Home

RTaW-Sim User Manual - RealTime-at-Work

1. Architectures ShortName C Bit Rate 1 CAN FD Bit Rate Periodic Load 3 Bit Stuffing 10 OffsetConfig Architecture Buses BusSimConfig ad CAN LS BusEvalConfig ad Ecus ComPatterns Sample Time Scope and Kind of Statistics End to End Canld Cant ShortNa Sender Receivers Paylo TxMo Ro Period Mini Dea TxO S 36 0x24 2 0A frame88 Ecu_12 5 bytes P E 200 ms 50ms a 59 Ox3b 2 0A frame95 Ecu_12 8 bytes P E 100 ms 50 ms 7 73 0x49 2 0A frame32 Ecu_1 8 bytes P 200 ms 119 0x77 2 0A frame26 Ecu_2 8 bytes P 100 ms 122 0x7a 2 0A frame40 Ecu_1 8 bytes P 50 ms 135 0x87 2 0A frame35 Ecu_1 1 bytes P 100 ms 196 0xc4 2 0A frame22 Ecu_2 8 bytes P 200 ms 198 0xc6 2 0A framel Ecu_3 8 bytes P 200 ms 201 0xc9 2 0A frame91 Ecu_11 7 bytes P E 200 ms 50 ms 208 0xd0 2 0A frame62 Ecu_15 8 bytes P 200 ms 209 0Oxd1 2 0A frame52 Ecu_1 3 bytes P 100 ms 215 0xd7 2 0A frame30 Ecu_1 6 bytes P 500 ms 235 0xeb 2 0A frame42 Ecu_1 8 bytes P 500 ms 241 0Oxf1 2 0A frame79 Ecu_17 1 bytes P E 500 ms 50 ms 271 0x10f 20A frame45 Ecul 8 bytes P 500 ms i 274 0x112 2 0A frame96 Ecu_10 5 bytes P E 100 ms 50ms Specific MOSS 295 0x127 20A _frame69__ Ecu 0 Bbytes P E 500 ms 50 ms v Results and Graphics Ut Click on the Architecture panel to get an abstract view of the topology 2009 2014 RTaW 15 182 RTaW Sim V1 4 7 User Manuat
2. 36 37 38 39 40 41 42 Response times in ms Results and Graphics 2009 2014 RTaW 165 182 RTaW Sim V1 4 7 User Manuat 4 7 5 5 Frame dialog response times Frame dialog related response time statistics are accessible under the node DialogTxTimeDatas besides the OccurrenceModels node RTaW Sim v14 7 Pro C Users jorn svn pegase netcar svn NetcarSiml docs help odt material DigAndSegTx DigAndSegTx sim ae a File Samples Simulation Worst Case Optimization What If Plot Tools Architecture A Response times for dialog Dialog after1d Architectures DialogQverCAN CANL Dialog Architecture Buses CANI CAN2 Ene Q 7 252 ms Qs 7 593 ms ECU ECU Q 8 231 ms Max 8 604 ms ECUS GateWay ali Entries Value Count CANL Dialog Dre cere eee eet neater erent a Min 2 110 ms Average 2 983 ms Q2 5 333 ms g3 6 352 ms Sample Size 6640000 UntavorableScenarioConfig UnfavorableScenarioConfig a O CCUITENC ell a gd gl 5 EAST EEA OE E E OA CEA AO E DialogTimeDatas A Response times for dil CANL SegmentedTx 2 146 ms 2 159 ms 2 171 ms 2 184 ms 2 196 ms 2 209 ms T s 2 222 ms Specific Models 2 234 ms 4 mT Results and Graphics Data Histogram Observed Unfavorable Scenario _ The lower part shows the table view of the histogram which can also be visualized as a graphic by selecting the Histogram tab 2009 2014 RTaWw 166 182 RTa
3. RTaW Sim v1 4 2 Pro Model not yet saved n File Samples Simulation Worst Case Optimization What If Plot Tools Architecture Architecture amp CAN LS Architectures Architecture Ecu_4 Buses Ecu_5 Ecu_l2 CAN LS E j g ComPatterns Specific Models Results and Graphics The window shows the name and the speed of the bus in kbits s as explained by the tool tip of the label Speed Furthermore the periodic bus load is displayed for the chosen bit stuffing 10 means that the bit stuffed part of the frame is 10 longer than nominal With the worst case option Bit Stuffing 10 ka the increase is about 25 Below the frames that are transmitted of the bus are displayed in a table with a column for each of the frame related properties Now we want to perform a simulation in order to obtain statistics on frame response times but first we need to save the sample file into one of your personal folders sample files are stored in a directory that depending on the OS may not be writable by a regular user 2009 2014 RTaW 16 182 RTaW Sim V1 4 7 User Manuat Samples Simulation W Open Import Arc Recent files New Merge Import Save Save As iS Quit Then open the simulation configuration dialog for a CAN bus Simulation Worst Case Optimization What If Plot Tools Simulation Architecture mj Initialize from existing second Run Simula
4. 2009 2014 RTaw 96 182 RTaW Sim V1 4 7 User Manuat archi 2s An ECU can also be connected to bus in two ways either in the exploration tree or in the architecture graph panel For the first alternative right click on the BusConnections node under Ecul in the exploration tree and select Add New 4 Architectures 4 archl 4 Buses gt busl 4 Ecus 4ecul BusConnectiong GateWays Add New gt ecuZ ComPatterns This brings up the following dialog New BusConnec BusConnection Bus eel Cancel The only possible choice busl is selected by default click on Create in order to make the details pane appear on the right 2009 2014 RTaw 97 182 RTaW Sim V1 4 7 User Manuat archl ecul busl amp Ecu ecul Bus bus SentFrames ShortNa Canld CanType Payload TxMode Period Minimu Deadline ReceivedFrames ShortNa Canld CanType Payload TxMode Period Minimu Deadline For the second alternative right click on an Ecu node in the architecture graph panel and select Connect to archl amp ecul busl ecu Show details Connect to bus iS Delete The new bus connection appears archl amp ecul busl To disconnect an ECU from a bus two methods are possible 2009 2014 RTaw 98 182 RTaW Sim V1 4 7 User Manuat e right click on the ECU to disconnect on the right panel and sel
5. RandomComOffset0 RandomConstantDrifts 1000ppm convergence of statistic Q 1 10 3 2 5 ms 2 25 ms 2 ms J 1 75 ms a LA in J 1 25 ms j 1 ms 4 Response Times in ms 0 75 ms 4 0 5 ms 1 0 29 Ms 0 ms i T np T T T T T 7 T T T T T T T T T T T F T T T T T T T T T T T T T T 1 21 791031942843244235 0553459865 16836957208 00887 90094 597A 24829 33314565375 88005 03874 F829 37952014032 Frames in increasing order of identifiers Q 1 10 3 1m O 1 10 3 1h Of1 10 3 12h Of 1 10 3 22h O 1 10 3 14 In order to check this click on the Data tab on the left bottom 2009 2014 RTaWw 53 182 RTaW Sim V1 4 7 User Manuat 0 00 1 224 lt lt lt lt Data Graphic and hide the curves with the shorter sample time 1m 1h and 12h Curves ShortName DisplayC Hidden Sco Qi1 10 3 1m X black QO L 10 3 1h red Q 1 10 3 12h green Q 1 10 3 22h blue Q 1 10 3 1d silver The resulting graph shows that between 22h and 24h the statistics have changed very little meaning that they are very likely to be close to the actual values A bus_2 Stuffing 10 HPF SOA5 RandomComOffset0 RandomConstantDrifts 1000ppm convergence of statistic Q 1 10 3 2 0 M5 2 29 M5 2ms 1 75 ms 1 9 ms 7 1 29 ms 1 ms Response Times in ms 0 73 ms 3 0 9 ms 0 25 M57 0 ms
6. cc cececececeeeeeeeeeeeeeeteaesteeeaneateaneateaneaneas 100 a P E AVG E T EE ENEE E E E E 102 443 Frame Gateway S orerar EEEE 106 4 4 4 Bus interfaces configuration s sssesssesrrrssrrrrrrssrrrrrrrrserrrrrrrerne 107 4 4 5 Transmission error occurrence ModelS ssssssssesrrresrrrsrrrreeen 112 4 4 6 Communication pattern sssssssssssrresrrrrrrrrsrrrrrsrrrrrrrrrerrrrrrnserne 112 4 4 6 1 Segmented TTANSMISSION sssssssssrrsrrrrsrrrsrrrrsrrrrrrrrrrssrrrrrrrerne 113 4 4 6 2 Frame DialogS ssssssserssrrrsrrrerrrrsrrrrerrresrrnerrrrrerrrrrrnrrsrrrrrrrenne 118 4 5 Open or import an existing Mmodel sssssssssssrrresssrrrrrrssrerrrrreserrno 123 2009 2014 RTaW 2 182 RTaW Sim V1 4 7 User Manuat 4 6 Performing a SIMULATION ccccececcee sees eeeeeeeeeeeeeseeaeeeeeeeateaeeaneaneas 123 A Ou Status UCA CONCEDES serere AETA AEON 123 4 6 2 Required system model ssssssssssressrrssrrrrrsrrrrrrrssrrrrrrerserrrrrresrne 127 4 6 3 Configuring and running a bus SIMULATION cceceee eee ee eee eae ees 128 4 6 4 Configuring a second run SIMULATION s ssssssserssrrssrrrrrrrrrrrerne 133 4 6 5 Command line execution of a simulation ssssesssssesssrrerrresen 134 4 7 Exploring performance evaluation resultS ssssssessssrrrrrrrseren 138 4 7 1 Table view of frame response tIMES ccccceceeeeeeeeseeeeeeeeeeeeneaes 138 4 7 2 Histogram view Of statiStiCS sssssssresrrrssrr
7. 2009 2014 RTaW Specified then the ComOffsetGenerationSeed property must be set The random number generator seed used to generate COM task offsets Only needed if ComOffsetConfig is not provided The name of a clock drift configuration to be used in the simulation If no such configuration is specified then the ClockDriftMode property must be given There are two modes e NODRIFT clocks are ideal and run exactly at the nominal speed e CONSTANTDRIFT every clock has a fixed drift factor that is generated based on the ClockDriftBound and the ClockDriftGenerationSeed which must both be specified Remember that the 7 clock drift factor defines the actual speed of a clock where 0 8 means 20 slower whereas 1 5 means 50 faster Generated clock drift factors are randomly chosen in an interval of the form 1 delta 1 delta with delta ClockDriftBound 106 because the unit of the ClockDriftBound is ppm parts per million The random number generator seed used to generate clock drift factors The name of the frame transmission error model to be used to be used in the Simulation The name of the occurrence model for event driven transmissions to use in the Simulation List of names of communication pattern occurrence models to used in the 137 182 RTaW Sim V1 4 7 User Manuat Simulation configuration properties Simulation GatewayX Name of gateway that needs to be configured GatewayX GatewayFra Name of t
8. A1 Second Statistics for System Simulation A 1 Staying above bus response time quantile for frame4 on body aft 2 F framed Specific Models y le AboveQ2 Stay times above Q2 BelowQ2 Stay times below Q2 Results and Graphics a a Simulations AboveQ3 Stay times above Q3 BelowQ3 Stay times below Q3 SIMUations A System Simulation AboveQ4 Stay times above Q4 BelowQ4 Stay times below Q4 a SecondRunsimulations AboveQ5 Stay times above Q5 BelowQ5 Stay times below Q5 4 A 1 Second Statistics for System Simulatio 4 BusSimulations AboveQ6 Stay times above Q6 BelowQ6 Stay times below Q6 4 A 1 body Stuffing 10 Interf 4 BusStatistics 4 12d 4 FrameRespTimesAboveC forwarded_frame238 forwarded_frame2 6 forwarded_frame284 forwarded_frame287 forwarded _frame300 framed framel frame2 frames framed framed frame Cn Ea dii p For each of the quantiles there is a histogram for the length of the intervals where the response times are consecutively above the quantile AboveQn and where they are consecutively below the quantile BelowQn The following shows the histogram of the intervals above the quantile Q4 The frame frame4 has a period of 50ms and thus a value of about 100 ms means that two consecutive response times exceed the quantile Q4 As can be seen most of the times the value is about 50 ms meaning that a response time above t
9. HPF SOAS RandomComOffset0 Ra Statistic Scope And Kind of Times Bus Response Ti Sample Time 1d and the second Bus 2 end 2 end delays with Scope and Kind set to End to End Response Time wn LZ es FrameStatCurve ShortName Bus_2 local delays DisplayColor blue Bussimulation A bus_2 Stuffing 10 HPF SOAS RandomComOffset Ra Statistic Maximum gt Scope And Kind of Times k Sample Time 1d Notice that if you want to change the parameters of an already created curve you can open the editing pane through the context menu Curves ShortName DisplayC Hidden Scope n Bus Bus 2 local delays black show Bus Resp bus_2 a ee Adapt column widths to contents Edit selected Curve Delete selected Curve Add Mew Curve Copy to Clipboard in CSV Format The result should be the following after having clicked on the Graphic tab in the lower left corner of the pane 2009 2014 RTaWw 56 182 RTaW Sim V1 4 7 User Manuat Bus_2 end 2 end delays 14 ms 13 ms 12 ms 11 ms Ji w lO ms 9 ms 8 ms 7 ms 6 ms 5 ms Response Times in m 4ms 1 3 ms J 2 ms 1 ms 4 0 ms 7 1 79 i94 3234 505 598 683 720 887 5 12343 1331 1537 1605 1747 Frames in increasing order of identifiers Bus_7 local delays Bus_2 end 2 end delays The black curve shows the response time of the frames on bus
