJ 45 for Cat 3 UTP or STP MIC for shielded twisted pair 10 1 6 7 ATM and DXI Interfaces The Data Exchange Interface DXT is a simple transmission protocol developed for interfaces such as V 35 EIA TIA 449 530 or EIA TIA 612 613 HSSD Origi nally DXI was conceived as a data tributary protocol for metropolitan area networks SMDS DXI SMDS Interest Group Technical Specification SIG TS 005 1993 For lack of an ATM specific DXI to connect their systems to ATM networks many manufacturers began by implementing the simpler SMDS DXI TROUBLESHOOTING LocAL AREA NETWORKS 1 0 324 as interface protocol The ATM Forum subsequently drafted a DXI variant tailored to the requirements of ATM af dxi 0014 The functions of ATM DXI are divided into user end processes Data Terminal Equipment or DTE Modes 1a and 1b and network end processes Data Commu nications Equipment or DCE Mode 2 In Mode 1a user interfaces support up to 1 023 virtual connections with AAL5 Protocol Data Units or PDUs and data packets Service Data Units or SDUs of up to 9 232 bytes The DTE SDUs which correspond to AAL5
136. nd and conventional LAN topologies such as Ether net and Token Ring on the other When the LANE protocol is used in an ATM network segment every ATM end system in this network can communicate directly with every other Ethernet or Token Ring station connected through an ATM router Furthermore selected stations in the LAN can be assigned to an ATM LANE workgroup That is end systems can be grouped regardless of their location or network interface into virtual LANs VLANs MPOA provides transparent interconnection of several ATM LANE segments Note that because LANE is a Layer 2 link layer protocol emulated LANs ELANSs are either Ethernet or Token Ring based not a mixture of both Layer network layer routing is necessary to provide communication between Ether net and Token Ring based ELANS via for example the MPOA protocol am 10 TROUBLESHOOTING LocAL AREA NETWORKS 300 XJS YOUMS WLY X Dujeub g 4 ti m H saul pesee Bury u yoL La sa
137. nd loopback cables TROUBLESHOOTING LocAL AREA NETWORKS 1 0 397 splitters Note that the transmitter ports of many analyzers equipped with single mode lasers When actively monitoring multimode fiber components the transmitters must be fitted with 10 dB attenuators in order to avoid overdriving the receiver electronics This does not result in damage to the multimode receivers however Because lasers still emit a small amount of light when off that is when sending 0 sensitive multimode receivers misinterpret this as a 1 and never see the difference between 0 and 1 Attenuators can correct this 10 2 1 1 Analyzing Physical Layer OAM Information Flows ATM s integrated error detection mechanisms are contained in the OAM infor mation flows F1 to F5 Flows F1 F3 yield information about the operating state of the SDH SONET transport structure while FA and F5 contain the corre sponding ATM layer data F4 concerns ATM virtual path connections VPCs the virtual channel connections VCCs The error management function in ATM is based on two types of alarms Alarm Indication Signal AIS and Remote Defect Indicator RDI The AIS is sent by the VC or VP node which recognizes the error condition to all upstream nodes so long as the error condition persists Immediately after the AIS an RDI signal is sent upstream to the end nodes of the connections affected These signals are also sent periodically unti
138. nd receiving by a green LED SDH SONET level problems by a yellow LED and complete signal loss by a red LED The ATM interface can also be tested using a fiber loop or a UTP loopback connector If the loopback connection is in order but the loopback test fails the firmware may not be loaded correctly or may not detect the hardware If the TROUBLESHOOTING LocAL AREA NETWORKS 1 0 396 loopback test is successful the ATM interface the firmware in order In this case you must test whether the physical layer connection exists between the network interfaces in the problem domain This can be done by inserting a protocol analyzer in pass through mode into the connection between the two nodes The analyzer may be inserted directly into the ATM connection active monitoring or may be connected in passive mode by means of optical power 1 1 Rx 2 Tx 2 Rx 3 3 4 4 ATM 5 5 ATM switch 6 6 7 Rx 7 Tx 8 Rx 8 Tx ATM Direct CONNECTION 1 Tx 1 Rx 2 Tx 2 Rx 3 3 ATM 4 end system 5 5 end system 6 6 7 Rx 7 Tx 8 Rx 8 Tx ATM Crossover Tx Rx 8 Rx ATM LooPBACK Figure 10 60 UTP Cat 5 pin assignments for ATM direct connection crossover a
139. ned to be a legal value that is one belonging to an existing virtual connection the cell with the corrupt header would be misinserted into this other virtual connection and of course lost from the Examine bit by bit 1 valid HEC pattern Incorrect HEC fields ALPHA successive incorrect HEC fields Examine cell by cell DELTA successive correct HEC fields DELTA 6 ALPHA 7 Figure 10 24 State diagram of the HEC cell synchronization algorithm TROUBLESHOOTING LocAL AREA NETWORKS 1 0 332 connection to which it really belongs However if new VPI VCI value was not supported the switch would not forward it but would discard it because no translation table entry for it would exist Note that some interfaces restrict the use of the HEC to detecting errors but not correcting them this is because certain types of physical layer scrambling can give rise to error multiplication making reliable correction impossible In addition to bit error detection and correction the HEC field also allows receiving stations to synchronize with the beginning of the cell This process is called cell delineation The following algorithm is applied in order to detect the beginning of an ATM cell in a bit stream In the HUNT state the incoming signal is analyzed bit by bit to determine whether it could be part of an HEC pattern As soon as a potential HEC pattern has been found the receiver ch
140. nsidered equivalent See Figure 10 27 Physical Layer OAM Cells In the cell based physical layer special cells can be inserted up to once in every 27 cells to transmit operation and maintenance information concerning the physical layer These are called PL OAM cells When received these cells are used by the physical layer and not passed along to the ATM layer Their purpose is to convey some of the information such as alarms normally carried by the overhead of frame based physical layers such as SONET TROUBLESHOOTING LocAL AREA NETWORKS 1 0 336 IMA Control Protocol ICP Cells These are special control cells used with implementations of the ATM Forum s Inverse Multiplexing for ATM specification VP OAM and VC OAM Cells VP OAM and VC OAM cells are used to transport the F4 and F5 maintenance flows respectively This information allows the network to monitor and test the capacities and availability of ATM virtual paths and virtual channels VP OAM cells use reserved values of VCI 3 and 4 for each VPI value for segment and end to end F4 OAM flows respectively VC OAM cells use a reserved value of PT field 100 binary 4 and 101 binary 5 for each value of VPI VCI for segment and end to end F5 OAM flows VP VC RM Cells These are resource management cells used so far for implementations of the available bit rate ABR traffic type VP RM cells have a reserved value of VCI 6 VC RM cells have a reserved PT field va
141. nts and the ARP server should succeed If the clients are still unable to communicate with one another make sure the ILMI stack is active on the switch and on the clients If the problem still persists you must monitor and analyze the signaling process with a protocol analyzer Finally the switch ports and the switch configuration should be examined in detail 10 2 2 2 Verifying ATM Performance Parameters If the connections are set up successfully but problems persist on the applica tion layer and during data transfers the next step is to examine the ATM performance parameters Oftentimes functional tests of ATM connections such as sending pings between two nodes setting up and clearing down SVCs etc seem to show that everything is working fine but once application data is transmitted at higher traffic loads problems arise Reasons for this type of behavior can be excessive cell loss or cell delay values traffic contracts that provide insufficient bandwidth and therefore cause cells to be discarded or simply an overloaded switch Measurements that determine these types of ATM performance parameters can be made either in service or out of service Out of service measurements are performed using special out of service test cells de fined in ITU T Recommendation 0 191 5 bytes 4 bytes 4 bytes 37 bytes 1 byte 2 bytes eos time stamp not used TCPT CRC10 header number MG PM Test Cell Payload
