Monitoring and Troubleshooting VoIP Networks With a
Contents
1. pd Pas Time Time Inte Operation Request Response Transport Information 1 19 07 22 705480 0 000000 INVITE INVITE sip 55555 l1 Src IP 866 140 116 2 2 19 07 22 879612 0 174132 100 Trying Src Port 3091 3 19 07 22 879698 0 000086 amp 407 Proxy Authentir Dest IP 69 90 155 70 4 19 07 22 891367 0 011669 gt ACK sip 5S555 Fwc Dest Port 5060 5 19 07 22 900382 0 009015 INVITE gt INVITE sip 55555 l1 Protocol UDP 6 19 07 23 085670 0 185288 100 Trying Timing 7 19 07 23 085727 0 000057 100 trying your c Start Time 8 23 2006 7 07 22 PM 8 19 07 23 403051 0 317324 180 Ringing End Time 8 23 2006 7 07 40 PM 9 19 07 23 819086 0 416035 E lt a 200 OK Duration 0 00 17 8 10 19 07 23 825379 0 006293 gt ACK sip 55555 83 Quality 11 19 07 40 259384 16 434005 BYE BYE sip 55555 83 MOS Score 4 4 12 19 07 40 558462 0 299078 K 200 OK R Factor 93 2 SIP Call ID 24170 192 168 131 70 Calling Party Src Display Name Src SIP Address 794379 fwd pulver com Src Tag 30604 Src User Agent PortSIP softphone 2 0 Called Party Dest Display Name Dest SIP Address 555550 fwd pulver com 0 6 2 3 Bandwidth Utilization The need for monitoring network bandwidth and controlling network conditions is paramount for delivering the expected voice quality in a VoIP system VoIP adds strict requirements for bandwidth utilization making it necessary to co2. Transport Information 1 06 08 32 92 0 000000 30059 1080704222 ITU T G 729 a Src IP 213 46 11 154 2 06 08 32 94 0 019999 30060 1080704382 ITU T G 729 Src Port 4030 3 06 08 32 96 0 019975 30061 1080704542 ITU T G 729 Dest IP 212 46 11 66 4 06 08 32 98 0 019985 30062 1080704702 ITU T G 729 Dest Port 31954 5 06 08 33 00 0 020027 30063 1080704862 ITU T G 729 Protocol UDP 6 06 08 33 02 0 019992 30064 1080705022 ITU T G 729 Timing 7 06 08 33 04 0 020045 30065 1080705182 ITU T G 729 Start Time 9 13 2006 6 0 8 06 08 33 06 0 019956 30066 1080705342 ITU T G 729 End Time 9 13 2006 6 0 9 06 08 33 08 0 020025 30067 1080705502 ITU T G 729 Duration 0 00 11 9 10 06 08 33 10 0 019985 30068 1080705662 ITU T G 729 Quality 11 06 08 33 12 0 020036 30069 1080705822 ITU T G 729 MOS Score 4 1 12 06 08 33 14 0 020020 30070 1080705962 ITU T G 729 R Factor 83 2 13 06 08 33 16 0 019965 30071 1080706142 ITU T G 729 RTP Statistics 14 06 08 33 186 0 020015 30072 1080706302 ITU T G 729 RTP Packet Count 599 15 06 08 33 20 0 019995 30073 1080706462 ITU T G 729 Lost Packets a 16 06 08 33 22 0 020015 30074 1080706622 ITU T G 729 Duplicate Packets 0 17 06 08 33 25 0 028216 30075 1080706782 ITU T G 729 Sequence Errors a 18 06 08 33 27 0 011824 30076 1080706942 ITU T G 729 Network Utilization 19 06 08 33 286 0 019966 30077 1080707102 ITU T G 729 Total Traffic bytes 44 326 20 06 08 33 30 0 020012 30078 108070
3. Cable Or Other Broadband Network y Detected xDSL or Other No inexpensive solution to greatly improve the quality of VoIP calls Frame Relay Network When you configure QoS on a wireless network you specify and prioritize the VolP Name Priority Protocol p g fargos oa 1 255 7 lt lt fur 3 traffic In effect a wireless access point will Local IP Range Local Port Range appl congestion management and u ho00 to 255 255 255 255 1024 tofessas PPlY 8 8 Ee congestion avoidance methods to ensure 15 the priority transmission of the VolP packets as opposed to the regular best effort delivery technique Implementing QoS in the wireless environment network allows getting more predictable performance and amp CommYiew for WiFi D Link DWA File Search View Tools Settings R me i BOPC y 2 Nodes AN Channels Pa Latest El Wireless Packet Info Signal level 80 Signal level in dBm 47 Noise level in dBm 82 Rate 130 0 Mbps Band 802 11ng Channel 2 2417 MHz Date 18 Jan 2008 Time 14 35 43 229552 Delta 0 002818 Frame size 124 bytes Frame number 7600 E 802 11 Frame Control 04188 16776 Protocol version 0 To DS 1 From DS 0 More Fragments 0 Retry 0 Power Management 0 More Data 0 Protected Frame 1 Order 0 Type 2 Data Subtype 8 QoS Data Duration 0x002C 44 BSS ID 00 19 5B 57 B8 2D Source Address 00 13 E3 40 8F B Destination Addr
4. around the network Endpoints Top Callers Chart IP Address MAC Address Description Placed Recei Succ Failed Total Tak 3 07 40 PM 192 168 10 2 WelltechCo 01 42 3D SIP201 Ip201si 0 1 1 0 0 00 10 8 3 07 40 PM 192 168 10 1 WelltechCo 01 42 3C SIP201 Ip201si 1 D 1 0 0 00 10 8 1 31 56 PM 68 142 233 154 AsustekCom 0D D4 4D 0 1 1 0 0 01 13 6 6 08 44 AM 213 46 11 154 AudioCodes 06 D0 C8 Gateway 4 804 3 3 6 0 0 01 21 9 6 08 44 AM 213 46 11 14 Cisco 4F 4D 01 netH323v4 7 6 2 3 3 6 0 0 01 21 9 7 07 40 PM 69 90 155 70 30 48 20 00 01 00 snom320 6 2 3 D 1 1 0 0 00 16 2 6 54 20 PM 213 53 35 219 30 48 20 00 01 00 CommuniGatePr 0 2 1 1 0 00 34 0 6 54 20 PM 886 140 116 2 00 00 01 00 00 00 PortSIP softpho 3 0 2 1 0 00 50 2 3 32 57 PM 86 192 4 88 D Link CO E8 FE CommuniGatePr 1 1 1 0 0 01 38 0 3 32 57 PM 213 53 35 219 Compex 36 FF 4C PortSIP softpho 1 D 0 0 0 00 24 3 Typical VoIP Problems Due to human perception VoIP is much more sensitive to certain network conditions that are considered well within spec for most applications Network issues such as packet loss jitter and packet sequence errors are inherent to IP networks and are well corrected and tolerated by data transfer protocols Voice transmissions are real time by the nature hence the different approach to handling the network issues Packet loss jitter and out of order packets are tied closely to each other Taking care of a single network issue