10. RTaW Sim V1 4 7 User Manuat Low Speed ii ShortName Low Speed S5peed 125 Periodic Load 0 Bit Stuffing 10 BusSimConfig BusEvalConfig a Sample Time hd Scope and Kind of Statistics Bus Response Times bd Canld n ShortName Type Sender Receivers Paylo Period Mini Offsets Min Average Q The creation of a new data entity can also be initiated through the context menu of the tables that can be found in the details panes Let us take the example of frames in the bus details pane right click anywhere in the Frames table and select the Add New Frame entry Canld ShortName Type Sender Show Defavorable Scenario Adapt column widths to contents Edit selected Frame Delete selected Frarne Add New Frarne Copy to Clipboard in CSV Format to bring up the frame creation dialog New Frame E Canld shortName Framel Payload Deadline Type P E Period 100 MinimumDelay 20 Create Cancel b Remember that the same frame creation dialog can also be obtained through the context menu of the relation node Frames under the bus entity node 2009 2014 RTaW 88 182 RTaW Sim V1 4 7 User Manuat Low a Architecture Buses E Low Speed Frames riFee Fins abk ambtine As soon as you have successfully created a new frame it appears in the table Canld ShortName Type Sender Receivers Payload Period MinimumDelay 123 07
11. Target i a SentFrames CANL A Le Only bridged frame can be target of a frame mapping The reason is then that the frame has not the required TxMode B for this See Section 4 4 2 on how to change the TxMode 4 4 4 Bus interfaces configuration The role of a CANBusInterfacesConfigs entity is to describe efficiently one specific configuration set for all bus interfaces connected to a certain bus For this reason these entities are shown below the concerned bus node 2009 2014 RTaWw 107 182 RTaW Sim V1 4 7 User Manuat Architecture a a Buses a CAN LS gt Frames gt BusConnections 4 CanBusInterfacesContigs Default gt OffsetConfigs gt ClockDriftContigurations gt ComOffsetsContigs a Ecus gt Ecu_0 gt Ecu 1 gt Ecu 2 A CANBusInterfacesConfigs entity defines the configuration of each bus interface on the bus and this in a compact way through the usage of a default configuration that applies unless a more specific configuration is defined By double clicking on the node that represents a CANBusInterfacesConfigs called Default in the screenshot above one can visualize its details CAN LS Default 25 ShortName Default DefaultConfig Queuing HPF TxBufferCount 3 UseHwCancellation W Queue BufferDelay us SingleShot T KOrder Type Highest Priority First Order Ti SpecificConfigs BusConnection Queuing TxBufferCount UseHwCancellation S
12. gt CanBusinterfacesContigs Pitt ab mi a is All contained sub entities such as buses ECUsS communication patterns frames communication stack related configuration parameter sets etc are also duplicated Thus copies of architectures are independent and can be modified separately This is useful for analyzing and comparing architectural alternatives 4 4 1 2 Creation from scratch Let us build a system from scratch right click on the Architecture node in the exploration tree Architecture a Architectures 2009 2014 RTaWw 93 182 RTaW Sim V1 4 7 User Manuat and select Add new New Architecture Architecture ShortName archl In the creation dialog provide a name for the new architecture and select Create Architecture arch 23 4 Architectures 4 archl Buses ECUS ComPatterns A new panel archl is created There are two alternatives for creating new buses and Ecus Either through a right click on the Buses or Ecus node in the exploration tree Architecture amp J archl 4 Architectures 4 archl Buses l Ecus Add New ComPatterns or through the context menu of the architecture panel on the right by selecting the Add new bus entry 2009 2014 RTaWw 94 182 RTaW Sim V1 4 7 User Manuat arch es Add new ECU Add new bus Layout as ci
13. 4 2 1 4 4 From Tracelnspector This functionality allows to import parameter sets that have been estimated from a communication trace with RlaW Tracelnspector The parameter sets include e clock drifts e transmission offsets e inter occurrence time distribution of event driven transmissions e inter occurrence count and burst size distribution of transmission errors In order to do so make first sure to have opened the RTaW Sim file into which the parameters shall be imported Then selected the 2009 2014 RTaW 70 182 RTaW Sim V1 4 7 User Manuat menu entry File gt Merge Import gt From Tracelnspector You will be asked to select the RlaW Tracelnspector file insp form which to import the parameter sets Then the specified file will be parsed and you will be invited to select the trace inspection containing the parameter sets to import 8 Selection of paramter sets to import em Traces CAN LS simulation trace 4 Tracelnspections Inspection0 Inspection1 Then you will be asked to specify the corresponding bus in RlaW Sim where the parameters sets shall be copied to Selection of the corresponding bus m Architectures 4 Architecture 20A a Buses body50 4 Architecture 206 Architecture 204 amp 3 4 Buses body50 Architecture FD basic a Buses body50 Architecture FD extended a Buses body50 The matching of frames and ECUs between RlaW Tracelnspector and RTaW Sim is ba
14. 4 2 6 5 Scale payload CAN FD This function allows to increase the frame payloads by multiplying them by the factor CAN FD Speed Standard CAN speed The rationale of this function is to evaluate the extent to which the additional bandwidth offered by CAN FD can actually be taken advantage of to transmit larger data payloads Some good insight into this question can be obtained with the response times of the CAN FD network with the larger data payload When you select this menu entry the following dialog appears CAN FD payload scaling Architecture Eaa Yes Cancel You have to select an architecture Notice e all frames of the selected architecture must have the type CAN FD e the functions fails if due to the scaling a payload would be higher than 64 bytes 4 2 7 Plot See Section 4 7 4 1 4 2 8 Tools Several useful tools are available in the menu Tools 2009 2014 RTaw 83 182 RTaW Sim V1 4 7 User Manuat Plot Anonymize Names Generate Random Offsets Generate Frame Deadlines Make all frames CAN2 0B Scale Load 4 2 8 1 Anonymization of Names This function changes potentially confidential names into anonymous names of the form Type N For example frames will be renamed to Frame_1 Frame 2 etc This function is useful when a configuration file has to be transmitted outside a company institution and the names of the frames and nodes cannot be c
15. Offsets DOAS Com Offset RandomComOftset0 Drift RandomDrift lppm Tx Errors NonBur uffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift lppm Tx Errors BurstyT t To finish change the default Statistic to Maximum the default Sample Time to 1d and click Create 2009 2014 RTaW 45 182 RTaW Sim V1 4 7 User Manuat Nes a ShortName Wo errors DisplayColor black BusSimulation Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset Re Statistic Maximum Scope And Kind of Times Bus Response Ti Sample Time Then create a second graph but enter Non bursty errors as name and select red as color As entry for the BusSimulation field select the simulation that corresponds to the case with non bursty transmission errors A5 Com Offset RandomComOffset0 Drift No drift AY Com Offset RandomComOffset Drift RandomDrift 1ppm AS Com Offset RandomComOffset Drift RandomDrift lppm Events InFrequentEvents AS Com Offset RandomComOftset Drift RandomDrift lppm Events FrequentEvents AS Com Offset RandomComOffset Drift RandomDrift lppm Tx Errors NonBurstyT ransmissionErrorModel AS Com Offset RandomComOftset Drift RandomDrift 1ppm
16. RTaWw Sim Wser Manual Contact timing tools realtimeatwork com Product homepage at http www realtimeatwork corm software rtaw sim R Lavv Real Time at Work We help designers build truly safe and optimized systems R law Real Time at Work RTaW Sim V1 4 7 User Manual Date 18 6 2014 Contents sl aye gee 0 cio 6 eee enn ne ee ee ne re nee ee ee ee ee ee 5 1 1 License Of the SOTTWALE cc cecceceec ee eceeteeeseeeeaeeaneaeeaeeaeeaneaneaneaneaes 5 1 2 PROV ING CO WS ria saa ceseaens race OTIT 6 Lo WSCA AE O er rra EEEE E AE T 8 LA UDD O T ea E E AE re 3 1 5 New releases ANd UPCAtESS ccceccccececeeeeceeeeeeeeeeeeeeeeeegeegeegeeeneaneaneaes 8 LO CM AGEGE A E E 9 2 GOSS AY ar E T EEE EE A 12 BO E E E E A EE E 13 3 1 Evaluating frame response tiMe S s ssssssrrrssrrrrrrrsrrrrsrrrerrrrsrrrrene 14 Sl ASUT S a U a E E A E S 20 3 3 Analyzing the effects of clock driftS sssssssssssrrrrrrrssrrrrrrrsserrere 26 3 4 Analyzing the effects of event triggered transmiSSiOnsS 34 3 5 Analyzing the effects of transmission errOfS sssssssrrrerssseerrrrrrne 41 3 6 Simulation with gatew ayS sesssssssrsssrrssrrrsrrresrrrenerrrrrserrrrrneeserrne 48 4 Reference Manual ssssssssssrresrererrrrsrrrerrrrsrrrnsrrrerrrnerrrnrrrrerrrrerrrnrrrrr gt 58 4 1 Overview of the GUl ssssssssssssssressnnssnnnssnnnnnnnsennnnsennennnnennnnnereesenneenne 59 eZ WON S ooreo rrr E T E
17. the Sample interval being the entire simulation horizon Additional times can be specified by entering values in the fields to the left of the Intermediate Statistics label with the time units described in Table 2 For intermediate statistics the sample interval spans from the beginning of the simulation until the specified time After having entered a value do not forget to push the Add button to actually 2009 2014 RTaw 131 182 RTaW Sim V1 4 7 User Manuat add the value to the list The buttons Update and Delete work on selected entries of the list Include bus off blocked is the recommended default behavior It means that response times of bus off blocked frame instances are included into the statistics As a result if bus off occurs frames that have been blocked will have response times larger than their period since the transmission of the data had to wait at least one period If this option is unselected response times of blocked frames are not accounted for in the statistics while the sender is in the bus off state The Frame delay distribution parameter determines how the delays induced by the gateway are generated by the simulator a Simulate D 7 So File Length amp Sample Times Simulate the traffic on bus bus_2 gt bus_1 4 Buses bus_1 Frame delay distribution Uniform L00 200 bus 2 4 Gateways bus 2 gt bus 1 bus 1 gt bus Notice that the poss
18. 1 21 79 1031942843244 2350553499805 16836957208 0688790094 597 A 24829933145 65375 89005 038 74 F82 893795201203 2 Frames in increasing order of identifiers p 1 10 3 22h Q 1 10 3 1d Let us now create a graph that allows to study end to end response time that is response times that include gateway delays 2009 2014 RTaW 54 182 RTaW Sim V1 4 7 User Manuat when the receiver is not located on the same bus as the sender For this purpose left click on the Results and Graphics tab in the lower left corner ecu Iy T Ecu 20 0 Ecu 21 0 Specific Models Results and Graphic and then right click on the Graphics node in order to create a new graphic 4 Simulations gt A System Simulation Evaluations 4 Graphic TA called Bus 2 end 2 end delays Then left click on its Data tab Data Graphic in order to add curves ShortName Bus_2 end to end delays Relative to period Ty Curves ShortName DisplayC Hidden ScopeAn Adapt column widths to contents Copy to Clipboard in CSV Format Show selected Curve Delete selected Curve Add Ne y Curve Entitle the first one Bus 2 local delays with Scope and Kind set to Bus Response Time 2009 2014 RTaw 55 182 RTaW Sim V1 4 7 User Manuat FrameStatCurve ShortName Bus_2 local delays DisplayColor black BusSimulation A bus_2 Stuffing 10