142. odsuey ejep yodsuey ejep SLYS SeyAq 881 p9 99110 1014 peunjonus sq sige wat 59 WAL yaaviang K pe ee ees J 155 99 4970 ouso espo 4ese ANS NS 150 27 0 ANS NS jio A glepeeu Sd Jepeeu fad dvs Japesy VS eg m m IE i i Sy Sy NOlLVOl lddy pog pog 19 8 nas wy 198 nas 4 nas Ex ejeq oipny oeptA OLLOL 00 eur e Jun WV Figure 10 33 Structure of AAL 1 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 349 ATM network transport not only the data but also the clock information For this reason the AALI protocol is capable of transporting both continuous bit streams and byte structured data such as data based on an 8 kHz sampling interval Lost or erroneous data is not repeated or corrected Events such as cell loss or the transmission of incorrect service data units 5005 loss of synchro nization or clock signal buffer overflow or the occurrence of invalid AAL header information AAL Protocol Control Information or AAL PCI are passed from the user layer to the management layer The AAL1 is composed of two sublayers the Segmentation and Reassembly Sublayer SAR and the Conver gence Sublayer CS The da
143. of a CPS packet that extends across a CPS PDU boundary indicates a transmission error in the header If the offset is less than 47 processing continues at the byte indicated by the offset Figure 10 35 AAL2 error messages to the layer management The history of AAL2 is complex It was originally proposed by the ITU T in the late 1980s as the adaptation layer to handle Class B services variable bit rate connection oriented services specifically video then withdrawn It returned through the initially separate then later joint efforts of the ATM Forum and the ITU T in the latter half of the 1990s to solve the problems inherent in AAL1 for low and sometimes variable bandwidth delay sensitive services such as com pressed voice and the inefficiency of AAL5 for handling very short data packets It was at one point nearly called AAL6 by the ATM Forum before it was concluded that it really did fulfill the original requirements for AAL2 defined a decade earlier At the time of this writing therefore two service types have been defined for AAL2 trunking and SAR ITU T Recommendation I 366 2 describes the AAL type 2 Service Specific Convergence Sublayer for Trunking the circuit emulation service used for the transportation of narrow band delay sensitive traffic AAL2 also provides the means to handle 1 TROUBLESHOOTING LocAL AREA NETWORKS 352 IN LV vs uaAviang AT8N3SSV3 NOILVIN3
144. of cells received valid or not This quantity is also influenced by the bit error ratio of the transmission path Loss of Cell Delineation Rate The loss of cell delineation rate is the number of cell synchronization losses over a certain time interval Mean Loss of Delineation Duration The mean loss of delineation duration is defined as the number of missing Cell Received Events CRE due to cell synchronization loss within a given time interval divided by the total number of expected CKE events in this time interval see Figure 10 66 Demux Error Ratio The demux demultiplex error ratio is the number of all correctly transmitted cells containing an invalid VPI value divided by the total number of correctly transmitted cells 10 2 1 3 Causes of Problems in the Physical Layer Physical layer problems in ATM networks may have a variety of causes Many problems occur on the connections to ATM switches where a change of the physical transmission media is required connectors fiber switch port Trans mission errors can arise due to aging humidity dust or material flaws Further more the signal quality of the cabling determines the bit error ratio in the transmission framework SDH SONET T3 etc and consequently the per formance of the ATM network 10 2 2 Troubleshooting the ATM Layer If no faults can be detected in the physical layer the ATM layer must be examined This involves checking whether the required PVC
145. on of receipt SDP PDUs Such confirmation is sent by the receiver in a STAT PDU After the receiver has analyzed the sequential numbers of the PDUs received it can send USTAT PDUs if necessary to report lost PDUs indicating the number range missing Data can also be sent in an unassured mode using Unnumbered Function Description Establishment Request initialization Request acknowledgement 0010 Release Disconnect command END 0011 Disconnect acknowledgement ENDAK 0100 Resynchronization Resynchronization command RS 0101 Resynchronization acknowledgement RSAK 0110 Reject Connection reject BGREJ 0111 Recovery command ER 1001 Recovery acknowledgement ERAK 1111 Assured data transfer Sequenced connection mode data SD 1000 Transmitter state information with POLL 1010 request for receiver state information Solicited receiver state information STAT 1011 Unsolicited receiver state information USTAT 1100 Unacknowledged Unnumbered user data UD 1101 data transfer Management Unnumbered management data MD 1110 data transfer Figure 10 39 SSCOP PDUs and their functions TROUBLESHOOTING LocAL AREA NETWORKS 1 0 357 Data UD PDUs Figure 10 39 lists the 15 types of SSCOP PDUs and their functions The SSCOP Timers The SSCOP protocol defines four timers that control the protocol process These are the POLL KEEP ALIVE NO RESPONSE and CONNECTION CONTROL CC timers
146. onnections across the ATM network to the end system called There the NNI TROUBLESHOOTING LocAL AREA NETWORKS 1 0 374 connection is translated back into a UNI connection However B ISUP is unable to fulfill functions such as Bandwidth management Route selection Routing table maintenance OAM process control for existing virtual connections The implementation of these functions is left to the component manufacturers or the operators of B ISDN networks The B ISUP protocol builds on Message Transfer Protocol Version 3 MTP 3 specified especially for ATM whose data packets are transported in turn by means of the NNI SAAL layer User Switch A Switch B Switch C User Start SETup T303 Start T40b ca PROC NE cO Start T40b Stop 0 1 T303 Start Pa T310 Stop T40b pROC Stop T40b ACM ee m 6 a Start Tob a Start puer s Stop Stop T9b 1313 SETUP NECT Stop T9b Stop T310 CONNECT Figure 10 51 Connection setup tear down in the UNI NNI network PNNI Signaling PNNI Private Network to Network Interface signaling consists of two proto cols a topology protocol which distributes information about the network topology to the individual network stations and a signaling protocol which is necessary for connection setup between PNNI nodes The signaling pro
147. ons can be set up The AAL3 4 protocol is thus suitable for transportation of non connection oriented communications services such as SMDS CBDS Switched Multi Megabit Data Service Connectionless Broadband Data Service metropolitan area networks or Frame Relay Like AAL Type 1 the AAL3 4 protocol consists of two sublayers the Segmentation and Reassembly SAR sublayer and the Convergence Sublayer CS although the CS includes both a Common Part Convergence Sublayer CPCS and a Service Specific Convergence Sublayer SSCS The variable length data packets 1 to 65 535 bytes of the application building on AAL3 4 are first padded to an integer multiple of 32 bits to permit an efficient hard ware based implementation of the AAL processes and a header and trailer are added The resulting CS PDU is then split into 44 byte fragments each of which receives another header and trailer and is then passed to the ATM layer as a 48 byte SAR PDU In reality this AAL is now only used for legacy SMDS services which are themselves obsolescent AAL5 has overtaken AAL3 4 for interworking with Frame Relay and for the support of almost all other services because it is so much more efficient 10 1 8 5 Type 5 AAL Type 5 was invented by the ATM Forum in the beginning of the 1990s to simplify the handling of data over ATM The ITU T later adopted it and it is now defined in ITU T Recommendation 1 363 5 AAL5 corresponds to a highly simpli fied implement