5. As described in the previous chapters VolP discards packets that are received too late or out of order As such wireless packet retries contribute to voice quality issues such as jitter and drops in audio streams caused by dropped frames High numbers of retries reduce the effective speed of the wireless network lowering available bandwidth and making the data transfer speed variable which in turn contributes to jitter and makes it even tougher to maintain an acceptable quality of conversation Using the statistics proved by a Wi Fi network analyzer you can determine the number of retries on a per node basis as illustrated by the screen shot above While some retries inevitably happen in any 802 11 network a high number of retries compared to the total number of packets typically indicates that the clients are located too far away from the access point s or the level of radio interference is too high Wireless QoS WMM Wi Fi Multimedia WMM is a Wi Fi Alliance certification based on the IEEE 802 11e draft standard It provides basic QoS features to IEEE 802 11 Enable QoS Engine y networks Prioritizing VolP traffic over less ii ion 17 time sensitive transmissions allows reducing Dynamic Fragmentation Y Automatic Uplink Speed F variability in the transmission of the VoIP Measured Uplink Speed Not Estimated packets Using QoS is a simple and Manual Uplink Speed f125 kbps lt lt Select Transmission Rate Connection Type
6. 0 057876 100 Trying 19 22 46 057917 0 000061 da 401 Authentication required codecs In order to successfully 19 22 46 067018 0 009101 ACK sip 84952492679 sipnet t communicate the two 19 22 46 073539 0 006521 INVITE D INVITE sip 84952492679 sipne 19 22 46 142806 0 069267 100 Trying endpoints must negotiate a 19 22 46 347880 0 205074 da 488 Codec Mismatch 19 22 46 352509 0 004629 gt gt ACK sip 84952492679 sipnet 1 common codec they will both use While issues resulting from HAAA 1 2 3 4 5 6 7 8 codec incompatibility or the lack of common codecs among heterogeneous infrastructures are rare these problems are also hard to detect One can identify SIP session errors by looking at codec mismatches as illustrated by the CommView screen shot Monitoring VoIP Networks This chapter overviews major network parameters that shall be monitored by a network traffic analyzer in wired and WiFi VoIP networks WiFi VoIP networks are often referred to as VOWLAN or VoFi While the cost and technological benefits of VoIP infrastructure are a notch above the POTN networks the end users of the technology are used to high quality and reliability of conversations If significant sacrifices have to be made regarding the call quality many users will not be willing to switch to the new technology In many cases especially where local area network LAN infrastructures are involved i
7. 3 are dominating the VoIP arena as the most commonly used signaling protocols Choosing one signaling protocol over another when developing a VoIP solution is a matter of a set service requirement and the choice of equipment SIP is commonly chosen among the full scale VoIP carriers to make use of the abundance of SIP compatible VoIP devices including the numerous inexpensive SIP phones and adapters When multimedia communication over IP networks is required including video conferencing and data calls in addition to audio transmission H 323 becomes the natural choice As defined by the scope of the signaling protocols there are numerous potential problems that may arise because of compatibility or networking problems The dreadful unable to connect problem lies frequently in the domain of signaling protocols The two peers once located may be unable to connect because of compatibility problems between the two endpoints various SIP implementations are especially prone to this problem as well as the lack of the required features such as conference calling in one of the connecting devices A network analyzer should recognize and support both SIP and H 323 signaling protocols allowing the detection of problems that occur on the signaling phase early during the implementation of a VoIP system Throughout this white paper we ll be illustrating the problems and solutions with the help of CommView and CommView for WiFi software based network analyzer