19. 2009 2014 RTaw 76 182 RTaW Sim V1 4 7 User Manuat Randomly generated offsets are in general much less effective than optimized offsets or offset generated by a heuristic but they are useful for benchmarking offset configurations and gaining a good insight into how useful it is to implement offsets 4 2 5 3 SOA Offsets This functionality allows to generate very good transmission offsets with the help of RTaW s SOPA heuristic Offset generation configuration Architecture Architecture CAN LS Basis Modify Basis Offset Granularity 25 Ves Cancel b First select an Architecture and then the Bus for which the transmission offsets should be generated The granularity must be a common divider of the periods of all frames the combo box only contains these kinds of values and all offsets will be a multiple of the granularity The parameter Ecu allows to restrict the generation to a single Ecu The parameter Basis allows to specify an existing transmission offset configuration that will be used as basis If Modify Basis is checked then the offset configuration specified as Basis will be modified otherwise a copy of the Basis will be created and modified Application notes In order to generate a partial offset configuration and or an offset configuration with different granularities for certain ECUs you can start with a zero offset configuration which
20. Com Offset RandomComOffset Drift No drift Mono Bus CAN LS I Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift and you will obtain the following 2009 2014 RTaWw 30 182 RTaW Sim V1 4 7 User Manuat Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift 1ppm convergence of statistic Maximum 3 0 ms Response Times in ms a 36 135 209 274 314 348 460 487 324 568 590 627 709 756 786 876 915 O42 O83 1211 1445 Frames in increasing order of identifiers Maximum 5s Maximum ih Maximum 6h Maximum 1d This time we see a graphic where the points of the graphs that correspond to longer sample times are situated above or at the points of the graphs that correspond to longer sample times Notice that you may have a closer look by zooming on a specific area of the graphic For this purpose select an area with the left mouse button 487 524 568 590 627 709 756 res i Frames in increasing order of identifiers 2009 2014 RTaWw 31 182 RTaW Sim V1 4 7 User Manuat and only the selected portion will be shown Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift lppm convergence of statistic Maximum wow fk n mM HA p O A O A ase O earl i Response Times in ms mO wW amp in 560 568 579
21. The Csv Import samples can be used as basis for CSV import files see Section 4 2 1 1 4 2009 2014 RTaWw 72 182 RTaW Sim V1 4 7 User Manual Simulation Worst Case Optimization What If Plot Tools Case Studies e fe Tutorials Csv Import P SampleCsvForlmportxls B l SampleCsvForlmport_CAN FD xls SampleCsvForlmport_CANZOB xls 4 2 3 Simulation This menu contains several entries for configuring and starting Simulations Simulation Worst Case Optimizatic Simulation Initialize from existing Second Run Simulation The Simulation entry allows to configure a simulation from scratch see Section 4 6 for more details It contains a sub menu with an entry for every existing Architecture File Samples Worst Case Optimization What If Plot Architecture Simulation i 0 Architecture Initialize from existing aL ba gp Second Run Simulation Selecting for example the entry SO opens the simulation configuration dialog for S0 The Initialize from existing menu entry allows to initialize the simulation configuration from existing simulation results This allows to easily and quickly perform a simulation with few configuration changes The Second Run Simulation menu entry allows to perform a second run simulation for statistical purpose see Section 4 6 4 for more details 4 2 4 Worst Case This menu contains an entry for comput
22. 111 182 RTaW Sim V1 4 7 User Manuat fi to a transmission buffer 4 4 5 Transmission error occurrence models Transmission error models are defined in the Specific Models section under the node TxErrorModels RTaW Sim v1 4 2 Pro Model not yet saved File Samples Simulation Worst Case Optimization What If Plot Tools Architecture BurstyTransmissionErrorModel amp Pad Ph specific Models ShortName BurstyTransmissionErrorMor gt ProbabilityLaws l l InterBurstCount Exp lambda 1 100 BurstLength BurstLength gt EventOccurrenceModels 4 TxErrorMadels ErrorRecoverTimelnBits BurstyTransmissionErrorModel FER 1 1 BER 1 2e 04 NonBurstyTransmissionErrorMode A transmission error occurrence model is mainly characterized by two probability distributions see also Section 5 4 e InterBurtsCount number of error free transmissions between to error bursts e BurstLength number of consecutive transmission errors The field FER shows the frame error rate induced by these two distributions whereas the field BER shows the bit error rate based on the FER and an assumed frame size of 100 bits Actual BER values vary in a range of 50 around the displayed value An optional third parameter is the time in bits until normal Operation after the occurrence of a transmission error If the parameter is not provided then RTaW Sim provides a
23. 2 whether the frame is sent by a local ECU or is actually forwarded by a gateway Notice that all response times are below 3 25 ms The blue curve shows the response time from the sender to the receiver on bus 2 which includes the response time on bus 1 and the gateway delay if the frame is forwarded by the gateway otherwise the two response times are the same Thus the forwarded frame can be identified by a higher blue curve value Notice that for these frames the end to end response time is generally more than the double of the bus 2 local response time This might suggest that the maximal response time of the two buses can simply be summed up in order to find the end to end response time In order to check this let us also draw the bus _ 1 local times FrameStatCunve ShortName Bus 1 local delays DisplayColor green r BusSimulation 4A Bus bus_1 Stuffing 10 Offsets SO4 5 Com Offset Rando Statistic Maximum Scope And Kind of Times Bus Response Ti Sample Time W Gene Cone 2009 2014 RTaWw 57 182 RTaW Sim V1 4 7 User Manuat This should result in the following graphic Bus_2 end 2 end delays Response Times in ms 5 2 37 74 1141381732202593013293603 934294624 p641 5075900200877 14704797 84787 2919946981121401 Frames in increasing order of identifiers Bus 2 local delays Bus_ 2 end 2 end delays Bus_1 local delays Notice that the frames which
24. 410 019a Frame 19a P Ecu 12 Add New Frame 413 0x19d Frame 19d P Ecu_ 8 Martie Famelbc B Ecu4 Gat COPY to Clipboard in CSV Format The considered frame is highlighted in orange in Gantt chart of the unfavorable scenario Frame_15c 248 0x15 ER Frame 15a 346 0x15a __ Frame_357 855 0357 TEDES eee eeeen wiueamne ced Frame_299 665 0x299 PNA rele eal see Ems Frame Ide 478 0x1 de Ea ae x Frame el 225 0xe1 l easa I Frame el 224 0xe0 Frame 18e 398 0118 T rae ne sss Frame 18d 39 0x18d a Jaume Im 14s 42ms Ims 4 7 4 Frame statistics visualized as graphs The values of the statistics can also be shown as graphs where the X axis represents the frames in increasing order of their identifiers displayed in decimal format and the y axis represents a statistic of their response times in milliseconds Each point on a curve is the value of a certain statistic for a certain frame The following graphic shows for each frame x axis the corresponding maximal green graph and average black graph response time for a simulation length of 1 day 2009 2014 RTaWw 143 182 RTaW Sim V1 4 7 User Manuat Maximum of response times after 1d 9 5 ms 9 ms 8 9 ms Sms 7 3 ms a 6 9 ms on a 5 9 M5 4 9 ms 3 0 Ms 3 ms End to End Response Times in ms LA a 2 0 MS 2ms 1 5 ms 1 ms 0 5 ms 0 ms F T T T T T T T T T T T T T T T T T T T T 36 135 209 274 314
25. 431 2 104 2 710 2 971 3 077 0 572 1 186 3 111 3 932 3 784 3 974 100 60 0 702 1 004 2 648 3 148 3 394 3 394 100 40 0 702 1 005 2 648 3 111 3 258 3 394 100 0 0 302 0 550 2 128 2588 2173 2 804 100 20 0 702 1 007 2 648 3 221 3 491 3 636 100 50 0 218 0 439 2 178 2 904 3 258 3 404 100 45 0 148 0 360 2 009 2 618 3 006 3 109 100 90 0 182 0 402 2 153 2 904 3 221 3 295 100 80 0 236 0 457 2 228 2 937 3 258 3 675 100 70 0 218 0 438 2 178 2 937 3 373 3 563 The reader interested in more information about how to use quantiles for validating complex automotive CAN based communication architectures can refer to reference 6 4 6 2 Required system model To be able to perform a simulation of a system the description of the later must be complete to a certain degree The following entities are the strict minimum that needs to be defined e abus e the frames that are sent over the bus e abus interfaces configuration see Section 4 4 4 e an offset configuration for the frames non specified offsets are considered to be zero e and finally the ECUs their connections to the bus and the frames they send Please consult Section 4 3 on how to create and edit data entities in general Section 4 4 about additional information regarding the creation of some complex data entities and Section 4 2 1 1 on how to 2009 2014 RTaw 127 182 RTaW Sim V1 4 7 User Manuat import a system description available in other format such as for example CSV Notice
26. 938 ms p ECUS gt GateWay Sample Size 2045581 UnfavorableScenarioConfig i UnfavorableScenarioConfig 4 ComPatterns t CANIL Dialog 4 CANL SegmentedTx t OccurrenceModels 32 111 ms 4 SeqmentedTxTimeDatas 32 297 ms A Response times for E 32 483 ms 32 671 ms 32 859 ms 33 049 ms 33 239 ms 33 431 ms 33 624 ms 33 818 ms Entries o The observed Throughput is displayed in the upper left corner of the Data tab It is the observed average payload transmission rate In the example of the screenshot the payload was 111 bytes which are transmitted on average in 36 586 ms which means approximately 111 bytes 0 036586 s 3033 bytes s 3 0 kbytes s The lower part shows the table view of the histogram which can also be visualized as a graphic by selecting the Histogram tab RTaW Sim v1 4 7 Pro CAUsers jorn svn pegase netcar svn NetcarSiml docs help odt material DigAndSegTx DigAndSegTx sim File Samples Simulation Worst Case Optimization What If Plot Tools Architecture Architecture A Response times for Segmented Transmission SegmentedTx after 1d ei Architectures ee Response times for segmented transmission Buses CAN 1 Seg mentedTx CANI Ecus ECUL ECU ECU GateWay ComPatterns CANL Dialog CANI SegmentedTx OccurrenceModels Segmented TxTimeDatas 0 020 4 A Response times for 5g 0 0154 0 010 0 005 4 0 000
27. Ecu 3 Adapt column widths to contents 590 hale 2 04 frame Ecu_3 Copy to Clipboard in CSV Format 333 d d 2 04 frames Ecu_3 Show Timing Chain 6 0 0x566 20A framed Ecu F Show Histogram eee anti aan a Show Unfavorable Scenario 578 0x24 2 04 frame Ecu_ 3 876 086c 20A frame Eou 7 ae 756 024 20A framed Ecu_2 Create or 314 Oxd3a 20A framed Ecu_ 2 Delete selected Frame 709 Ox2cS 20A framel0 Ecu_2 Add New Frame 414 01 9e 20A framell Ecu 7 se Before actually creating the copy a dialog allows to set the name of the duplicata which you should change to be able to distinguish the copy from the original Name of copy 2009 2014 RTaW 90 182 RTaW Sim V1 4 7 User Manuat 4 3 2 Modification Existing data entities can be modified in their details pane on the right hand side of the user interface Remember that the details pane of a data entity can be accessed by double clicking on the corresponding node in the relation tree under the expand bars on the left The details panes can also be accessed through the context menu in a table right click on a line Below the example of the Frames property in a bus details pane Canld m ShortName Type Sender Recervers Paylo Period Mini 123 Ox b Framel P E T 100 20 231 Oxe7 Frame P e amm l 371 hd4 Frame p Show Defavorable Scenario Q Adapt column widths to contents Edit selected Frame Delete s