148. ored They are identified as F4 cells by the reserved VCI value 3 for F4 segment cells or 4 for F4 end to end cells F5 OAM cells have the same VPI TROUBLESHOOTING LocAL AREA NETWORKS 1 0 342 values the user cells of the virtual channel connection to which they are related They are identified by the payload type field PT The PT value 100 decimal 4 denotes F5 segment OAM cells 101 decimal 5 indicates an F5 end to end OAM cell End to end OAM cells must pass unchanged through all net work nodes between the two connection endpoints Only the endpoints may remove them from the cell stream Segment OAM cells must be removed from the stream at the end of the given segment There are five types of ATM layer OAM cell fault management APS coordination protocol performance management activation deactivation cells and system management The OAM cell type is indicated by the first 4 bit field in the OAM cell payload F4 F5 OAM Cell ATM cell header ATM info field lt lt 4 bits 4 bits 45 bytes 6 bits 10 bits G vpi VCI PT HEC OAM Fufictierespeeifie reserved 10 cell type type fields 8 G X x 9 Xo 1 Same value VCI 3 Segment as User cells VCI 4 End to End F4 OAM cell 0001 Error Management 0000 AIS PT 100 Segment F5 OAM end Fr 100 End to End 0001 FERF 0010 Loopback
149. ork for Integrated Services and RSVP over ATM E Crawley L Berger 5 Berson Baker M Borden J Krawczyk August 1998 RFC2383 ST2 over ATM Protocol Specification UNI 3 1 Version M Suzuki August 1998 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 393 RFC2417 Definitions of Managed Objects for Multicast over UNI 3 0 3 1 based ATM Networks C Chung M Greene September 1998 Obsoletes RFC2366 RFC2492 IPv6 over ATM Networks G Armitage P Schulter M Jork January 1999 RFC2512 Accounting Information for ATM Networks K McCloghrie J Heinanen W Greene A Prasad February 1999 RFC2515 Definitions of Managed Objects for ATM Management K Tesink Ed February 1999 Obsoletes RFC1695 RFC2601 ILMI Based Server Discovery for ATMARP M Davison June 1999 ASCII RFC2684 Multiprotocol Encapsulation over ATM Adaptation Layer 5 D Grossman J Heinanen September 1999 ASCII Obsoletes RFC 1483 RFC2761 Terminology for ATM Benchmarking J Dunn C Martin February 2000 RFC2823 PPP over Simple Data Link SDL using SONET SDH with ATM like framing J Carlson P Langner E Hernandez Valencia J Manches ter May 2000 ASCII RFC2844 OSPF over ATM and Proxy PAR T Przygienda P Droz R Haas May 2000 RFC2955 Definitions of Managed Objects for Monitoring and Controlling the Frame Relay ATM PVC Service Interworking Function K Rehbehn O Nicklass G Mouradian October 2000 ITU T Recommendat
150. otocols can be transported over a single VCC When a new LE Client is added to an ELAN the LE Client first sets up a Configuration Direct VCC to the LE Configuration Server in order to register as a member of a particular ELAN At the same time it may optionally use the LE configuration protocol to negotiate a variety of parameters addresses name of the ELAN maximum frame size Next the Control Direct VCC to the LE Server is set up At this point the LE Client should be in possession of all information necessary for participation in the LAN emulation service including the LE Client Identifier LECID the LAN type 802 3 Ethernet 802 5 Token Ring etc If necessary the LE Server may TROUBLESHOOTING LocAL AREA NETWORKS 1 0 380 also set up a one way point to multipoint Control Distribute VCC to several LE Clients Point to multipoint connections are supported in LANE Version 2 0 but not in LANE 1 0 Finally communication with the BUS is opened over a Data Direct VCC The LE Client is then ready for operation under the LAN emulation service see Figure 10 56 The various connections are set up and cleared down by means of the ATM Forum UNI signaling protocols UNI 3 0 UNI 3 1 UNI 4 0 In LAN emulation data flows either between the LE Client and the BUS or between one LE Client and another If an LE Client needs to send a packet to a station whose ATM address is unknown it first sends an LE ARP message to the LE Server
151. path The information flow F5 is used for segment or end to end management at the virtual channel level in this case end to end refers to the entire communication channel between the two endpoints for the complete virtual channel which may be longer than the virtual path The purpose of F4 and F5 OAM cells is to provide the measurement data necessary to monitor the availability and performance of a given virtual channel or virtual path Figure 10 29 lists the individual OAM functions of the ATM layer OAM function Use 15 Alarm Indication Signal Error reports in the direction of transmission RDI Remote Defect Indication Error reports from the receiving system Continuity check Continuous monitoring of cell transport Loopback Monitoring connections as needed Localizing errors Testing connections before making a link available Forward performance monitoring Estimating performance in the direction of transmission Backward reporting Returning results of forward performance monitoring Activation Deactivation Activating or deactivating performance monitoring and continuity checking System management Managing by the given end systems APS ATM protection switching Figure 10 29 ATM layer OAM functions OAM Cell Format The F4 OAM cells that transport management and monitoring information for virtual path connections have the same VPI value as the user cells on the virtual path being monit
152. ported in specially indicated physical layer OAM cells PL OAM TROUBLESHOOTING LocAL AREA NETWORKS 1 0 326 OAM F1 F3 for PDH Based ATM Systems In PDH networks F1 and F3 information can be transported in the PDH header The frame alignment byte of the header is analyzed for F1 functions while F3 parameters are taken from the remaining header bytes There is no provision for an F2 flow OAM F1 Cell Based Physical Layer In cell based physical layer networks the OAM information flows are transmit ted by means of special physical layer OAM cells PL OAM cells As described in L 432 2 evaluation of OAM F1 and is provided for but there is no provision for an F2 flow The corresponding parameters are communicated as part of F3 PL OAM cells may be inserted in the cell stream no more than once every 27 data cells and no less than once in 513 cells The F1 cell contains OAM parameters for the regenerator section F3 cells are used to monitor the trans OAM flow Byte 1 Byte 2 Byte 3 Byte 4 Byte 5 F1 0000000 0000000 0000000 000001 1 Valid HEC 0101110 F3 0000000 0000000 0000000 0001001 Valid HEC 0110101 Because the PL OAM cells are not passed to the ATM layer the field values are of no significance as ATM cells Figure 10 19 PL OAM cell headers Bye 1 R Byte 17 R Byte 33
153. r National number Network specific number Subscriber number Abbreviated number The format of the address types abbreviated number and network specific number is different from case to case depending on the network Network specific numbers can be used for administrative services such as operator numbers Abbreviated number addresses contain a shortened form of a complete ATM address and may be defined by the network operator according TROUBLESHOOTING LocAL AREA NETWORKS 1 0 368 to the internal structure of the network other address types use of the following address formats Unknown SDN Telephony Numbering Plan E 164 150 NSAP Addressing ISO 8348 AD2 Private Numbering Plan The address format unknown is used when the user or network does not know any address format The use of the ISO NSAP address format is optional and takes the ATM address type unknown 10 1 9 4 Connection Clear Down Before describing the processes involved in clearing down a connection we must define the three states Connected Disconnected and Released An ATM virtual channel is in the Connected state when it forms part of an active B ISDN connection The virtual channel is in the Disconnected state when it is not part of such a connection but is not available for any other connection When an ATM virtual channel is in the Released state then it is not part of a