8. 7262 ITU T G 729 work Transport 25 158 56 7 21 06 08 33 32 0 02000 y a 68 16 2 la lee RTP does not use TCP protocol for transmitting voice packets Despite the fact that TCP guarantees the delivery of the packets its session initiation time and associated delays are unacceptable for transferring multimedia data in real time Therefore UDP is the natural choice here As there is no recovery for the packets not delivered to the recipient a certain percentage of voice packets are typically lost While VoIP does provide means to reconstruct the lost packets without significant loss of voice call quality packet loss beyond a certain level starts to noticeably degrade the quality of the conversation The illustration above displays the numbers of lost packets allowing identifying network problems on the streaming phase RTP encapsulates additional information in every voice packet including payload type identification to identify the type of content being transmitted sequence numbering that is used to detect and identify the lost packets and time stamping to allow synchronization and jitter calculations The additional information is extremely handy when analyzing RTP streams in order to identify the source of quality issues Codecs In order to convert an analog voice signal into a set of digital packets and then reconstruct the packets back into audible voice special voice codecs were developed Codecs are used to encode voice into d
9. Monitoring and Troubleshooting VoIP Networks with a Network Analyzer Executive Summary Voice over IP VoIP is rapidly changing the way we use the telephone for voice communications The term Voice over IP defines the transport of VolP based networks including the signaling and streaming protocols as well as describing the codecs VoIP is widely used by all kinds of consumers ranging from computer enthusiasts who are excited to get free long distance calls over the Internet to full scale enterprise solutions targeted to replace the entire infrastructure inherent to analog telephony There is no doubt that VoIP technology has everything in it to reduce communication costs significantly when compared to traditional analog telephony Achieving call quality comparable with the quality of calls carried over PSTN networks is another matter Deploying Voice over IP solutions requires careful analysis of network requirements and current conditions in order to provide call quality comparable to analog PSTN carriers This White Paper describes the potential quality issues that must be addressed when developing or deploying a VoIP solution over both wireless Wi Fi and wired Ethernet network infrastructures OFT Copyright 2008 TamoSoft All Rights Reserved No part of this work can be reproduced or duplicated in any form without the expressed written permission of TamoSoft PFOLOCOMOVERVIO Weeds sites o e e eo Leds eno lt do a e O 3 SEN
10. age Jitter 10 20 s 5 a Oo 0 w o mn o un o Ln o w te p o o p o o o o o o o i So j Ss Ss a 6 5 6 Jitter Jitter is a specific VoIP Quality of Service issue that may affect the quality of the conversation if it goes out of control Unlike network delay jitter does not occur because of the packet delay but because of a variation of packet delays As VolP endpoints try to compensate for jitter by increasing the size of the packet buffer jitter causes delays in the conversation If the variation becomes too high and exceeds 150ms callers notice the delay and often revert to a walkie talkie style of conversation There are several steps to be taken to reduce jitter both on the network level and in the VolP endpoints such as VolP software IP phones or dedicated VolP adaptors By definition reducing the delays on the network helps keep the buffer under 150ms even if a significant variation is present While the reduced delay does not necessarily remove the variation it still effectively reduces the degree to which the effect is pronounced and brings it to the point where it s unnoticeable by the callers Prioritizing VolP traffic and implementing bandwidth shaping also helps reduce the variation of packet delay At the endpoint it is essential to optimize jitter buffering While greater buffers reduce and remove the jitter anything over 150ms noticeably affects the perceived quality of the conversation A
11. arious conversations and give their judgment on the acceptable quality The ability to play back actual conversations is essential for successful VoIP deployments A network analyzer provides the means to record and play back calls to assess voice quality by simply listening to the voice as shown on the screen shot 13 VoWLAN Specific Factors Wireless networking introduces specific problems affecting the quality of VolP conversations There are a limited number of users who can use a single wireless access point at the same time while maintaining acceptable voice quality Limited reception and low signal strength alter available bandwidth and affect the number of retries Adjusting the settings of endpoints and wireless access points to permit prioritizing voice traffic over data with the use of Quality of Service QoS allows for increased call quality with no extra effort VolP over wireless networks may suffer from numerous problems that are nonexistent on wired networks This chapter discusses the various factors that affect the quality of VolP calls transferred over wireless networks Number of Clients per AP Wi Fi access points have limited bandwidth effectively restricting the number of concurrent VolP users Depending on the choice of codecs by the endpoints and the 802 11 standard used in the hardware the number of simultaneous VolP calls supported by a wireless access point may vary Analyzing the average and peak num