28. Frame TxMode is Seg CF Consecutive Frame REQ Request frame Must be provided if RES Response frame TxMode is Dig P TxMode Undefined value Disables checking of consistency constraints P Periodic transmission Requires values for payload and period No role must be provided E Event driven Requires values for transmission Payload and 2009 2014 RTaw 104 182 RTaW Sim V1 4 7 User Manuat MinimumDelay No role must be provided P E Mixed transmission Requires values for Payload Period and MinimumDelay No role must be provided B Forwarded by frame No values must be gateway provided neither for Payload Period minimum delay nor Role since they are inherited Seg Part of segmented Requires a value for transmission Role No values must be provided neither for Payload Period nor minimum delay since they are defined by the corresponding communication pattern Dlg Part of frame dialog Requires a value for Role and payload No values must be provided neither for Period nor minimum delay since they are defined by the corresponding communication pattern 2009 2014 RTaw 105 182 RTaW Sim V1 4 7 User Manuat 4 4 3 Frame gateways Frame gateways are defined at the level of ECUs In order to create a new frame gateway right click on GateWay node below the node of the ECU to which you wish to add it Architecture x a Buses CANT C
29. Gate 8 m 0 225 xel Frame el P Ecu 5 Ecu 4 Gate B 10 0 P 0 Ecu 9 397 Oxl d Frame_l d Ecu 3 396 de Frame 18e Ecu_3 404 0x94 Frame_194 Ecu_3 410 0x19a Frame_19a Ecu 12 oer errr Tee ere ere rtPrrrrrrer Eererrerr TePerrerenererrrererer Ererrerrrererrrrr i Peererrern lerererrerrrerrerrt Peer rerrr errrrrre rrr errrnt freer rrr re ty i a i i r i a POPP CEP Oe CUPP OEP OCEP PPO PEPOP OL PPP EECEPP SCOPE POPErCEPPCePerreer EPP e Cer C CPP EOOEPOLOPEPEEEeer PEPE rPOCePOCEPrCEePCEereLEeCerroeroEe TEP PPOOCEOPEOCOPOPEOLEPrerrrrrs e ee a aaa aa as raaa i i EEEF ETTE Tf er rerrerrrrrrrrery errerrrrr ve Terrerererrerrrererrerrerrrrrrrrrrrrrrrereererrrrn COrrrrrrerPeErerrrrerererrerrr CEerrrrereerrrerrrrererrrerrr T sence ee dee eee eee Bee ee eee eh eee ee ee ee eee bee ee eee eh ee ee ee ee de eee ee ee bee ee ee te eee ede ee eee eb ee ee ee te ee ee eee de ee ee eck ee eee ee de a i 045 0 50 0 55 0 60 0 65 0 70 0 75 0 60 0 85 0 90 0 95 1 00 1 05 1 10 1 15 1 20 Values By clicking on the Data tab in the lower left corner one can also visualize the corresponding numerical values of the histogram entries 2009 2014 RTaWw 141 182 RTaW Sim V1 4 7 User Manuat bus 1 Bus response times for Frame_15c afterld 2 Resolution 10 Min 0 442 Average 0 496 Quantigo 0 925 Quantil999 1125 Max 1 167 UnfavorableScenarioContig UnfavorableScenarioContig
30. If a frame is forwarded by a gateway then the displayed response times span form the instantiation of the frame in the first ECU that sends the frame into the network of buses until the transmission end on the considered bus As particular case if a frame is not forwarded by a gateway on the considered bus then the displayed response time is actually the bus local response time Inter Transmission Time The value displayed for a frame is the delay between two Successive transmission ends of 2009 2014 RTaw 146 182 RTaW Sim V1 4 7 User Manuat that frame on the considered bus Every curve of a graphic has its own ScopeAndKind property The value set at the level of graphic is used as default value when curves are created Enter Statistics as short name select Bus Response Time and click on Create AS a result an empty graphic will appear Statistics 23 Statistics 1 05 ms 1 ms jf 0 95 ms 0 9 ms 0 85 ms 4 0 6 ms 4 0 75 ms 0 35 ms 0 3 ms 0 25 ms 7 0 2 ms 7 0 15 ms 0 1 ms 0 05 ms 0 ms Frames in increasing order of identifiers Data Graphic To create a graph left click on the Data tab Data Graphic In order to create curves right click in the curves table and select Add New Curve 2009 2014 RTaw 147 182 RTaW Sim V1 4 7 User Manuat Statistics 5 ShortName Statistics Style Absolut
31. Simulation Worst Case Optimization What If Plot Tools Case Studies Tutorials Can Bus Tutorial CAN landmark 0 sim Gateways Tutorial CAN landmark 1 sim z Tutorial CAN landmark 2 sim N Tutorial CAN landmark 3 sim Tutorial CAN landmark 4 sim Event triggered transmissions can occur with mixed frames i e frames with a periodic transmission pattern but for which additional instances can be transmitted between two periodic transmissions such as AUTOSAR mixed transmission mode or pure event triggered frames They have respectively the types P E and E In our example we have both periodic and mixed frames as can be seen by looking at the TxMode column 2009 2014 RTaw 34 182 RTaW Sim V1 4 7 User Manuat RTaW Sim v1 4 2 Pro Model not File Samples Simulation Worst Case Optimization What If Plot Tools Architecture amp Architecture CAN LS Z A Architectures z Architecture Buses BusSimConfig A CAN LS Stuffing 10 Default DOAS RandomComOffset0 No drift CAN LS Ecus Ecu_0 Sample Time aos Canld Cant Paylo Role cu Ecu 3 LOL da 20A gt bytes P periodic transmission Ecu 4 59 0x3b 20A frame95 8 bytes E event driven transmission Ecu 5 73 0x49 2 0A 8 bytes P E periodic and event driven transmission Ecu_6 119 0x77 20A 8 bytes Ecu 7 122 0x7a 2 0A 8 bytes Ecu_8 135 0x87 2 0A 1 bytes Ecu 9 196 Oxc4 2 0A 8
32. Tx Errors BurstyTransmissionErrorModel To finish change the default Statistic and Sample Time as before and click Create New Curve M ieee ee es ShortName WNon bursty errors DisplayColor red ka BusSimulation Mone Bus CAN LS Stuffing 10 Offsets DOAS Com Offset Rz Statistic Maximum Scope And Kind of Times Bus Response Ti Sample Time T Then click on the Graphic tab on the left bottom 2009 2014 RTaW 46 182 RTaW Sim V1 4 7 User Manuat t Graphic oata to see the result Effect of errors on maxima 9 5 ms 9 ms 8 5 ms 8 ms Fa MS 7 7 ms 6 9 Ms ma i E Om Ln iA F 5 ms se Times in u 4 5 ms 4ms v 2 2 ms Respon 3 ms 2 0 M5 2 m5 1 9 ms 1 ms 0 5 Ms 36 135 209 274 314 348 400 487 324 508 590 627 709 756 786 870 915 942 O85 1211 1445 Frames in increasing order of identifiers No Errors Non bursty errorsNon bursty errors N As expected the response times are longer with transmission errors For some frames the difference reaches up to 5ms which corresponds to approximately 5 frame length at 125kbit s Notice that when a frame is corrupted by a transmission error it is sent again as soon as possible but since higher priority frames might have become ready since the start of the initial transmission the additional delay induced by the error can be much longer than just the time b
33. and thus on average there is a transmission error burst every 273 milliseconds The BurstLength field shows the chosen distribution for the length of a burst in number of consecutive faulty frames The fields FER Frame error rate and BER Bit error rate are computed by the tool from the distributions To make the details of the chosen distribution visible right click on the field Burstlength MoBurst b Show details Remove Set 2009 2014 RTaw 42 182 RTaW Sim V1 4 7 User Manuat and chose the Show details entry of the pop up menu and then click on the Data tab of the appearing pane HonBurstyTransmissionErrorhodel WoBurst p4 ShortName MoBurst HistoLawEntries pain width Probability 1 1 1 0 Here the distribution is actually non random always exactly 1 transmission error per burst When a transmission occurs the simulator chooses randomly uniform distribution one of the bits of the frame as detection time of the error The ErrorRecoverTimelnBits field shows the chosen distribution of the delay in bits until the error is eliminated and the bus is back to normal functioning When no distribution is specified then the Simulator uses an uniform distribution over the set of possible values 14 15 27 28 Let us now run a Simulation a Simulate File Length amp Sample Times Simulate the traffic on bus CAN LS V Buses CAN LS Bit Stuffing 10 G
34. are forwarded from bus _ 1 to bus 2 are those for which there are points on all three curves As can be seen the end to end response times blue curve are only a little higher than the bus 1 local response times and thus the maximal end to end response times are generally not the sum of the local maximal response times and the gateway delay This means that the maxima of each bus are unlikely to occur at the same time 4 Reference Manual This section provides a detailed description of all functionalities offered by RlaW Sim e Section 4 1 gives an overview of the GUI e Section 4 2 describes all available menus e Section 4 3 tells how data editing works in general e Section 4 4 provides a detailed description of specific system entities e Section 4 5 is about opening and import system descriptions e Section 4 6 explain how to perform a simulation 2009 2014 RTaW 58 182 RTaW Sim V1 4 7 User Manuat e Section 4 7 describes how to visualize simulation results e Section 4 8 explains how date can be exported 4 1 Overview of the GUI Besides the classical menu bar at the top see the figure below the user interface is subdivided into two parts On the left there are expand bars that show the different parts of the data under the form of relation trees and on the left in a tabbed folder are displayed the various panes that display details about the different data entities RTaW Sim v1 4 2 Pro Model not y
35. be below 300ppm 2009 2014 RTaWw 50 182 RTaW Sim V1 4 7 User Manuat Length amp Sample Times Simulate the traffic on bus bus_1 Buses Bit Stuffing bus 2 Gateways Bus Interfaces Configuration bus 1 gt bus 2 Com Offsets Generator Seed 30092011 Drift Mode CONSTANTDORIFT Drift Bound ppm Drift Generator Seed Frame Error Model Communication Pattern Models t Pause Resume Stop Do the same for bus 2 If you click on bus 1 gt bus 2 gateway node you will see that the unique FrameDelayDistribution is already chosen K File Length amp Sample Times Simulate the traffic on bus bus_2 gt bus_1 a Buses bus 1 Frame delay distribution Uniform 100 200 bus 2 4 Gateways bus_2 gt bus 1 bus 1 gt bus 2 fp ll Pause Resume Stop 2009 2014 RTaW 51 182 RTaW Sim V1 4 7 User Manuat Now we can start the simulation which will take longer than in the other tutorials because not only we have chosen a longer simulation time but also because the frames have shorter periods ex 10ms instead of 100ms and thus there are much more events to be handled per second As can be seen from the simulation report on longer simulations the number of events treated per second is even larger than in previous simulations pS Simulation Report l Dh Simulated 1d in 20m 35s 822ms es P Simulation speed 1 2 mega events second Speedu
36. be provided because it is not contained in the DBC file Second there are no standard attribute names for frame periods and frame transmission offsets The dialog proposes by default the custom attribute names that are used by the Vector Tool Chain You can edit the fields and change the names according to your needs 4 2 1 1 6 FIBEX file FIBEX Field Bus Exchange Format is an XML format used to describe message oriented communications systems It is an ASAM standard that now supports FlexRay CAN MOST and LIN networks and that became the de facto standard for FlexRay tool support RTaW can provide a converter allowing to import FIBEX files into RTaW Sim OEM specific extensions can be supported Contact the Support if you need this functionality 4 2 1 1 7 Autosar file Autosar iS an ongoing automotive industry initiative aimed at Standardizing electronic architectures and their development process The system template XML description contains the description of the communication systems RTaW can provide a converter allowing to import AUTOSAR XML files into RTaW Sim Contact the support if you need this functionality 2009 2014 RTaw 67 182 RTaW Sim V1 4 7 User Manuat 4 2 1 2 Recent Files The Recent Files menu entry provides a quick list of recently opened files Open Import Recent files CARTaW Sim Samples Tutorial Gateway landmark 0 sim New CARTaW Sim Samples Tutorial C4N landmark 1