154. r incorrectly to the VPI VCI as an address The value of the VPI VCI field changes as cells of a given virtual connection pass through ATM switches The translation of the VPI VCI label value is determined by the contents of a translation table whose contents are determined by the signaling process in the case of switched virtual connections SVCs network manage ment provisioning in the case of semi permanent virtual connections PVCs or a label distribution protocol in the case of multiprotocol label switching MPLS An ATM virtual channel VC refers to a two way transmission link for ATM cells although both directions may not always be used All cells in a given virtual channel have the same VCI value in a given link Several ATM virtual channels may be carried within an ATM virtual path VP As with the virtual channel all cells transported by a given ATM virtual path are identified by a particular VPI Rather than the physical channels and paths familiar from older telecommuni cations techniques the ATM concept of virtual paths permits substantially more efficient use of available bandwidth vc ve P vc Communication medium Figure 10 23 Virtual channels VCI and virtual paths VPI The Payload Type Field PT Three bits of the ATM header are used to identify the type of data the cell payload contains This makes it possible to distinguish betwe
155. r ratio is the number of invalid cells divided by the sum of the number of successfully transmitted cells and the number of invalid cells Suc cessfully transferred cells tagged cells and errored cells contained in severely errored cell blocks are excluded from the calculation of cell error ratio 403 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 f Test 1 Test cell insertion Test ATM testing Intrusive Test Test cell removal t Passive cell stream monitoring ATM testing system n T Passive cell stream monitoring e d Clock timing common to both ports under test Non Intrusive Test Figure 10 64 Intrusive and non intrusive tests in ATM networks TROUBLESHOOTING LocAL AREA NETWORKS am 10 404 Cell Loss Ratio The cell loss ratio is the
156. ration to be carried out automatically a separate protocol for the exchange of MIB parameters was created called the Integrated Local Management Interface ILMI Because most network applications today build on the Internet Protocol IP and few native ATM applications are available to date ATM is also able to serve as a fully transparent transport layer for IP However the Internet Protocol was originally designed for connectionless Ethernet networks and therefore uses the broadcast principle for a number of functions In the Classical IP over ATM protocol defined for ATM networks broadcasts cannot easily be translated to connection oriented ATM this is possible only with the LAN Emulation LANE protocol Nonetheless IP can still be transported transparently over pure ATM networks Address resolution is performed by specially defined ATM Address Resolution Protocol ATMARP and Inverse ATM Address Resolution Protocol InATMARPD functions other broadcast types however are not supported 10 1 2 Heterogeneous LAN Environments ATM is used not only in new networks planned exclusively on the basis of ATM but often in combination with evolved existing network structures In order to integrate ATM smoothly in such heterogeneous networks the protocols LANE and Multiprotocol over ATM MPOA have been defined for ATM These proto cols permit complete transparent interoperability between ATM networks and end systems on the one ha
157. reams for transport which are then reassembled into the original stream at the receiving end This makes it possible to transport high bandwidths over several bundled data links of lower capacity For example bandwidths of 4 to 34 Mbit s can be realized by bundling the appropriate number of 1 links though it becomes uneconomical to consider bundles or link groups with more than about 6 8 links The ATM cells are transported one at a time over each of the available lines in turn For every M cell a special IMA Control Protocol ICP cell is inserted The ICP cell together with the corresponding ATM user data cells comprise an IMA frame Each ICP cell contains a Link Identifier LID indicating the individual line an IMA frame sequence number and various information fields used to monitor and synchro nize the transmission Figure 10 9 9 bytes 261 bytes STM 1 e SOH 1 AU 4 PTR I eee VC 4 1 i B3 oy SOH G1 F2 Y STM 1 VC 4 24 53 bytes Figure 10 10 Transportation of cells in the 1 transport module TROUBLESHOOTING LocAL AREA NETWORKS 1 0 314 10 1 6 3 ATM in SDH and SONET Networks The transportation of ATM cells over SDH and SONET networks ATM mapping is specified in ITU T Recommendation G 707 and ANSI T1 105 respectively
158. resses summary addresses correct Once again the fastest way to diagnose the problem is to monitor the PNNI messages during connection setup with a protocol analyzer 10 2 4 Cabling Problems Cabling problems are also very common in ATM networks Typical causes include bad splices low quality cables wiring faults excessive segment length and for UTP incorrect characteristic impedances or noise due to electromag netic interference caused by air condition systems photocopiers pagers eleva tors or production environments These types of problems are discussed in detail in Chapter 6 10 2 5 Problems with ATM Interface Cards The typical symptoms of defective interface cards in ATM networks are high rates of cell errors or complete loss of cell synchronization Because of ATM s connection oriented architecture it is easy to determine whether the problem is caused by a network interface card Starting in the middle of the affected connection the number of cells received is compared with the number of cells transmitted at each ATM interface If the numbers do not match the QoS parameters of the given interface must be examined If no restrictions can be detected a loopback test shows whether or not cells are being lost due to NIC problems When changing ATM interface cards care must be taken due to the high temperatures at which these cards normally operate Either wait until the card cools or avoid touching the chips on the card
159. riation Figure 10 66 One point CDV TROUBLESHOOTING LocAL AREA NETWORKS 1 0 406 the actual arrival time that is y 7 c a where the reference arrival time is defined as follows T ifc gt a T in all other cases Two Point Cell Delay Variation The two point cell delay variation v for cell k between MP1 and MP2 is defined as the difference between the actual cell delay x and the reference delay 9 between the two measurements points v d The actual cell delay x is defined as the difference between the actual cell arrival time at MP2 a2 and the actual arrival time at MP1 al that is a2 al The reference cell delay 9 between 1 MP2 equals the actual cell delay of cell 0 between the two measurement points MP1 MP2 Vy XS di Bg eec Actual arrival time of cell k at MP1 Basti Actual arrival time of cell k at MP2 MP1 Measurement point 1 E Actual cell delay MP2 Measurement point 2 di ese Actual cell delay of cell 0 Two point cell delay variation Figure 10 67 Two point CDV Figure 10 68 shows an overview of the ATM layer performance parameters for the various QoS classes that can be attained in ATM wide area networks with the reference diameter of 27 500 km TROUBLESHOOTING LocAL AREA NETWORKS 1 0 407 QoS Class Limit Default Stringent Toler