12. at the client is not WMM aware on the driver level Reporting Real time analysis is important for performing on site adjustments to a VolP system being developed or deployed A more comprehensive analysis provides benefits of greater statistics samples better usability and visibility of essential issues While real time analysis is mostly usable to technical specialists the team leaders and the management can use a properly formatted report A processed report makes for a great presentation or a brief report during a corporate meeting The ability to generate custom reports highly benefits a network analyzer The following screen shot demonstrates typical information available in a VoIP system performance report 16 IV Generate reports Data to include JV SIP Sessions JV Endpoints Y H 323 Sessions JV Registrations JV RIP Streams JV Errors Save to Report Format P Projects C Source REPORTS VoIP report htm B HTML View Clear C Overwrite C csv Append z Interval o hours 30 min o se Next report in hh mm ss 00 30 00 Save Report Now Reports are generated only if capturing is on Turning off capturing suspends the timer About TamoSoft TamoSoft develops cutting edge security and network monitoring software for the Internet and Local Area Networks providing clients with the ability and confidence to meet the challenges of tomorrow s technology Keeping pace with in
13. bers of connected clients is crucial for the deployment of a VoIP system allowing making a weighed choice of hardware and codecs The following screenshot illustrates a few clients connected to wireless access points as well as a number of other parameters discussed in the next chapters File Search View Tools Settings Rules Help PETIT Adidas 2 Nodes Channels Pa Latest IP Connections T Packets El VoIP T Logging Rules E Alarms S D Link 57 B8 2D 2 AP DOTI1N WPA CCMP 48 68 85 1 106 97 300 1 740 704 5 608 113 609 amp amp p Link E9 05 00 11 AP OFFICES WEP 46 57 68 1 13 99 54 123 346 1 190 89 0 D IntelCorpo A0 8F B5 2 STA WPA 38 52 83 18 56 89 65 570 414 2 876 17 0 D D Link 03 96 1F 2 STA WPA 43 58 80 1 150 79 300 977 582 2 428 230 697 D Netgear D2 37 53 11 STA WEP 23 36 56 11 43 59 54 83 704 1 150 130 0 With the advent of 802 11n technology that presently supports speeds up to 300 Mbps and will soon support 600 Mbps the issue of the number of concurrent VoIP calls becomes less critical but the older 802 11b 11 Mbps 802 11g 54 Mbps and 802 11a 54 Mbps gear is still far more common than the newer 802 11n devices Online bandwidth calculators are widely available using one such calculator the network administrator can get necessary metrics that depend on the 802 11 standard and codes being used For example for an 802 11g WLAN and G 729 as the VolP codec 27 simultaneous VolP calls are recomm
14. can often reduce all three problems and significantly improve the quality of voice calls This chapter discusses network issues that affect the perceptual quality of VoIP calls Packet Loss Packet loss occurs in every kind of network All network protocols are designed to cope with the loss of packets in one way or another TCP protocol for example guarantees packet delivery by sending re delivery requests for the lost packets RTP employed by the VoIP protocol does not provide delivery guarantee and VoIP must implement the handling of lost packets While a data transfer protocol can simply request re delivery of a lost packet VoIP has no time to wait for the packet to arrive In order to maintain call quality lost packets are substituted with interpolated data A technique called Packet Loss Concealment PLC is used in VoIP communications to mask the effect of dropped packets There are several techniques that may be used by different implementations Zero substitution is the simplest PLC technique that requires the least computational resources These simple algorithms generally provide the lowest quality sound when a significant number of packets are discarded Waveform substitution is used in older protocols and works by substituting the lost frames with artificially generated substitute sound The simplest form of substitution simply repeats the last received packet Unfortunately waveform substitution often results in unnatural rob
15. daptive algorithms to control buffer size depending on the current network conditions are often quite effective Fiddling with packet size or using a different codec e g G 711 often helps control jitter While jitter is caused by network delays more often than by endpoints certain resource struggling systems that are executed in concurrent environments such as VolP soft phones may introduce significant and unpredictable variations in packet delays While developing VolP endpoints or examining call quality problems within existing VoIP infrastructure it is very important to isolate the cause of jitter A network analyzing tool can be extremely handy in localizing the source of the problem quickly and efficiently A good network analyzer is capable of calculating jitter for every RTP stream and building jitter and jitter deviation charts along the time axis Sequence Errors Data packets travel independently of one another and are subject to various delays depending on the exact route they take Out of sequence packets are not considered a problem for data transfers as data transfer protocols can re order packets and reconstruct data without corruption Due to the time sensitive nature of voice communications VoIP systems are required to handle out of sequence packets in quite a different manner Some VoIP systems discard packets received out of order while other systems discard out of order packets if they exceed the size of the internal buffe