37. ends either no more frames are waiting and the bus switches back into the idle state or a new transmission end is scheduled as described above Duration The following actions take zero simulation time 2009 2014 RTaw 176 182 RTaW Sim V1 4 7 User Manuat e determining the frame instance that wins the arbitration i e the waiting instance with the highest priority and the Subsequent scheduling of the transmission end e scheduling of the inter frame period end 5 1 4 Time drift of local clocks Each ECU has a local clock that is used to periodically instantiate frames The clocks may drift against each other which means that a clock may run faster or slower than others and thus the actual instantiation period of frames with same nominal period may differ between ECUs The supported clock drifts are the following DriftMode Name Meaning NODRIFT Clock frequency at nominal speed CONSTANTDRIFT Clock frequency at a fixed value typically below or above nominal Speed A second effect is the progressive modification of the initial time offsets between the clocks and thus the inter ECU frame offsets 5 1 5 Transmission delay statistics The transmission delay of a frame instance spans from the moment when it is instantiated until the moment when its transmission over the bus ends In particular if a frame is instantiated when the bus is idle and no other frame is waiting then the transmission delay is equal to the time need
38. laws The Results and Graphics tree covers the simulation results related data entities such as response time statistics and graphics 4 2 Menus In this section are described all menus and their entries available from the menu at the top of the main window of the tool RTaW Sim v14 2 Pro Model not yet saved File Samples Simulation Worst Case Optimization What If Plot Tools A ey rh ae ren wh 4 2 1 File The entries of the File menu are described in the following sections Samples Simulation Open Import Recent files New Merge Import d Save Save As 2009 2014 RTaw 60 182 RTaW Sim V1 4 7 User Manuat 4 2 1 1 Open Import In this section are described which kind of files the tool is able to open or to import When the menu entry Open Import is chosen a file selection dialog pops up If you want to import data from files with a different format than the default format of the tool then you need to select the appropriate file name filter as shown below for the case of csv files Choose file to open Regarder dans d import i Ee E 4 CsvFormatSampleFile csv eine Mes documents r cents Bureau B Mes documents cj Poste de travail te Mom du fichier Favoris r seau Fichiers de Lye ATaw Cey file Annuler AT a Sim sml File ME TLAA Aral zer sml file AT aw Cev file Vector Dbe file h The following
39. occurs at the point in time where the instance has been released or a buffer has been freed Rational The zero simulation time assumption is made to simplify the implementation of the simulation and because otherwise many different elementary timings would be needed and more detailed information about the exact behavior of the CAN controller software and hardware Also it would be difficulty to know to which degree the resulting impact on the transmission delays is representative With the zero delay assumption the following impacts on transmission delays can be simulated with sufficient degree of representativity 1 Outer priority inversion as defined in 1 may be simulated by using a non null Queue2BufferDelay and too few transmission buffers Then outer inversions occur when a controller has frames to send but is not able to present any frame to arbitration because the copying of a frame to the transmission buffers is not yet completed This situation may occur when only one transmission buffer is used or when transmission cancellation is used with less than 3 hardware 2009 2014 RTaw 175 182 RTaW Sim V1 4 7 User Manuat buffers because then refilling a buffer or replacing the contents of a buffer takes time 2 Inner priority inversions as defined in 1 may be simulated by using a FIFO software queue as a result a higher priority frame located in the queue may be blocked by a lower priority frames lo
40. of Consecutive Frames that may be sent after the reception of a Control Frame The value 0 means 256 Payload The number of bytes to be transmitted Pause Pause between two transmission sessions i e the delay between the transmission confirmation of the last CF frame and the sent request of the next FF The granularity of time is 1 microsecond FclxDelay Delay between the reception of a First Frame or the last Consecutive Frame of a block and the instantiation of the expected Flow Control Frame The granularity of time is 1 microsecond cflxDelayAfterF Delay between the reception of a Flow Control CRX Frame and the instantiation of the first Consecutive Frame of the next block The granularity of time is 1 microsecond The following diagram illustrates the response delays FclxDelay and cflxDelayAfterFcRx of the ECUs which specify delays that are not covered by ISO 15765 2 standard but needed for the simulation 2009 2014 RTaw 117 182 RTaW Sim V1 4 7 User Manuat Sender Receiver First Frame FClxDelay Flow Control Frame CFIxAfter CRX Consecutive Frame STmin Consecutive Frame eet ateteletatateletatetetetatetetetatetetetetebehahekabatata al Consecutive Frame STmin FClxDelay Flow Control Frame CFixAfterF CRx Consecutive Frame How to integrate a segmented transmission into a simulation is explained in Section 4 6 3 How to visualize the response time Statistics and th
41. of a frame from some other bus or is itself retransmitted on some other bus then also all preceding or Successive frames are duplicated In other words the whole chain of frames is duplicated On all other buses duplication of a proportional part of all the frames that are neither retransmissions nor themselves retransmitted on some other bus Notice if a system consists of two buses with for example the respective loads 40 and 30 and if the target load on the first bus is 60 1 5 40 then resulting load on the second bus will be 45 1 5 30 4 2 6 2 Change identifier types to CAN2 0B This function allows to change all CAN2 0A identifiers of a bus or of all buses of an architecture into CAN2 0B identifiers This is useful for transforming CAN2 0A_ configuration files into CAN2 0B configuration files studying the effects of using long identifiers on the global network load When you select this menu entry then the following dialog is displayed Migrate to CAN2 0B Architecture Architecture Bus lca ki Ves Cancel 2 EE a E First you have to select an architecture If you do not select a bus then the change is applied to all buses of the architecture otherwise only to the selected bus Notice The extended part of the migrated identifiers is set to 0 2009 2014 RTaw 81 182 RTaW Sim V1 4 7 User Manuat 4 2 6 3 Change frame types to CAN FD This function al
42. sim Merge Import CARTaW Sim Samples CAN HS WFCS 2010 sim Save aa Notice that the currently opened file does not appear in the list 4 2 1 3 New Allows creating a new empty and unsaved model If at the time of invocation the currently opened model has unsaved changes then the tool asks first if these should be saved 4 2 1 4 Merge Import In the following sub sections we describe the different possibilities to merge data from other model files into the currently opened model Merge Import From Tracelnspector Save Selective Save As Sample Tx Error Models Quit Entire file 4 2 1 4 1 Entire File This functionality allows to import all data entities from a specified file into the current model The imported entities are put side by Side with the existing ones 4 2 1 4 2 Tx Error Models This functionality allows to import into the current model some typical transmission error models together with the underlying probability laws The imported entities are then shown in the Specific Models section 2009 2014 RTaw 68 182 RTaW Sim V1 4 7 User Manuat lt lt Architecture Specific Models 4 ProbabilityLaws Error Bursts Exp lambda 1 100 Exp lambda 1 1000 Exp lambda 1 150 Exp lambda 1 1500 Single Errors EventOccurrenceModels 4 7TxErrorModels Error Burst BER 1 10 4 Error Burst BER 2 10 5 Single Error BER 1 10 4 Single Error BER 1 10 5 lt lt Results an
43. stored anywhere on the file system The configuration file may also be edited by hand The following table describes all simulation configuration properties All references to model elements such as the bus the frame offset configuration etc must be be part of the referenced model file Simulation configuration properties ModelFile The path to the RTaW Sim model file The path may be absolute or relative to the configuration file Under Windows the backslash path separator must be escaped but the normal slash can also be used C RTaW Sim Tutorial landmark 2 xml StatTimes Comma separated list of times where Snapshots of the frame response times shall be made The format of the times must include units see Table 2 Duration Length of the time interval to simulate expressed with time units from Table 2 Architecture The architecture to which the buses and the gateway belong BusxX The name of bus to be simulated BusxX OffsetConfig The name of the frame offset configuration to be used for the simulation Busx ComOffsetConfig The name of an offset configuration for the COM task If no such configuration is 2009 2014 RTaWw 136 182 RTaW Sim V1 4 7 User Manuat Simulation configuration properties Busx ComOffsetGenera tionSeed BusX ClockDriftConfigu ration BusxX ClockDriftMode BusxX ClockDriftBound BusX ClockDriftGenerat lionSeed BusX FrameErrorModel BusX EventlxModel BusxX ComPatterns
44. that frame periods and offsets are specified in milliseconds ms This information is also provided by the tool tips of the table column headers and the labels of the parameter fields 4 6 3 Configuring and running a bus simulation A simulation is configured and its execution is controlled through the Simulation configuration dialog which can be opened by selecting the Architecture entry in the Simulation menu Worst Case Optimization What If Plot Tools Simulation Architecture Initialize from existing Second Run Simulation The Length amp Sample Times dialog allows to configure the simulation times E Simulate File Length amp Sample Times Length of Simulation a Buses Gateways Intermediate Statistics Include bus off blocked W Generate Trace Start Pause Resume To specify the other characteristics of the simulation click on the bus name in the left tree 2009 2014 RTaw 128 182 RTaW Sim V1 4 7 User Manuat E Simulate File ie a alas Simulate the traffic on bus CAN LS C 4 Buses CAN LS Gateways gt Bit Stuffing 10 Bus Interfaces Configuration Default Offset Configuration Inter ECU COM Offset Configuration Com Offsets Generator Seed 30092011 Clock Drift Configuration j Drift Mode CONSTANTDRIFT x Drift Bound ppm 200 Drift Generator Seed 1102007 Event Occurrences Model Frame Error Model Communication Pattern