160. rity bits determined on the basis of the BIP 16 code in the corresponding forward monitoring cell Performance monitoring is done over blocks of cells the size of the blocks N being related to the peak cell rate of the virtual connection being monitored A forward performance monitoring OAM cell is sent after each N user cell provid ing information about the cell block At the receiving end the same calculation is performed on the cells of the block and the results compared with the informa tion in the forward performance monitoring OAM cell The results are reported to network management and optionally via a backward reporting OAM cell to the originating end The reason for having both forward performance monitoring and backward reporting cells is that the endpoints of the monitored virtual path or virtual channel may lie in the domains of different network operators Normally one TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 346 service provider does not have access to the information the network manage ment system of another network operator so the backward reporting cell pro vides the means to transfer the results back across the domain boundary boundaries between the relevant network operators When no domain bound aries are crossed backward reporting cells need not be sent because network management can access all nodes in their networks Activation and Deactivation of OAM Functions The performance monitoring and cont
161. s wall jacks MAUs hubs bridges routers Misconfigured ATM interface card interrupts drivers timers Misconfigured routing protocol entries address tables mapping tables subnet masks default gateways routing tables timers PNNI Hello protocol on the PNNI interface is not active PNNI Misconfigured peer group leader PGL is not active or no designated parent LGN PNNI Misconfigured PNNI port parameters cell rate cell transfer delay bit rate PNNI PNNI addresses prefixes or summary addresses are incorrect PNNI The PNNI Routing Control Channel SVCC RCC is inactive PNNI Topology information on the switch port is incomplete or outdated PNNI Uplinks to neighboring peer groups are inactive or not defined PNNI Wrong route selection due to incorrect ATM addresses for the destination node PVC not set up invalid VCI Faulty physical router or switch installation cables connectors plug in modules are loose backplane connections are miswired Problems on the physical layer cabling connectors ATM port Switch is overloaded Too many nodes along the transmission path of a VPI VCI connection Traffic contract is exceeded cells are being discarded Figure 10 73b The most frequent causes of trouble in ATM networks TROUBLESHOOTING LocAL AREA 5 1 0 1 Index of chapter 10 A AAL Type 0 347 AAL Type 1 347
162. s and Maintenance OAM 1 325 Physical Medium Dependent PMD sublayer 304 PNNI Private Network to Network Interface Signal 415 PRESYNC state 332 Principles of ATM 302 0 2931 message format 360 Quality of Service 005 parameters 339 R Remote Defect Indicator RDI 316 Reserved header bytes and VPI VCI values 333 Route selection in MPOA networks 383 TROUBLESHOOTING LocAL AREA 5 1 0 5 5 SAAL connection setup 358 Segmentation and Reassembly Sublayer SAR 349 Selected Multicast Server SMS 377 Selective discarding of cells 340 Self Synchronizing Scrambling SSS 309 314 332 Service classes and AAL types 347 Signaling ATM Adaptation Layer SAAL 355 Single Longitudinal Mode SLM laser 316 SSCOP error messages 358 SSCOP timers 357 Switched Multimegabit Data Services SMDS 301 Synchronous Payload Envelope SPE 305 314 Timers for signaling processes 370 Traffic shaping 340 Transmission Convergence TC sublayer 304 Troubleshooting ATM 394 Troubleshooting the ATM layer 399 Two point cell delay variation 406 U Unassigned cells 305 335 UNI cell header 329 UNI signaling 410 V Verifying ATM cell transmission 397 Virtual Channel Connections VCCs 328 337 Virtual Path Connections VPCs 328 338 VP OAM and VC OAM cells 336 VP VC cells 336 VP VC RM cells 336
163. s and SVCs are working and whether the ATM addresses are correct The OAM PM cell streams are then monitored using protocol analyzers to determine whether the ATM TROUBLESHOOTING LocAL AREA NETWORKS 1 0 400 layer traffic conforms to the traffic contract parameters for the connections applications in question Parameters to examine include User cell rate Cell loss ratio CLR Cell transfer delay CTD Cell delay variation CDV Number of cell sync losses Number of cells with corrected headers Number of cells that violate the traffic contract non conforming cells or NCC 10 2 2 1 Verifying PVCs SVCs and Addressing The first step in troubleshooting the ATM layer is to verify whether the ATM connections are working at all Configuring one ATM node to send a constant stream of pings to another can do this If the ping does not get through to its destination use the activity LEDs on the ATM interfaces or a protocol analyzer to determine whether the cells leave the initiating system reach and leave the switch and arrive at the interface of the destination node If no ping packets Internet Advisor ATM Run Time Signalling Results ls File Run View GoTo Setup Window Help a mi N 4j Signalling Statistics Total Messages Tz Rz Link Release 1 Calls Progressing Calls Rejected Signalling State Informat
164. s marked by the control symbol pair sequence TT In TAXI the JK sequence is used as an idle symbol and is sent when there are no assigned cells to send unassigned and idle cells are not used in the TAXI interface this is a truly asynchronous interface In case of noise interfaces resynchronize only when the next idle signal is received If higher cell losses are tolerable fewer idle symbols may be required but not less than one every 0 5 seconds The MIC connector described in ISO 9314 3 customary in FDDI networks is also used in TAXI The TAXI interface is now obsolete as STM 1 OC 3 interfaces have become commonplace in the local area offering a higher bandwidth 10 1 6 6 ATM Cell Streams at 25 6 Mbit s Originally developed to allow the economical connection of workstation com puters to ATM networks but more recently associated with ATM over ADSL modem interfaces a further specification for a 25 6 Mbit s ATM interface was developed independently of the cell based physical layer interfaces described previously The ATM Forum specification calls for Cat 3 100 Q UTP unshielded twisted pair 120 Q Cat 4 ISO IEC11801 or 150 Q STP shielded twisted pair As for the TAXI interface the bit stream is 4B 5B encoded the actual coding is different from TAXTI and transmitted asynchronously without further framing The transfer rate of 25 6 Mbit s with 4B 5B encoding implies a line rate of 32 Mbaud This interface is closely based on a
165. s of virtual channel connections VCCs and virtual path connections VPCs multiplexed in a continuous cell stream During the connection monitoring and control mechanisms operate to ensure that the connection parameters agreed upon in the ATM switching negotiation are maintained Bit 8 7 6 5 4 3 2 1 Header 5 bytes Info field Cell 48 bytes 53 bytes CO m n ll 53 Figure 10 21 ATM cell TROUBLESHOOTING LocAL AREA NETWORKS 1 0 329 An cell consists of a 5 byte header 48 byte information or user data field There are two basic types of cells UNI cells and NNI cells UNI cells are transferred at user network interfaces NNI cells at network node interfaces The two cell types differ only in 4 header bits which are specified for flow control in UNI cells though in reality not used for this and to extend the virtual path identification VPI field to 12 bits in NNI cells Bit 8 is the most significant bit in all fields The bits within each byte are therefore transmitted beginning with bit 8 The bytes in turn are transmitted in ascending order that is beginning with byte 1 Figure 10 22 illustrates the header structure of UNI cells The header consists of the six fields GFC 4 bits VPI 8 bits VCI 16 bits PT 3 bits CLP 1 bit and HEC 8 bits 4 3 2 1 1 Generic Flow Control GFC Virtual Path ID VPI 2 Virtual Path ID