16. dustry trends we offer professional tools that support the latest standards protocols software and hardware in both wired and wireless networks With a portfolio including such companies as Motorola Siemens Ericsson Nokia Cisco Lucent Technologies Nortel Networks Unisys UBS Dresdner Bank AG Olympus and General Electric TamoSoft is one of the fastest growing IT application development firms in the marketplace today TamoSoft products are available through this Web site as well as through a network of distributors and resellers Founded in 1998 as the software division of a Cyprus based business consulting company TamoSoft is a privately held company based in Christchurch New Zealand TamoSoft employs an international team dedicated to the creation of high quality software that customers from over 100 countries rely on and partners with industry leaders in technology and services such as Zone Labs Visualware and The CWNP Program TamoSoft PO Box 1385 Christchurch 8140 New Zealand www tamos com 17
17. ecause of bandwidth limitations and lost packets rather than codec quality If there is less than 64 Kbit s of available bandwidth picking a low bitrate high compression G 729 or G 723 codec is much more appropriate Note that while a local area network LAN may provide high bandwidth external calls may be subject to bandwidth bottleneck in the upstream ADSL and cable network providers often offer limited upstream bandwidth which results in upstream congestion if multiple VoIP calls are carried concurrently In this case low bandwidth codecs may provide better results G 711 PCM a high bandwidth codec provides the best audio quality yet consumes the most bandwidth G 729a CS ACELP G 723 1 MP MLQ and G 726 ADPCM offer varying conversation quality sorted by decreasing relative quality The codec choice will not take place automatically A system administrator must specify and prioritize codecs available to the particular VoIP system Taking care of the wrong choice of codec may significantly improve conversation quality By logging and displaying the session flow a network analyzer allows seeing the codec negation process i e the codecs available to the endpoints as well as the final negotiated codec choice Pa Time Time 1m Operation RequestjResponse 1 15 07 20 968571 0 000000 INVITE A INVITE sip 2222 192 168 10 2 1 INVITE Header Content v 0 o SIP201 12367 0 IN IP4 192 168 10 1 s 5IP201 Session i Audio Sessio
18. ended with an anticipated MOS of 3 8 with the maximum being 98 simultaneous VolP calls with an anticipated MOS of 3 2 Using 802 11b instead of 802 11g decreases the number of simultaneous calls by approximately five times Using a wideband G 711 codec with an 802 11g access point 15 simultaneous VolP calls are recommended with the anticipated MOS of 4 1 the maximum being 53 simultaneous VolP calls with the anticipated MOS of 3 4 14 Signal Strength The signal strength affects the effective transmission speed of a wireless connection having immediate effect on the quality of a VoIP call Low signal causes endpoint hardware to choose lower transmission speeds possibly invalidating the initial choice of codec Low signal levels affect the number of packet retries causing out of order packets and contributing to jitter Needless to say lower than usual signals may drop transmission rates to the point where voice transmission quality becomes unacceptable A network analyzer continuously monitors and displays signal levels and transmission rates of all wireless stations allowing network administrators detect insufficient rates make necessary hardware relocations and or perform codec adjustments to match the effective bandwidth Number of Retries Wireless data transmission is much more susceptible to errors than wired Packets sent over wireless networks are often lost causing wireless hardware to retransmit the same packets over and over again
19. ess 00 30 05 40 4 Fragment Number 0x0000 0 Sequence Number 0x0038 56 QoS Control 0x0000 0 SERENE better effective transmission rates reducing VoIP jitter and the effects caused by dropped and out of order packets Enabling QoS is generally as simple as adjusting the settings in the endpoints and the wireless access point The screen shot above illustrates a typical QoS engine configuration of an access point QoS has arguably the best price performance ratio of all adjustments to the configuration of a wireless network While configuring QoS in an access point appears to be a simple task it s important to verify that the QoS works as intended i e the QoS engine rules are correctly applied VoIP packets are prioritized while others are not and that the wireless nodes involved in VoIP communications support WMM on the driver level Since a wireless network monitoring tool can capture and decode individual packets it is quite possible to examine the captured packets for QoS compliance Specifically for the packets that are used in VoWLAN communications the Data Subtype should be set to QoS Data and the QoS Control field that contains the data priority level should be set to a non zero value Respectively in non VolP packets the QoS Control field should be set to zero If the Data Subtype is set to Data rather than QoS Data for some of the clients this typically indicates th