45. the frame is always the same of course this is not the typical behavior and it is due to the fact that station clock drifts variable bit stuffing for the frames and transmission jitters are not modeled in this example Before trying to understand this phenomenon let us draw a graphic that shows how statistics evolve with increasing sample size For this purpose choose the corresponding entry from the Plot menu for the Maximum statistics Tools Create graphic with all statistics gt Offsets DOAS Com Create graphic showing convergence for a statistic Minimum Worst Case Scenario Average Observed Unfavorable Scenario Q 1 10 2 Of1 10 3 Q 1 10 4 Of1 10 5 Q 1 10 6 Maximum ty 2009 2014 RTaW 24 182 RTaW Sim V1 4 7 User Manuat This will automatically select the simulation we have performed before since it is the only one and create the following graphic A CAN LS Stuffing 10 Default DOAS RandomComOffset0 NoDrift convergence of statistic Maximum 4 ms j 3 75 ms 3 0 ms 3 20 M5 7 Response Times in ms fa wn 35 122 208 271 308 334 458 483 517 560 586 618 697 743 782 870 907 940 972 1045 1398 Frames in increasing order of identifiers Maximum 5s Maximum im Maximum 15m The resulting graphic seems to show only one curve because they are all identical As in the previous graphic the x axis re
46. the new frame instance is the following 1 if there is a free transmit buffer then the frame is stored in such a buffer 2 if there is no free transmit buffer then 2 1 if the new frame has a lower priority than all frames already in the transmit buffers then the frame is stored in the software queue according to its queuing policy 2 2 if the new frame has a higher priority than at least one frame in the transmit buffers except the frame being sent if this is the case two cases arise 2 2 1 With cancellation the lowest priority frame in the transmit buffers except the frame being sent if this is 2009 2014 RTaw 174 182 RTaW Sim V1 4 7 User Manuat the case is replaced with the new frame the replaced instance is put back to the software queue 2 2 2 Without cancellation the new instance is stored in the software queue according to its queuing policy When a frame has been successfully transmitted and thus a transmission buffer becomes free again the frame located at the head of the software queue is copied to the freed transmission buffer as soon as the buffer becomes free Constraints Transmit buffers with cancellation is only allowed with HPF software queue Duration The creation and storing of a new frame instance in the software queue or the hardware buffers the cancellation of transmission requests and the time needed to copy a frame from the queue into a freed buffer takes zero simulation time and
47. transmission start until the beginning of a following burst of higher priority periodic frame instances e if the frame is mixed then an event triggered transmission may occur during a burst of higher priority frames which was previously prevented by the transmission offsets for the time triggered frame instances 2009 2014 RTaWw 40 182 RTaW Sim V1 4 7 User Manuat Performing a similar simulation based on the FrequentEvents model and adding a similar graph to the previous plot gives the following graph Effect of event triggered transmissions on maxima 1 ms 16 ms 7 15 ms 14 ms 10 ms 9 ms B ms 7 ms Response Times in ms 6 ms 5 ms 4 ms 3 ms 2 ms i 1 ms Oms l l 36 135 209 274 314 348 460 4867 324 568 590 627 709 756 786 876 915 942 983 1211 1445 Frames in increasing order of identifiers Ho Events Infrequent events Frequent 3 5 Analyzing the effects of transmission errors After having investigated the effects of clock drifts let us study the effects of transmission errors If you have performed the previous part of the tutorial just continue otherwise you can open the Sample file that corresponds to the beginning of the third tutorial Samples Simulation Worst Case Optimization What If Plot Tools Case Studies Tutorials Can Bus Tutorial CAN landmark 0 sim Architecture Gateways Tutorial CA
48. 0 0 105 0 800 1 087 3 3 742 0x2e6 frame eee 1000 0 825 0 944 1 036 18 743 Ox2e7 frame56 Pora et a 500 0O 240 0 728 0 950 2 8 756 Ox2t4 frames P E Ecu 2 8 1000 50 580 0 944 1 067 2 2 T58 Ox2f6 framme25 p Ecu 2 5 1000 0 180 0 728 0 836 2 0 769 001 frarmed4s p Ecu_1 a 200 0 120 1 416 1571 3 3 TTT y 0309 frame 4 P E Ecu 9 2 500 50 255 0 520 0 757 3 3 TAJ A Owie frames P F Feu 15 1 1000 40 ASS 0 AAR Noss 41 7 33 182 RTaW Sim V1 4 7 User Manuat framel4 549 225 Ei frames 329 001 49 frame6l 560 0x230 frames 069 0x79 frames36 531 0x213 framed 235 0xeb 23h 59m 45s 979ms fms Sms 4ms The waiting period transparent dotted rectangle and the transmission period plain rectangle of the frame37 are underlined in orange It can be seen that frame36 and frame42 which are sent by the same ECU1 may be instantiated at the same time than frame37 Furthermore the clock drifts lead to a situation where 4 other ECUs distinguished by colors instantiate higher priority frames before frame3 is able to win the arbitration 3 4 Analyzing the effects of event triggered transmissions After having investigated the effects of clock drifts let us study the effects of even triggered transmissions If you have performed the previous part of the tutorial just continue otherwise you can open the sample file that corresponds to the beginning of the third tutorial Samples
49. 3 8 579 586 590 596 601 615 618 627 669 682 692 697 709 721 739 742 743 750 758 769 7 7 Frames in increasing order of identifiers Maximum 5s Maximum ih Maximum 6h Maximum 1d Now the graphs can more easily be compared one can clearly see that a longer sample implies a higher response time maximum This phenomenon is the result of the clock drifts that make the inter ECU frame offsets vary over time and produce other scenarios than those in the periodic behavior encountered in the first simulation The longer the simulation the more different trajectories of the system are simulated and the higher the maximal response times of the frames Readers can learn more about the effects of clock drifts on response time distributions in 5 To get back to the initial zoom select use the auto range functionality from the context menu which can be brought up through a right click 2009 2014 RTaW 32 182 RTaW Sim V1 4 7 User Manuat Properties T Save 45 Print zoom In ie zoom uk Auto Range Domain Axis Range Axis On the previous graphs one can see for example that the maximal response time of the frame with decimal id 669 is somewhat higher than that of the frames with similar priority Let us have a look at the scenario that has lead to this high response time For this purpose turn back to the details pane of bus CAN LS and locate the line of frame37 Make sure that the
50. 30 31 2006 3 N Navet Y Q Song F Simonot Worst Case Deadline Failure Probability in Real Time Applications Distributed over CAN Controller Area Network Journal of Systems Architecture Elsevier Science vol 46 n 7 2000 Available at url http www realtimeatwork com page id 5 4 N Navet A Monot J Migge Frame latency evaluation when Simulation and analysis alone are not enough 8th IEEE International Workshop on Factory Communication Systems WFCS2010 Industry 2009 2014 RTaw 181 182 RTaW Sim V1 4 7 User Manuat Day May 19 2010 Available at url http www realtimeatwork com page id 5 5 A Monot N Navet B Bavoux PSA Impact of clock drifts on CAN frame response time distributions Proc of the 16th IEEE International Conference on Emerging Technologies and Factory Automation ETFA 2011 Industry Practice track Toulouse September 2011 Available at url http www realtimeatwork com page id 5 6 N Navet S Louvart J Villanueva S Campoy Martinez J Migge Timing Verification of Automotive Communication Architectures using Quantile Estimation Proc Embedded Real Time Software and Systems ERTS 2014 Toulouse France February 5 7 2014 Available at url http www realtimeatwork com page id 5 7 N Navet H Perrault CAN in Automotive Applications a Look Forward 13th International CAN Conference Hambach Germany March 5 6 2012 Available a
51. 348 460 487 524 568 590 627 709 756 786 876 915 O42 983 12111445 Frames in increasing order of identifiers Average Maximum Data Graphic 4 7 4 1 Generation of frame response time graphics The Plot menu allows to create a number of predefined graphics Tools Create graphic with all statistics l Create graphic showing conwergence for a statistic Minimum Worst Case Scenario Average Observed Unfavorable Scenario Q 1 10 2 i esa Qft 10 3 O 1 10 4 1 10 5 Q 1 10 6 Maximum The following tables describe the predefined graphics in more details Create graphic with all statistics 2009 2014 RTaWw 144 182 RTaW Sim V1 4 7 User Manuat Description Creates a graphic with a graph for each of the predefined statistics for a certain sample time length Allows to see how the different statistics compare to each other for a specific sample time Parameters e ResponseTimeStatistic specifies the common Sample time of the statistics Create graphic showing convergence of a statistic Description Creates for a certain statistic graphic with a graph for each of the available sample time lengths It allows to see how the statistic evolves with increasing sample time for the different frames Parameters e BusSimAnalysis specifies the bus and the used configuration parameters e Statistic spe
52. 4 7 User Manuat Maximum of response times after id End to End Response Times Period in 0 36 135 209 274 314 348 460 487 324 568 590 62 7 709 756 786 876 915 O42 983 12111445 Frames in increasing order of identifiers Average Maximum Creating graphics by copying The Data tab of Graphics panels allows to create new graphics and curves based on copies of existing ones In order to create a copy of an existing graphic right click on the corresponding node and select Create copy This will bring up ae Results and Graphics x a Simulations gt A System Simulation SecondRunsimulations Evaluations a Graphics Statistics the dialog for providing the name of the duplicata change the name to Relative Loads 2009 2014 RTaw 157 182 RTaW Sim V1 4 7 User Manuat Name of copy Relative Loadd Then click Yes Statistics _ Relative Loads 5 Relative Loads 15 ms 14 ms jf 13 ms 12 ms 1 11 ms 7 10 ms 4 9 ms 5 ms 7 ms 7 6 ms 7 5 ms Response Times in ms 4 ms 3 ms jf 2 ms 1 ms 0 ms E T T T T T T F T T F T T 4 74 144 240 323 392 465 962 623 faz 836 910 986 Frames in increasing order of identifiers Data Graphic In order to change some characteristics click on the Data tab In the following example we change the style of the graphic to Relative to Pe
53. 4 BusStatistics gt 12h InterErrorPassiveTimes LengthOfStayInBusOff 4id gt FrameRespTimes InterErrorWarning Times LengthOfStaylnErrorPassive InterBusOffTimes gt FramelnterTransTimes 4 QueueStatistics Ecu_1 bus_1 Ecu_10 bus_1 Ecu_11 bus_1 Ecu_2 bus_1 Ecu_3 bus_1 Ecu_4 bus_1 Two kinds of statistics are available e Backlog maximal number of frames waiting at the same time in the software queue or the hardware transmission buffers e Residual Backlog maximal number of frames still waiting when a new batch of frames is about to be instantiated In order to visualize a statistic click into the corresponding field Backlog Backlo ErrorWarning show details i Delete nErrorPassive Create New StavInBusOff and select the Show detail entry 2009 2014 RTaWw 160 182 RTaW Sim V1 4 7 User Manuat Min 1 Average 1 13843200188444 Q 3 Q3 4 Q4 4 O05 4 Q 4 Max 5 Entries Value Count Width 2 103231683 1 3 15419564 1 4 52984 1 5 6709 1 6 5 1 The lower part shows the table view of the histogram 4 7 5 2 Error Warning Error Passive and Bus Off related statistics Under the QueueStatistics node of a BusStatistics node double click on the node that corresponds to the ECU for which you wish to see Statistics about CAN error counter states A hect y Ecu_4 Gateway bus eA Te y Buslnterface Ecu_4 Gateway bus_2 Backl