166. se and fall time 10 90 1 6 1 6 ns Maximum systematic interface peak to peak jitter 0 5 0 5 ns Maximum random interface peak to peak jitter 0 15 0 15 ns Minimum eye diagram aperture at receiver 0 31 0 31 ns Software based MMF parameters Transmission characteristics 62 5 um MMF 50 um MMF Units Wavelength 770 to 860 770 to 860 nm Maximum spectral width 9 9 nm Mean optical power 10 to 4 10 to 4 dBm Minimum extinction rate 9 9 dB Maximum rise and fall time 10 9096 0 75 0 75 ns Maximum systematic interface peak to peak jitter 0 35 0 35 ns Maximum overshoot 25 25 96 Receiver characteristics Minimum sensitivity 16 16 dBm Minimum overload 0 0 dBm Maximum rise and fall time 10 9096 1 2 1 2 ns Maximum interface peak to peak jitter 0 55 0 55 ns Minimum eye diagram aperture at receiver 0 31 0 31 ns Figure 10 15 Optical transmission and reception parameters for the two 622 Mbit s ATM multimode fiber optic interfaces plastic fiber 793 2 Section 4 A4d and for 225 um multimode polymer fiber is used 793 2 Section 3 A3d The maximum loss is 9 1 dB for POF media and 1 8 dB for HPCF The minimum modal bandwidth for both fiber types must be at least 10 MHz km at a wavelength of 650 nm The data stream is NRZ encoded the PN and F07 connectors used must conform to IEC 1753 BB FFS Figure 10 16 lists the transmission and reception parameters for POF and HPCF interfaces ATM over 75
167. ses Questions to ask at this stage include Do the symptoms occur regularly or intermittently Are the symptoms related to certain applications or do they affect all network operations Do the symptoms correlate to other activities in the network When was the first occurrence of the symptom Was there any change in any hardware or software network component Has anyone connected or disconnected a PC laptop or desktop or any other component to or from the network Has anyone installed an interface card in a computer Has anyone stepped on a cable TROUBLESHOOTING LocAL AREA NETWORKS 1 0 421 Has any maintenance work been performed in the building recently by a telephone company or building maintenance personnel for example Has anyone including cleaning personnel moved any equipment or furni ture The following table lists the most common causes of problems in ATM networks ATM interface card defective ATM interface card incorrectly configured interrupt driver timers Cell streams with different priorities are being transmitted at high load and cells with low priority are discarded Classical IP ATM ARP server address not configured on the client systems Classical IP Misconfigured ATM ARP server clients are not registered at all or registered under a wrong address Faulty cable infrastructure see Chapter 6 Electromagnetic interference ATM
168. signaling mecha nisms signaling in ATM networks is an extremely complex procedure All kinds of traffic parameters such as the AAL type streaming or message mode as sured or unassured transfer cell rates peak cell rate and sustainable cell rate cell loss ratio cell delay cell delay variation tolerance and maximum burst size must be negotiated and guaranteed over all segments of the connection path Moreover novel connection oriented topologies such as point to multipoint or broadcast connections must also be handled For this reason the existing UNI and NNI signaling protocols Q 931 and Q 933 or ISUP on which B ISDN signal ing is based and which were originally developed for narrow band ISDN have undergone major extension for ATM networks to produce Q 2931 or B ISUP Signaling at the user interface takes place either based on ITU T Recommenda tion Q 2931 B DSS2 or by means of one of the ATM Forum specifications UNI 3 0 3 1 or 4 0 At the network interface public networks use B ISUP ITU T Recommendation 9 2761 Q 2764 or the corresponding ATM Forum protocols B ICI or PNNI Q 2931 ATM Forum UNI 3 0 ATM Forum UNI 3 1 Q 2931 ATM Forum UNI 4 0 ae ATM Forum P NNI Private UNI Br Private NNI SE Private ATM NETWORK E E E S Q 2931 amp ATM Forum UNI 3 0 N ATM Forum UNI 3 1 amp ATM Forum UNI 4 0 B ISUP Q 2761 Q 2764 m
169. sing AAL2 for Narrowband Services Low Speed Circuit Emulation Service LSCES Implementation Conformance Statement Performance Specification af phy 0143 000 af phy 0144 001 af pnni 0075 000 af pnni 0081 000 af pnni 0066 000 af saa 0049 000 af saa 0049 001 af saa 0069 000 af saa 0088 000 af saa api dlpi 0091 000 af saa 0124 000 af tm 0121 000 af vtoa 0085 000 af vtoa 0089 000 af vtoa 0113 000 af vtoa 0119 000 af vtoa 0120 000 af vtoa 0132 000 Approved Date Mar 2000 Mar 2000 Jan 1997 July 1997 Sep 1996 Jan 1996 Mar 1997 Nov 1996 July 1997 Feb 1998 July 1999 Mar 1999 July 1997 July 1997 Feb 1999 May 1999 May 1999 Oct 1999 390 10 1 12 2 1 113 1 121 1 150 1 211 1 311 1 321 1 327 1 361 1 363 1 1 363 2 1 363 3 1 363 5 1 610 1 356 1 371 1 350 1 555 1 365 1 1 370 1 372 1 233 1 F811 TROUBLESHOOTING LocAL AREA NETWORKS 1 0 ITU T ATM Standards Vocabulary of terms for broadband aspects of ISDN Broadband Aspects of ISDN B ISDN ATM Functional Characteristics B ISDN Service Aspects B ISDN General Network Aspects B ISDN Protocol Reference Model and Its Application B ISDN Functional Architecture B ISDN ATM Layer Specification B ISDN ATM Adaptation Layer Specification Type 1 AAL B ISDN ATM Adaptation Layer Specification Type 2 AAL B ISDN ATM Adaptation Layer Specification Type 3 4 AAL B ISDN ATM Adaptation Lay
170. t signaling Finally the ATM forum introduced UNI 4 0 which in most respects is a superset of both Q 2931 and Q 2971 but additionally contains specifications for leaf initially join point to multi point signaling the only signaling messages not contained in UNI 4 0 from the ITU T recommendations concern interworking with narrow band ISDN services The following sections are more detailed but are limited to a description of the major differences that remain between the signaling mecha nism described in UNI 3 0 UNI 3 1 and UNI 4 0 on one hand and in Q 2931 and Q 2971 on the other Signaling ATM Forum UNI 3 1 Versus ITU T Recommendation 0 2931 The for signaling SAAL specified in UNI 3 1 is based only on AAL5 The AAL sublayer definitions for this SAAL CP AAL and SSCS with the Service Specific Coordination Function SSCF and the Service Specific Connection Oriented Peer to Peer Protocol SSCOP are identical with those in the corre sponding ITU T Recommendation While the use of VPCIs is supported for identification of the ATM virtual path used for data transmission these VPCIs are limited to a length of 8 bits as opposed to 16 bits in Q 2931 in order to be identical with the VPI Furthermore no negotiation is possible between user and network regarding the VPCI VCI values to be used UNI 3 1 addressing also differs from the Q 2931 specification in that it uses only two number types TROUBLESHOOTING LocAL AREA NETWORKS
171. ta to be transported is first broken into 47 byte data blocks CS PDUs Each such data block is given a one byte header that contains a 3 bit sequence count and CRC 3 protection with parity plus a bit the CSI bit that is used to carry multi cell synchronous residual time stamp SRTS The 48 byte SAR PDU is then transported in the data field of an ATM cell At the receiving station the original data rate of the transmitting station can be synchronously regenerated using the clock information in the CS PDU header In simple terms the SRTS process compares the clock rate of the encapsulated service that is the constant bit rate service carried over AALI with the physical layer clock for example the SDH SONET clock at the source and derives the SRTS which is transported to the receiver The physical layer clock and the SRTS are then processed to recover the service clock An assumption is that there is a fixed relationship between physical layer clocks at both ends of the virtual channel carrying the AALI service normally the case within na tional networks at least 10 1 8 3 Type 2 The Type 2 ATM Adaptation Layer ITU T Recommendation 1 363 2 is used for efficient transportation of delay sensitive narrow band applications with vari able bandwidth such as telephony The network must guarantee certain QoS parameters such as maximum cell delay or cell loss ratio for each connection while making varying bandwidth available A