20. igital form and decode it back into audible analog form when received In VoIP codecs are used to encode voice for streaming across the IP network There are numerous codecs on the market many being publicly available at no charge Codecs vary in the sound quality they deliver using different bandwidth and computational requirements With few exceptions codecs employ compression to save network bandwidth at the expense of using more CPU and memory resources and or delivering lower voice quality on the receiving end The less bandwidth and computational power a codec requires for achieving identical sound quality the better it is considered to serve its purpose G 711 is a common open source and royalty free high bitrate codec This codec does not require licensing fees and uses very little computational resources while providing the best possible sound quality at the expense of higher than usual network bandwidth On the other hand G 723 and G 729 patent protected in some countries consume 3 to 4 times less bandwidth than G 711 at the expense of increased CPU and memory load and slightly lower sound quality There are numerous other free and licensed codecs on the market each offering a different tradeoff between computational requirements bandwidth and voice quality Pac Time Timeinte A A Each VolP service or endpoint 19 22 45 999980 0 000000 INVITE gt INVITE sip 84952492679 sipne supports several different 19 22 46 057856
21. ithin a SIP call with a measured MOS of 4 0 in one direction and 2 7 in the other direction According to the quality standard MOS of 4 0 means overall good sound quality and demonstrates that the one direction of the call has a perceptible level of distortion that will be noticed by some users but will only disturb a few At the same time the RTP stream in the other direction demonstrates poor quality MOS of 2 7 with varying MOS over time Analyzing network parameters and making certain modifications to the network or choosing a different codec may still improve the quality of this particular VolP network SIP Session x Session RTP Streams 2 Ld Duration RTP Packet aver Total Traffi max l Lost Packets mos Score R Factor Duplicate Sequenc A i RTP Stream x Stream Info Charts mos Score y Y MOS Score enue 588388588 SEES I EESERAENKRARRARA B a f f f fe GSP ral fay tS rs fa f G f re O tS fei Sy Si Si feb G rei gai fat SO f fe tsi si fs fs is tsi t ts ts f tsi tel tel t fey fey vey ted fay i fg o Sy The MOS and R Factor do not tell the whole story about the quality of a particular VoIP system These measurements are merely suggested baseline points for expected call quality in a VolP deployment In order to get an idea on the exact correlation between the numerical measures and the actual quality of the call one has to listen to v
22. ms AS AAA A A aes A a A datos as 3 RN 4 COS ii A Aia 5 Monitoring VolP Neri 6 Status Of CurrentiCall Si a A A A A casa 6 Details about Current Calls iria 6 Bandwidth Utiliza A ads 7 ENCIMA eA ees WE 8 Typical VOIP Problems teinne AA A A A Las Sees E ET as eae ea ad RR 8 Packet LOSS ii cea 8 MEA ii N 9 SEQUENCE EAS ada 10 Codec QUAY a A A A A A TES 11 Assessing Sound Quality ii A A AA AAA te 12 MOSand Ria AAA 12 Actual Call Playback ia iii 13 VoWLAN Specitit FACtorS siesta AA A AA A A dts whew aa ieaeenax evs 14 Number of Clients per AP cccscccccccsssssnsscecececsssssanseeceesesssnsaesececeeseneaeseseceesessueanseseessenesaeauseeeeeeseneaneeeenss 14 A A ea E en edete ieee 15 Nuimberot Retries acta A te ent data 15 Wireless QO0S WM Mii ac di a aa 15 REO A ee 16 About TAMOS dui e aaneen eat eah aae tag eE aege Aeaee E aea aeee Ea Aea aa Sears Aa aaa 17 Protocol Overview This chapter overviews the VolP transport protocol Voice over IP uses the Internet Protocol IP as an underlying transport base Voice is digitized converted to IP packets and transmitted from point to point over an IP network Signaling VoIP standards define numerous signaling protocols that are used to set up and carry out the calls transmit information required to identify and locate remote callers and negotiate carrier and endpoint capabilities Different companies developed numerous different signaling protocols within the VoIP scope SIP and H 32