54. 6 min of periods 1000 ms 0 73 ppm Ecu_6 min of periods 1000 ms 0 87 ppm As can be seen some drifts are positive i e the clock drift factor is above 1 and thus the clock faster than nominal and some are negative i e the clock drift factor is below 1 and clock slower than nominal and this within the specified bound of 1 ppm Now a second BusSimConfig appears in the combo box in the Bus panel CAN LS 23 ShortName CAM L5 Speed 125 Periodic Load 31 1 Bit Stuffing 10 BusSimContig BusEvalConfig Mono Bus CAN LS Stuffing Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOfftset Drift No drift 10 Offsets DOA Com Offset RandomComOffset0 Drift RandombDrift 1ppm Sample Time Choosing it and a sample time makes the corresponding statistics appear in the last columns of the table And now let us plot a convergence graphic for the statistic Maximum as before 2009 2014 RTaWw 29 182 RTaW Sim V1 4 7 User Manuat Plot Tools Create graphic with all statistics graphic showing convergence for a statistic Minimum Worst Case Scenario Average Observed Unfavorable Scenario Q 1 10 2 Q 1 10 3 Q 1 10 4 Q 1 10 5 Q 1 10 6 Maximum Select the newly created BusSimAnalysis Select a BusSimulation object a Simulations Mono Bus CAN LS Stuffing 10 Offsets DOAS
55. 74 ms 1 464 ms 2 104 ms 2 228 ms 2 515 ms 2 267 ms 2 267 ms 2 178 ms 2 954 ms 2 871 ms 2 165 ms 2 695 ms 2 347 ms 3 166 ms 2 515 ms 3 111 ms 3 741 ms 4 197 ms Q5 0 883 ms 1 295 ms 1 414 ms 1 681 ms 1 843 ms 1 886 ms 1 681 ms 1 740 ms 2 486 ms 2 648 ms 2 821 ms 2 558 ms 2 789 ms 2 695 ms 3 373 ms 3 166 ms c72 2 989 ms 3 431 ms 4 370 ms 5 046 ms Q6 883 ms 1 768 ms 1 864 ms 3 076 ms 3 166 ms 3 076 m 3 593 ms 4 370 ms Max 0 883 ms 1 295 ms 1 566 ms 1 903 ms 2 004 ms 2 013 ms 1 768 ms 1 875 ms 2 617 ms 2 696 ms 3 022 ms 3 096 ms 3 183 ms 3 107 ms 3 413 ms 3 783 ms 2 847 ms 3 308 ms 2 965 ms 3 649 ms 3 255 ms 3 787 ms 4 414 ms 5 456 ms Bound m If you have performed a Worst Case analysis see Section 4 2 4 the BusEvaluationConfig combo box will contain a corresponding entry Select one such analysis and the last column entitled Bound will display the corresponding worst case response time bounds The gray shades of certain values are explained in Section 4 6 1 The Scope and Kind of Statistics combo box provides the following options Name Scope amp Kind Bus Response Times Delay between the instantiation queuing of the frame until its transmission end End to End Response If the frame is sent by a gateway on the 2009 2014 RTaw 139 182 RTaW Sim V1 4 7 User Manuat Times considered bus then there e
56. 884 ms 1 013ms 1 730 ms 39 0x27 20A Frame_27 Ecu_1 8 bytes P 50 ms 25ms 0 472ms 0 583ms 1 232 ms 40 028 20A Frame_28 Ecu_2 8 bytes P 50 ms 25ms 0 472ms 0 562ms 1 204 ms 45 0x2d 2 0A Frame_2d Ecu_3 5 bytes P 50 ms 25ms 0 364ms 0 469ms 1 130 ms 59 0x3b 20A Frame_3b Ecu_2 2 bytes P 100 ms 50 ms 0 532 ms 0 899ms 1 919 ms 65 0x41 20A Frame 41 Ecu_6 8 bytes P 50 ms 25ms 0 708ms 0 847 ms 1 720 ms 69 0x45 2 0A Frame_45 Ecu_2 8 bytes P 200 ms 100 ms 0 472 ms 0 566ms 1 290 ms 74 0x4a 20A Frame_4a Ecu_2 2 bytes P 100 ms Oms 0 260 ms 0 615ms 1 643 ms 76 0x4c 20A Frame 4c Ecu_9 3 bytes P 200 ms 100 ms 0 296 ms 0 420 ms 1 190 ms 77 0x4d 20A Frame_4d Ecu_3 Ecul 6bytes P 200 ms 175ms 0 776ms 0 922 ms 1 897 ms 82 052 20A Frame_52 Ecu_10 6 bytes P 200 ms 100 ms 0 400 ms 0 521 ms 1 328 ms 86 056 2 0A Frame_56 Ecu_2 5 bytes P 50 ms 25ms 0 848 ms 0 986ms 1 986 ms 98 0x62 2 0A Frame 62 Ecu_1 8 bytes P 50 ms 25ms 0 956ms 1 138 ms 2 280 ms 100 0x64 2 0A Frame 64 Ecu_2 3 bytes P 100 ms 50 ms 0 840 ms 1 261 ms 2 618 ms Q3 0 882 ms 1 124 ms 1 013 ms 1 533 ms 1 366 ms 1 320 ms 1 268 ms 1 232 ms 1 832 ms 1 886 ms 2 116 ms 1 740 ms 1 720 ms 1 652 ms 2 486 ms 2 320 ms 1 801 ms 2 190 ms 1 801 ms 2 529 ms 1 963 ms 2 603 ms 3 006 ms 3 412 ms Bit Stuffing 10 OffsetConfig ee a oie SOA25 Scope and Kind of Statistics End to End Response Times Q4 0 883 ms 1 275 ms 1 143 ms 1 633 ms 1 605 ms 1 587 ms 1 3
57. ARN a Ecus ECU ECL a GateWay BusConnections am Then bate Ways gt CormPatterns Ke New The source and the target bus must be specified The name is optional if not provided it will be of the form source bus name gt target bus name New GateWay r Source Target Name Create Cancel In the details pane of the gateway you need to be define right click in the field at least one e FrameDelayDistributions probability distributions in us for the delay between the reception of the frame on the receiver side and the instantiation of the frame on the sender side e FrameMappings specify which frame is sent on the sender side as reaction on the reception of a frame on the receiver side 2009 2014 RTaw 106 182 RTaW Sim V1 4 7 User Manuat Architecture CAN2 gt CANI 23 Buses Name CAND Source GateWay CAN Target GateWay CANL CAN Dii FrameDelayDistributions ECUI BusConnections CANL GateWays ECU BusConnections CAN FrameMappings Source Target GateWays GateWay a Adapt column widths to contents CANL Copy to Clipboard in CSV Format CAN GateWays CANI gt CAN2 CAN2 gt CANT ComPatterns Show selected FrameMapping Delete selected FrameMapping While you are defining the frame mapping it might happen that you can not chose as target of the mapping a frame which is sent by the gateway select
58. B 10 Frame lde Eco S Ecu a FP 20 10 Frame_ ie Ecu 9 Ecu 4 a P 20 10 Frame 224 Ecu 14 Ecu 4 S P 20 10 4 2 9 Help The user manual is available in two formats e HTML accessible through the help menu gt Help Html e PDF accessible through the help menu Help Pdf ue RTaW Sim v1 4 2 Pro Model not yet saved m File Samples Simulation Worst Case Optimization What If Plot Tools Architecture Help Html Architectures Help Pdf N l Architecture Help WCRT Pdf About Help WCRT is the help for the Worst Case Analysis Plug in available in the Professional edition of RTaW Sim 2009 2014 RTaW 85 182 RTaW Sim V1 4 7 User Manuat Furthermore tool tips on property labels often provide a short definition To make them visible simply position the mouse pointer over the label Name Source GateWay CANL Fra dass celia Delay induced by the forwarding in microseconds 4 3 Data editing In this section is described how new data entities such as buses frames graphics can be created Section 4 3 1 or modified Section 4 3 2 4 3 1 Creation In this section we show how data entities can be created from scratch see Section 4 3 1 1 or by duplicating an existing one see Section 4 3 1 2 4 3 1 1 Creation from scratch The creation of a new data entity can be initi
59. BusSimulationConfig with drift is selected and then right click on the line in order to choose Show Unfavorable Scenario Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOftset Drift RandomDrift 1ppm c CAN LS a ShorntName CAN LS Speed 125 Perodic Load 31 1 Bit Stuffing 10 BussimConfig Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset Drift RandomDrift 1ppm As a result the following Gantt chart is displayed 2009 2014 RTaw BusEvalConfig v Sample Time 1d Scope and Kind of Statistics Bus Response Times ki Canld ShortName Type Sender Recervers Paylo Period Mini Offsets Min Average Quanti 586 Ox24a framet6 p Ecu_1 1000 g 10 0 344 1 007 1 8 590 Ox2de frame p Ecu_3 100 0 90 0 872 1 037 aT 596 0x254 frame89 P E Ecu_11 100 50 35 0 944 1 086 2 2 601 0x259 frame5 p Ecu 3 500 g 735 0 592 0 821 3 1 615 Ox267 frame51 P Ecu 1 500 0 130 0 872 1 001 1 8 618 Ox2ba frame l P Ecu 1 500 0 a0 1 624 1 883 3 5 627 0x273 frame58 P Ecu_6 1000 0 175 0 944 1 067 25 665 03d famel Eran De EE framel6 Show Histogram 500 0 145 0 448 0 584 1 6 frames2 Show Unfavorable Scenario 1000 50 320 0 944 1 083 1 93 framel Adapt column widths to contutis 500 0 480 0 944 1 024 21 framel0 Edit selected Frame 1000 0 380 1 696 2 008 41 3 frame EE E A EE 200 50 95 0 872 1 037 2 3 739 Ox2e3 framel0l 20
60. CARBENCH allows to generate realistic sets of messages that can be used for comparing configuration algorithms or making design choices it is especially useful early in the design cycle when the real message sets are not yet available for the embedded system architecture design or for assessing the maximum load that can be achieved on a certain bus Another interest of NETCARBENCH is that the message sets it generates can be communicated overcoming the confidentiality requirement one has with real sets of messages RTaW Sim can import NETCARBENCH generated files choose NETCAR Analyzer xml file format in the import windows since NETCARBENCH format is a subset of NETCAR Analyzer one In a first step you may for instance test this possibility using the example configuration files e g automotive powertrain and body networks available on the NETCARBENCH home page 4 2 1 1 4 CSV import file As its name says this kind of file contains Character Separated Values CSV but it is easier to understand as textual data under a tabular form as shown in Table 1 The current format 1 2 allows to describe and import an architecture consisting in one or several CAN buses connected by 2009 2014 RTaw 63 182 RTaW Sim V1 4 7 User Manuat frame gateways with periodic event triggered or mixed frames and one or several transmission offsets RiaVW Can Csv Format 1 4 Bus Name Speed FD Speed Frames Name Frame Al Fram
61. Data Graphic 2009 2014 RTaWw 150 182 RTaW Sim V1 4 7 User Manuat You should see the following Statistics 25 Statistics 3 20 M5 3 ms 2 7 Ms 7 2 2 M5 7 E 2 25 ms 2 MMs 1 9 ms 7 1 20 MS 1 1 ms 1f 0 5 ms jf Response Times in ms 0 5 ms 7 0 25 ms j 0 ms f T T T T T T T T T T T T T T T 12 59 106 128 179 242 282 359 432 465 580 2 700 722 750 869 Frames in increasing order of identifiers Local bus_2 Data Graphic Let us now add a second curve by copying For this purpose go back to the Data tab select the existing curve in the table and right click on the line 2009 2014 RTaWw 151 182 RTaW Sim V1 4 7 User Manuat Statistics 25 ShortName Statistics Style Absolute Value ScopeAndkind Curves ShortName Display Hidden ScopeAndkind prp hisek U a Bie Bernanee Timer Adapt column widths to contents Copy to Clipboard in CSV Format Show selected Curve Create Copy h Delete selected Curve Add New Curve The Create Copy entry allows to create a copy of the existing curve Statistics End 2 end bus 2 25 ShortName End 2 end bus_ DisplayColor green y Hidden BusSimulation A bus_2 Stuffing 10 HPF SOAS RandomComOttset RandomConstantDrifts Statistic Q 1 10 6 gt ScopeAndKind End to End Response Times ka l Bus Response Times Sample Time 1d a
62. Drift RandomDrift ppm 10 Offsets DOAS Com Offset RandomComOftset Drift RandomDrift lppm Events InFrequentEvents gt To finish change the default values of the Statistic and Sample Time field as before and click Create New Frametraph ShortName ax non bursky errors DisplayColor red F Analysis Result Cah LS f OffsetsDOA f WonBurstyTransmissionErrorModel f In Statistic Maximum Sample Time id ae 2009 2014 RTaWw 39 182 RTaW Sim V1 4 7 User Manuat Then click on the Graphic tab on the left bottom t Graphic Qata to see the result Effect of event triggered transmissions on maxima 9 5 ms 9 ms 8 9 ms amp ms 7 9 m5 7 ms 29 Ms Z ms 55 ms E 5m5 Time u 4 5 Ms 4 ms Respon 3 0 ms 3 ms 2 0 mS 2 m5 1 3 ms 1 ms 0 5 ms Oms 36 135 209 274 314 348 4600 487 524 568 590 62 7 709 756 786 76 915 942 O83 1211 1445 Frames in increasing order of identifiers No Events Infrequent eve ntg As expected the maxima of the response times increase for some frames the difference reaches 5ms which at 125 kbit s corresponds to approximately 5 additional frames that preempt the considered frame There are two kinds of causes for the high response time increase e ifthe frame is periodic then the occurrence of higher priority event triggered frame instances may delay the