172. terface The payload type field the priority bit and the HEC field correspond with the same fields in the UNI cell header Figure 10 27 lists the reserved byte values in the NNI header Byte 1 Byte 2 Byte 3 Byte 4 Reserved for physical layer 2 0000 0000 0000 0000 0000 0000 0000 PPP1 Physical layer OAM cell 0000 0000 0000 0000 0000 0000 0000 1001 Idle cells 0000 0000 0000 0000 0000 0000 0000 0001 IMA ICP 0000 0000 0000 0000 0000 0000 0000 1011 Unassigned cells AAAA 0000 0000 0000 0000 0000 0000 2 Bit available for use by the ATM layer pu Bit available for use by the physical layer preter In cells with the VPI VCI value equal to 0 0 the four bits normally used to represent the PT and CLP fields are reinterpreted to distinguish between different types of unassigned and physical layer cells 2 nr Cells identified by header information as physical layer cells are not passed to the ATM layer Figure 10 27 Reserved header bytes in the NNI ATM cell TROUBLESHOOTING LocAL AREA NETWORKS 1 0 335 10 1 7 1 ATM Cell Types In addition to user ATM cells cells whose VCI gt 31 there are several types of cells that are used not to transport user data but to perform certain operational functions These include idle cells unassigned cells PL OAM cells ICP cells RM cells and VP VC OAM cells Note that in cells having a VPI VCI 0 0 the least 4 bits in the 4 byte normally used
173. the effects of cell encapsulation in the physical layer trans mission framing Thus two cells within a single SDH container SONET SPE will have a smaller CDV relative to one another than two cells transported in different containers SPEs TROUBLESHOOTING LocAL AREA NETWORKS 1 0 408 Figure 10 69 lists the typical causes of problems in the ATM layer grouped by symptoms Cell Error Rate CER Cell Loss Ratio CLR Cell Misinsertion Rate CMR Mean Cell Transfer Delay MCTD Cell Delay Variation CDV Sources of error CER CLR CMR MCTD CDV Signal propagation delay x Fault in communication medium X X Switch architecture X X X Buffer capacity X X X Number of nodes X X X X X along a given VPC VCC connection Network load X X X X Error X Bandwidth allocation X X X to a given VPC VCC Figure 10 69 Symptoms and causes of problems in ATM networks 10 2 3 Troubleshooting Higher Layers If the ATM layer is working and the problem persists the higher layer protocols in use must be analyzed Two of the most common higher layer protocols besides UNI signaling are LAN Emulation LANE and the Private Network to Network Interface PNNI 10 2 3 1 LAN Emulation If problems occur in LAN emulation environments the first step is to make sure that the connected traditional LANs 10 100 1 000 Mbit s Ethernet FDDI etc and the LAN interfaces of the LAN ATM inter
174. to convey the PT and CLP fields are re assigned for other purposes see Figure 10 27 Idle Cells Idle cells are physical layer cells and carry no useful information They are used to adapt the cell rate to the bandwidth of the transmission medium If there are not enough ordinary cells to fill the allocated bandwidth idle cells are inserted by the transmission convergence Sublayer This permits alignment of the ATM cell stream with the throughput of the physical medium such as an SDH VC4 container SONET SPE Unlike unassigned cells idle cells are not passed to the ATM layer The ATM Forum following the Bellcore definition calls the idle cell header pattern an invalid header revealing some confusion in the minds of Bellcore and hence ATM Forum specifiers see Unassigned Cells for more on this See Figure 10 27 Unassigned Cells Unassigned cells VPI VCI 0 0 are ATM layer cells and carry no useful payload information They are used when no assigned cells are available to send from the ATM layer At the receiver they are treated similarly to idle cells and discarded In North America and implementations elsewhere based on certain ATM Fo rum specifications unassigned cells are used where the ITU T would specify idle cells in particular they are specified for rate adaption by the ATM Forum adopting Bellcore usage which strictly conflicts with ITU T usage Despite this confusion idle and unassigned cells can normally be co
175. tocol is mainly based on the ITU T Q 2931 signaling recommendation In addition to the signaling mechanisms drawn from Q 2931 however PNNI also contains a num ber of completely new functions These include searching for alternative routes which can involve crankback back tracking when an attempt to move for ward fails for various reasons maintenance of connection matrices for ascer taining routes and the operation of a dedicated routing control virtual channel used only to distribute routing information Furthermore all PNNI nodes save TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 375 updated topology state information at regular intervals These topology state parameters provide information about the availability of connections to adja cent nodes and are continuously updated by means of a special hello proce dure carried out between all PNNI nodes 10 1 10 ATM Interworking 10 1 10 1 LAN ATM LLC Encapsulation RFC 1483 The first specifications available for connecting LANs with ATM networks were the two Requests for Comments RFCs RFC 1483 superceded in 1999 by RFC 2684 and RFC 1577 superceded in 1998 by RFC 2225 published in 1993 by the IETF Internet Engineering Task Force RFC 2684 specifies two methods for the encapsulation of LAN data packets in ATM LLC encapsulation VC based multiplexing FE FE 03 ISO PDU PAD UU CPI LNG CRC packet 3 a
176. transmission path There is small probability that cells with errored headers may appear as valid cells and thus lead to incorrect transmissions misinserted cells The probability of such an event can be calculated from the number of errored headers containing more than two incorrect bits the HEC checksum of such a header no longer indicates 104 105 105 107 10 109 y 10 10 10 1072 1073 1074 1075 1076 1077 1078 1079 10 20 1021 102 10 23 1024 105 10 26 10 27 1025 with incorrect headers Probability of cell loss or of valid cells rr rer ti 10 108 107 10 105 10 Probability of bit errors 3 hml TTT e e 0 0 Probability of cell loss Probability of valid cells with incorrect headers Figure 10 61 Probability of the transmission of errored cell headers as valid cells in relation to the bit error ratio TROUBLESHOOTING LocAL AREA NETWORKS 1 0 399 whether it is corrupt and the ratio of valid header values to the number of all possible header values Figure 10 61 shows the probability of the transmission of errored cell headers as valid cells in relation to the bit error ratio Discarded Cell Ratio The discarded cell ratio equals the number of cells received with errored head ers that cannot be corrected and which are therefore discarded divided by the total number
177. ts Value transmission parameters WEE Mbaud 1944 Baud rate tolerance ppm 100 Optical signal power min dBm 20 max dBm 14 Wavelength min nm 1270 max nm 1380 Spectral width nm 200 Optical extinction rate 96 lt 10 Pulse rise and fall time min ns 0 6 10 90 ns 25 Pulse overshoot 10 Pulse undershoot 10 Optical transmission jitter due to data ns 0 5 Optical pulse phase shift ns 0 5 Optical pulse jitter ns 0 5 Eye diagram 171 for output signal 1 1 Level L 1 q 1 1 1 1 1 1 1 1 1 98 5 N J EE 0 Level Time gt Optical Parameters Units Value reception parameters Baud rate Mbaud 194 4 Baud rate tolerance ppm 100 Optical signal power min dBm 29 max dBm 14 Wavelength min nm 1270 max nm 1380 Pulse rise and fall time min ns 0 6 10 90 max ns 3 0 Optical transmission jitter due to data ns 1 0 Optical pulse phase shift ns t05 Optical pulse jitter ns 05 Figure 10 14 Optical transmission and reception parameters for the 155 Mbit s ATM multimode fiber optic interface TROUBLESHOOTING LocAL AREA NETWORKS 1 0 318 media found increasingly often in LANs as well Figure 10 13 lists the single mode fiber optic parameters as specified in ITU T Recommendation G 957 for 622 08 Mbit s interfaces ATM over Multimode Fiber For the transmission of ATM cells in SDH SONET