23. n c IN IP4 192 168 10 1 t 00 m audio 16384 RTP AVP 08 18 4 18 a rtpmap 0 PCMU 8000 1 a rtpmap 8 PCMA 8000 1 a rtpmap 18 G729 8000 1 a rtpmap 4 G723 8000 1 a rtpmap 18 G729 8000 1 15 07 21 322004 0 353433 180 Ringing 1 INVITE 15 07 29 146735 7 624731 200 OK 15 07 29 228749 0 082014 ACK sip 22220192 168 10 2 5060 15 07 40 127723 10 898 gt BYE sip 22220192 168 10 2 5060 15 07 40 202025 0 074302 200 OK 11 Assessing Sound Quality Measuring sound quality by placing a test call and listening to the remote party is highly subjective when developing or deploying a VolP system There are existing formal methods to give qualitative and quantitative assessments to voice quality This chapter explains the methods used to assess the quality of a VolP conversation MOS and R Factor In order to provide quantitative assessment of the quality of VoIP communications the Mean Opinion Score MOS has been introduced The MOS indicates the perceived voice quality of a VolP conversation ranking the call quality as a number in the range 1 to 5 Originally the MOS was meant to represent the arithmetic mean average of all the individual quality assessments given by people who listened to a test phone call and ranked the quality of that call Today human participation is no longer required to determine the quality of the audio stream Modern VolP quality assessment tools employ artificial software models to calcula
24. ntinuously monitor the network under normal and critical loads at least at the deployment stage and while troubleshooting call Stream Bandwidth quality problems Being able to see network bandwidth utilization in real WN o 0O time is essential for understanding the Bandwidth kbps difference between the optimum and current environment often enlightening 0 00 01 0 00 07 0 00 11 network administrators that a simple Packet Intervals 1 traffic shaping can essentially increase the quality of the calls transferred over that network z g a One can identify bandwidth utilization issues by looking at the shape of the Stream Bandwidth chart as illustrated on the screen shot A bandwidth utilization problem can be clearly seen with the odd shape of the bandwidth graph sharp spikes and uneven packet intervals Endpoints There are multiple types of end user endpoint devices such as PC software IP phones VolP adaptors and gateways Endpoints are responsible for originating handling and ending conversations Heterogeneous VolP networks may have multiple types of endpoints with varying degrees of compatibility with the VolP server and between each other This is especially true for SIP based VolP networks A network analyzer identifies and displays the model names of endpoint devices allowing tracking down connectivity problems and often fixing them with a simple software or firmware update without hunting
25. otic sound when a long burst of packets is lost The more advanced algorithms interpolate the gaps producing the best sound quality at the cost of using extra computational resources The best implementation can tolerate up to 20 of packets lost without significant degradation of voice quality While some PLC techniques work better than others no masking technique can compensate for a significant loss of packets When bursts of packets are lost due to network congestion noticeable degradation of call quality occurs In VoIP packets can be discarded for a number of reasons including network congestion line errors and late arrival A network analyzer displays the number of lost packets Better network analyzers display real time charts that allow visualizing the number of dropped packets as well as detect quality degrading bursts Seeing the exact shape of packet loss graphs allows network administrators to choose a Packet Loss Concealment technique that best matches the characteristics of a particular environment as well as to implement measures to reduce packet loss on the network Packet Count y Y Packet Sizes y Y y Total a 50 Total Out of Sequence o RTP Payload a i Duplicates 5 An ar an AR An n RTP Header 2 S z N A amp Network Header 8 8 8 8 8 8 a a Stream Bandwidth y Y Jitter y Y 25 gt 20 0 Jitter Deviation 15 E 40 O Aver
26. otocol RTP for streaming voice packets in real time RTP standard does not define and therefore does not require the use of any specific UDP port leaving it up to endpoints to agree on a certain port to commence a voice call The floating port implementation makes it difficult to traverse firewalls often requiring the use of dedicated STUN servers to synchronize the endpoints A frequent VoIP connectivity problem occurs while attempting to send or receive voice packets via a more or less random port that is blocked by a firewall A network analyzer allows the RTP streams associated with a specific signaling session to be clearly seen as well as the IP addresses and the ports being used for the VoIP call which helps faster and easier deployment of a VoIP system RTP Streams Destip__ StartTme End Time Duration RTP Packe Averag Total Traffi Max Jitter ms 212 55 35 219 86 192 4 88 3 32 26 PM 3 32 50 PM 0 00 24 3 A 11 29 a 213 46 11 66 213 46 11 154 6 08 32 AM 6 08 44 AM 0 00 11 9 3 57 193 232 243 1 86 192 4 88 3 27 58 PM 3 28 25 PM 0 00 27 2 863 19 27 67 314 10 80 212 55 35 254 86 192 4 88 3 09 46 PM 3 10 28 PM 0 00 42 0 2113 83 91 452 182 306 21 212 55 35 219 86 140 116 2 6 52 23PM 6 52 47 PM 0 00 23 3 19 63 58 734 56 28 y gt RTP Stream x Stream Info l Charts A ee Pac Time Time Inte Sequen RTP Timest Payload Name
27. r which in turn causes jitter as described in the previous chapter Sequence errors cause significant degradation of call quality Sequence errors may occur because of the way packets are routed Packets may travel different paths through different IP networks causing different delivery times As a result lower numbered packets may arrive at the endpoint later than higher numbered ones The packets are usually received in the buffer allowing the endpoint to rearrange out of order frames and reconstruct the original signal However the size of internal buffer is limited to control jitter and significant variance in the orderly delivery of packets may cause the endpoints to discard frames resulting in both jitter and dropped packet issues 10 Routing VolP calls through consistent routes to avoid spreading packets from the same call over different paths allows for significant reduction in sequencing errors Codec Quality A codec is software that converts audio signals into digital frames and vice versa Codecs are characterized with different sampling rates and resolutions Different codecs employ different compression methods using different bandwidth and computational requirements Choosing the best codec for particular network conditions may considerably increase the quality of voice calls If the network has low effective bandwidth choosing otherwise great lossless G 711 codec would be a mistake as the quality of the calls would suffer b
28. s for wired and wireless networks that include a VoIP analysis engine gt CommYiew File Search View Tools Settings Rules Help a m Intel R 82566DM Gigabit Network Connection gt E A Beh AA A nec ar Se PIC P A INE PIE E A Pa Latest IP Connections TT Packets Logging Rules Go Alarms El VoIP D SIP Sessions 3 SIP Sessions H 323 Sessions 0 SiclP DesiP Start Time End Time Duration Status ATP Streams 2 210 54 125 221 210 54 125 100 6 52 11 PM 6 52 48 PM 0 00 36 4 Completed Registrations 1 210 54 125 221 210 54 125 100 6 52 12 PM 6 53 51 PM 0 01 39 2 Nota call 210 54 125 221 210 54 125 100 6 54 20 PM 6 54 20 PM 0 00 00 1 Not a call Endpoints 2 Errors 6 Call Logging SIP Session x Report Call Info RIP Streams 2 INVITE sip 8495 a 100 Trying 401 Authenticatic ACK sip 8495245 INVITE sip 8495 100 Trying 183 Session Proc Timing a 18521 INVITE Start Time 8 23 2006 6 52 11 PM End Time 8 23 2006 6 52 48 PM Duration 0 00 36 4 iil yv ee Quality 52 1 INWITE MOS Score Als R Factor 7ARARHAAE Streams Once the peers are located and a connection is made the streaming of voice packets begins to occur In order for the voice conversations to sound natural without echoes and delays the voice packets must be transferred over the IP network in real time All VoIP standards use Real time Transport Pr
29. t is totally possible and relatively easy for a network administrator to control the quality of voice transmission by controlling network parameters such as bandwidth utilization packet loss and delays Therefore in addition to the many factors that are specific to VoIP applications a network traffic analyzer shall be able to monitor network conditions and bandwidth utilization Status of Current Calls A network analyzer shall be able to display the status of all VoIP calls providing additional call details by request The following screen shots illustrate the typical call details available to a network analyzer Details about Current Calls Being able to see the details about current calls allows identifying and resolving connectivity problems on the networking and protocol levels Call details include source and destination IP addresses time of the 6 beginning and end of a conversation call duration call status quality score user agent types as well as additional protocol specific details such as authentication information and codec types One can identify many types of network problems by looking at the call details as illustrated by the following CommView screen shot In addition to the high level and statistical data displayed on the left pane the user has access to the low level actual log of the signaling session on the right pane Session RTP Streams 2
30. te the MOS The MOS is highly subjective One should not make decisions on a VolP system based on the MOS alone Other measurable parameters should be analyzed such as network delay packet loss jitter and so on As an alternative to the MOS a different less subjective rating has been introduced R Factor is an alternative method of assessing call quality Scaling from O to 120 as opposed to the limited scale of 1 to 5 makes R Factor a somewhat more precise tool for measuring voice quality R Factor is calculated by evaluating user perceptions as well as the objective factors that affect the overall quality of a VolP system accounting for the Network R factor and the User R factor separately The following table demonstrates the effect of the MOS and R Factor on the perceived call quality User Satisfaction Level MOS R Factor Maximum using G 711 4 4 93 Very satisfied 4 3 5 0 90 100 Satisfied 4 0 4 3 80 90 Some users Satisfied 3 6 4 0 70 80 Many users dissatisfied 3 1 3 6 60 70 Nearly all users dissatisfied 2 6 3 1 50 60 Not recommended 1 0 2 6 Less than 50 12 Some users believe R Factor to be a more objective measure of the quality of a VolP system than MOS Still a network analyzer should be able to calculate both scores and produce the two assessments for better judgment of the call quality The following screen shot displays the details about two RTP streams one stream in each direction w
Download Pdf Manuals
Related Search