178. um delay between the transmission of two BGN END or RS PDUs if no answering PDU is received The CC timer must also have a value greater than twice the signal delay over the connection concerned SSCOP Error Messages to the Management Layer Error codes are used to forward SSCOP error events to the management layer The table in Figure 10 40 lists the various SSCOP error types SAAL Connection Setup The connection setup between two SAAL system components is triggered by an AA ESTABLISH message from the SSCF sublayer This command contains the parameters SSCOP User to User Parameter SSCOP UU and Buffer Release BR which are used in generating the SSCOP sublayer s BGN message The receiver decodes the BGN message and passes an AA ESTABLISH ind to the receiving SSCF This sublayer responds with an AA ESTABLISH res command which likewise contains the parameters SSCOP UU and BR The resulting BGAK message is sent back to the originating SSCOP which passes an AA ESTABLISH conf to the initiating SSCF see Figure 10 41 SSCF SSCOP SSCOP SSCF AA ESTABLISH req AA ESTABLISH ind AA ESTABLISH res Bont AA ESTABLISH conf Figure 10 41 SAAL connection setup TROUBLESHOOTING LocAL AREA NETWORKS 1 0 359 10 1 9 ATM Signaling Signaling in ATM refers to all processes necessary to set up a connection between two or more stations In contrast to conventional
179. ure 10 54 TROUBLESHOOTING LocAL AREA NETWoRKS 1 0 377 LLC Ethernet ATMARP a AA AA 03 000000 08 06 CRCIPAD UU CPI Figure 10 54 ATMARP and InATMARP RFC 2225 The following rules govern the design and operation of Classical IP subnetworks LISs All IP nodes of an LIS must be directly connected to the ATM network All IP nodes of an LIS must have the same IP network or subnet work address Network nodes outside the LIS must be accessible only through routers AILIP nodes in an LIS must be able to resolve addresses by Every IP node in an LIS must be able to communicate with all other IP nodes Address resolution must function both for PVCs and for SVCs The default packet size in CIPs is 9 180 bytes Adding the 8 byte LLC SNAP header yields a default AAL5 PDU size of 9 188 bytes in CIP networks The differences between RFC 2225 and its predecessor RFC 1577 lie mainly in the in the progress made in signaling procedures during the intervening 5 years 10 1 10 3 LAN Emulation LANE The most universal method of efficiently integrating ATM networks in existing conventional LAN structures involves a complete emulation of the LAN MAC layer which allows all existing LAN applications to be extended across ATM networks without modification From the point of view of traditional local area networking the L
180. utes from the MPS as necessary Prerequisites for the use of MPOA in the LAN include support for ATM UNI signaling UNI 3 0 3 1 4 0 LANE v2 Next Hop Resolution Protocol Basic MPOA Processes As in LANE all MPOA processes are managed through control virtual channels while the actual data transport takes place through separate data virtual chan nels All control and data flows are transported over VCCs with LLC encapsula tion RFC 1483 Channels are set up and cleared down in accordance with one of the UNI signaling specifications UNI 3 0 3 1 or 4 0 Four kinds of control virtual channels are defined MPS MPC configuration virtual channels MPC MPS control virtual channels MPS MPS control virtual channels MPC MPC control virtual channels MPOA components obtain configuration information from the LE Configura tion Server The MPOA clients obtain route information from the MPOA server over the MPC MPS control virtual channel The MPS MPS control virtual chan TROUBLESHOOTING LocAL AREA NETWORKS 1 0 383 nels are used by standard routing protocols and NHRP to exchange route information MPCs exchange control information with one another only if one MPC receives misrouted data packets from another In this case the sender MPC is notified so that it can delete the incorrect routing information from its cache MPOA networks are characterized by the following five basic operating pro cesses
181. work Interface UNI Layer 3 Specification for Basic Call Connection Control Q 2951 1 Stage 3 Description for Number Identification Supplementary Services Using B ISDN Digital Subscriber Signaling System No 2 DSS 2 Basic Call Clauses 1 6 and 8 DDI MSN CLIP CLIR COLP COLR SUB Q 2957 1 Stage 3 Description for Additional Information Transfer Supplemen tary Services Using B ISDN Digital Subscriber Signaling System No 2 DSS 2 Basic Call Clause 1 User to User Signaling UUS Q 2971 Broadband Integrated Services Digital Network B ISDN DSS 2 Digital Subscriber Signaling System No 2 User Network Interface Layer 3 Specification for Point to Multipoint Call Connection Control 10 1 12 3 IETF ATM Standards RFC2379 RSVP over ATM Implementation Guidelines L Berger August 1998 RFC2225 Classical IP and ARP over ATM M Laubach J Halpern April 1998 Obsoletes RFC1626 RFC1577 RFC2226 IP Broadcast over ATM Networks T Smith G Armitage October 1997 RFC2320 Definitions of Managed Objects for Classical IP and ARP Over ATM Using SMIv2 IPOA MIB M Greene J Luciani White T Kuo April 1998 RFC2331 ATM Signaling Support for IP over ATM UNI Signaling 4 0 Update M Maher April 1998 RFC2380 RSVP over ATM Implementation Requirements L Berger August 1998 RFC2381 Interoperation of Controlled Load Service and Guaranteed Service with ATM M Garrett M Borden August 1998 RFC2382 A Framew
182. working devices are functioning correctly This includes verifying the various LAN configuration settings and measuring basic network statistics with a protocol analyzer Examination of the LAN emulation components begins only after the LAN part has been proven to be working properly The first step in LANE troubleshooting is to send pings between two LE Clients and check whether a connection can be set up at all If the ping does not go through check the IP interfaces of the LE Clients Examine whether the IP interfaces are active at all and whether the IP addresses and TROUBLESHOOTING LocAL AREA NETWORKS 1 0 409 subnet masks correct and in the same subnet Then determine whether the LANE software on the clients is active and whether both LE Clients belong to the same ELAN If no error is found the following configuration parameters must be systematically checked through the system management interface of the ATM components Are both LE Clients registered on the same LE Server and Broadcast Unknown Server BUS Is the ATM address of the primary and secondary if configured LE Server correct and are the Configuration Direct VCCs set up Is the ATM address of the primary and secondary BUS correct and are the Multicast Send VCCs set up If it is still impossible to set up data VCCs between the LE Clients the last error cause to check for is a restricted VC capacity on one of the systems due to traffic contr
183. ws T y T awaya i M1 TT i 228 x 1 i mj E E INVI LY Da WLW v INN J ws INN dIN19 VOdN dVN1O eX Pao 1 i NN y TE Lu NN 1 ex i Ald NV SnO03N3908313H pua WLY paa noxae pi CE 3 gt i d ul INN INS INN WV Said MM EM MEM t pm INN INN INNa 101 8 INN 1262 O L 62 0 W LY i od WN 4051 9 VN 4 Je21sse 25 we SA NS INN INN 4 Syys INN alo gt NN i CN i NO Jf NN INNd IW i Voas J N Y x AHd WL NLY T asus dws 3544 KVS O v L 0 INN y i CUm 1 LL6Z D LE6Z O n d 0 7 0 INN RSEN Ly L262 02620 Swe 4 F snoeueBoJejeu snoeueBouiou NVM nand i NYM NLY 31VA 3g NYI NLY Figure 10 1 ATM in LAN and WAN General view TROUBLESHOOTING LocAL AREA NETWORKS 1 0 301 10 1 3 Public Wide Area Networks When used for data communication in public wide area networks ATM be transparently connected with Frame Relay and Switched Multimegabit Data Services SMDS networks the latter being o

Network Troubleshooting

Contents

Download Pdf Manuals

Related Search

Related Contents