63. E 60 2 ek WiC E A A A E E E A E E E 60 BZ dt le Open IM OOE eerren ENERET EErEE EEr ST 61 AA RW A US e A E E O OE TE 62 4 2 1 1 2 NETCAR Analyzer file n ssssnsnnsennnessrrrrrrrsrrrrrrssserrrrressrrrrrrreene 62 Ae ele NETCAR BENGE IIO rirnan e A EE EA 63 4 2 1 1 4 CSV import fil essssssssssssrssnnnssnrusnnnssnnnnnasseorennnennennnnunseerrennenne 63 d ELS DBC TIE oean a a E E E 66 Zadal O FIDEA IG anrr E EOE 67 Aa LT AULO ar UE EE T A EEA AE 67 4 2 1 2 Recent FileS ssssssssssssssssnnssnannsnsanasnnunannnasansnnnesnnnnansnnnnannnnnnnnnnne 68 BZ i NOW a E A E E E E E 68 4 2 1 4 Merge UMD OF Usecnerscevnsscnnentbesevecisatetscanetactadeedesbosdansecovernenteeestedas 68 ALAE ENUIS FIIO oaae E eee 68 4 2 1 4 2 TX Error ModelS sssssssssnrssssarusnnasnnnsnssnnnsnnnnunnnnnnannnnnnnnnnnannnnan 68 AA SOIC CU VG rarna a E E E AEE 69 4 2 1 4 4 From TracelNnSpectoOr sssssssssssrrrrrssrrrrrrrsrrrrnnerrrrrrrnsrrrrrrerrrnne 70 2009 2014 RTaW 1 182 RTaW Sim V1 4 7 User Manuat AZ a Mea I Cia ces began E E E 72 Aere dG I SG e E sacune gene esessamenue beeen ie ceoseaeee 72 AL QUI souapreceaiecmpontnnnes via E EEEE 72 A Dale Sarisccacrransp waren A T 72 AAS NETO U cractactnpntenrenssiintataadenacetecenequieabiadaa oes E 73 eZ oe NV OVS Ch GC octet ese AE E eres aitesauenenanersiceaaeee 73 A29 OBUMIZ AU ON ccrusaprieasiacacrewnmerrexs EE RT E ETEEN 75 4 2 5 1 Rate Monotonic CAN Identifiers sssssssssssrrrssssrrrrrrrs
64. ER and bit error rate BER estimations have been added to details pane of transmission error model see Section 4 4 5 Clarified meaning of bus interface specific configurations see Section 4 4 4 changes between Version 1 2 10 and Version 1 3 12 The main file format is changed to sim zipped xml to reduce file size The xml can still be opened and saved by selecting the corresponding file filter Simulation of segmented transmission has been added see Section 4 4 6 1 Simulation of frame dialogs has been added see Section 4 4 6 2 Flexible configuration of CAN bus interfaces see Section 4 4 4 2 GI ossary Term Meaning Frame instance Since frames are usually sent in some recurrent manner for example periodically several instances of a frame are created 2009 2014 RTaw 12 182 RTaW Sim V1 4 7 User Manuat at run time which usually differ by their contents and exist at different moments but may also exist concurrently when the transmission of a previous instance is still pending Frame instantiation time Time where an instance of a frame is ready data has been copied to the payload section and queued for participating in the CAN arbitration Once the frame has been instantiated its contents IS not modified anymore When the frame is actually participating in the arbitration and when its transmission exactly Starts depends on the other pending frame instanc
65. Ecu 3 FrameMappings Source Target Ecu_4 Frame_168 Frame_168 Ecu_5 Frame_365 Frame_365 Ecu_6 Frame_397 Frame_397 Ecu_ Frame _2de Frame _2de Ecu_8 Frame_1d1 Frame_1d1 Ecu_9 Ecu_10 Ecu_11 Ecu_12 Specific Models Results and Graphics Let us now perform a simulation For this purpose open the multi bus simulation configuration dialog Simulation Worst Case Optimization _What If_Plot_Tools Simulation A Architecture j Initialize from existing Name Second Run Simulation Source F 2009 2014 RTaW 49 182 RTaW Sim V1 4 7 User Manuat This dialog contains three sections E Simulate File Length amp Sample Times Buses Gateways K This time we will do a longer simulation to make sure that the statistics converge For this purpose left click on the Length amp Sample Times label and enter the following sample times for the statistics E Simulate File Length i Sample lames Length of Simulation Buses Gateways lm lh 22h 1d Include bus off blocked W Generate Trace Pause Resume Stop Now click on the bus 1 node and then in the right pane check the Simulate box choose the SOA25 offset CONSTANTDRIFT as Drift Mode and finally 1000 as Drift Bound We choose very high clock drifts rate in order to accelerate the convergence of the Statistics More realistic values would
66. Frame Frame 5c Entries Value Count 0 445 5648629 0 455 131000 0 465 131109 0 475 131127 0 485 131614 0 495 131104 0 505 131220 0 515 130899 0 525 131315 0 535 131046 0 545 131152 0 555 130954 0 565 124180 0 575 122843 0 585 118302 0 595 110163 0 605 98203 0 615 88805 Data Histogram Defavorable Scenario 4 7 3 Viewing the unfavorable scenario The column entitled Max in the table view see Section 4 7 1 provides the highest values that has been encountered during the sample time that was simulated It is possible to view the scenario that lead to this maximal value For this purpose right click on the line that contains the frame for which the unfavorable scenario shall be shown and choose Show Unfavorable Scenario 2009 2014 RTaw 142 182 RTaW Sim V1 4 7 User Manuat Canld ShortName Type Sender Receivers Paylo Period Mini 224 Owed Frame e0 P Ecu 5 Ecu 4 Gate B 10 0 225 Oxel Frame _el P Ecu 5 Ecu 4 Gate a 10 0 346 Ov15a Frame 15a P Ecu 9 Ecu 4 Gate 6 10 0 348 OxL5c Frame_15c P Ecu 9 o 349 015d Framel5d P Ecud Lota SL 350 0x1 5e Frame 15e p Ecu 9 Show Unfavorable Scenario 397 Ox18d Frame 18d P Ecu 3 Adapt column widths to mr 398 0xl8e Frame 18e P Ecu_3 Edit selected Frame 404 hd4 Frame 194 P Ecu_s Delete selected Frame
67. In the Data tab of a graphic the Style parameter specifies how curves are displayed Maiamum of response times after 1d 2 ShortName Maximum of response times after 1d Style Absolute Value Type of Cuna Absolute Value HIMES Relatively to Period Relatively to deadline ScopeAndl Da End to Ens Average black show End to Em Q 1 10 2 blue hide End to Em The following table explains the different options frame Value displayed This is the default behavior Absolute The actual values of the response time statistics are the period Relative to The response times are divided by the period and period displayed as percentage This allows to easily identify frames for which the response time statistics are close 4 2 8 2 Relative to The response time are divided by the deadline and deadline displayed as percentage This allows to easily identify frames for which the response time statistics are close to the deadline Notice that frames for which the deadline is not defined are ignored Notice also that deadlines can be defined individually see Section 4 4 2 or generated based on a global rule see Section The following graphic shows the response times relatively to periods which tells the designer about the load of the system taking into account latency constraints and whether it can accommodate additional frames and stations 2009 2014 RTaw 156 182 RTaW Sim V1
68. Models Pause Resume Stop The buttons Start Pause Resume and Stop at the bottom of the dialog allow to control the simulation The progress bar above the buttons provides feedback on the progress of the simulation achieved percentage of the total functioning time to simulate and the actual elapsed time The minimal set of configuration parameters for a simulation consists of e a bus a bit stuffing factor increase of frame length due to bit stuffing e a bus interfaces configuration e a frame offset configuration e a simulation length Table 2 describes the time units used in the specification of the simulation length For example ld 20m 7us means 1 day 20 minutes and 7 microseconds 2009 2014 RTaW 129 182 RTaW Sim V1 4 7 User Manuat Abbreviation Meaning US microsecond ms millisecond S second m minute h hour d day Table 2 Time units an their abbreviations The Bit Stuffing parameter allows to specify by how much the length of the frames increases due to bit stuffing Possible choices are 0 10 15 20 and worst case The worst case configuration corresponds to approximately 25 of increase The bus interfaces configuration determines the characteristics of the bus interfaces such as the queuing policy of the software queue or the behavior in case of transmission errors see Section 4 4 4 The inter ECU COM offset conf
69. N landmark 1 sim Tutorial CAN landmark 2 sim Tutorial CAN landmark 3 sim iS Tutorial CAN landmark 4 sim Let us first have a look at the proposed model for transmission errors for this purpose click on the Specific Models tab on left 2009 2014 RTaWw 41 182 RTaW Sim V1 4 7 User Manuat e Specific Models iS Y Results and Graphics lt and then double click on the NonBurstyTransmissionErrorModel entry 4 xErrorModels Bursty TransmissionErrorModel NonBurstyT i ErrorModel which allows to see the details of the error model NonBurstyTransmissionErrorModel 2 5 ShortName NonBurstyTransmissionErrorModel InterBurstCount Exp lambda 1 100 BurstLength NoBurst ErrorRecoverTimelnBits FER 1 0 BER 1 0e 04 In RTaW Sim two types of transmission errors are identified errors that corrupt a single frame and burst errors that corrupt several consecutive frames and translates into a period during which the bus is not accessible see Section 5 4 for more detail on that error model The InterBurstCount field shows the chosen distribution of the number of consecutive frames between two successive occurrences of transmission error bursts We have chosen an exponential distribution with parameter lambda 1 100 which means that on average there is a transmission error burst every hundred transmitted frames Notice that in our example there are approximately 366 transmissions per second
70. ND Show selected Frame e Remove selected Frame G aie m sid iame ha The source ECU of the segmented transmission sends the First Frame and the Consecutive Frames while it receives the Flow Control Frames The destination ECU receives the First Frame and the Consecutive Frames while it sends the Flow Control Frames In case of gateway crossing do not forget to specify the sending and receiving at the bus connections of the gateway and also to define 2009 2014 RTaw 114 182 RTaW Sim V1 4 7 User Manuat the frame gateways mappings in the two directions see Section 4 4 3 Next you need to define the corresponding communication pattern Architecture a CANI es a Buses Short _ BusSimc AR a Ecus BusEvalCc Sample gt ECU gt GateWay Canld as ComPattern Add New 57 E 100 0x64 In the creation dialog for CommunicationPatterns make sure to select the SegmentedTx type New CommunicationPattern es e S Segmented Ix ShortMame Segmented Tx StartBus CANL FirstFrarme CANIL_FF FlowControlFrame CANT FC ConsecutiveFrame CAN CF Create Cancel i A segmented transmission pattern is best defined by specifying first the bus where the original sender is located followed by the identification of the three kinds of frames Notice that e only frames with the correct inherited role may be selected e Also in case of
71. RTaw 180 182 RTaW Sim V1 4 7 User Manuat For the time being we believe that a uniform distribution over all bits of a frame is an acceptable model for the position of the faulty bit The time until resumption should be a modeled by a probability distribution over the set 14 15 27 28 We think that a uniform distribution is a good default value After the resumption time there is the inter frame time of 3 bits before the first arbitration phase after the error 5 4 3 Transmission delay statistics When a frame is corrupted by a transmission error its response time increases until it is correctly transmitted As a result the transmission error induces a longer response time This does not mean however that the response times of all frames become longer or remain equal when transmission errors are added to the Simulation even if all other simulation configuration parameters remain unchanged The reason is that a higher priority frame may benefit from the transmission error of a lower priority frame that would normally be transmitted just before the higher priority frame and whose transmission is interrupted by the transmission error 6 References 1 Specification of CAN Interface AUTOSAR v4 0 2009 2 M Grenier J Goossens N Navet Near Optimal Fixed Priority Preemptive Scheduling of Offset Free Systems Proc of the 14th International Conference on Network and Systems RINS 2006 Poitiers France May
72. Stuffing 10 2009 2014 RTaw 138 182 RTaW Sim V1 4 7 User Manuat In order to see the frame statistics one must alSo choose a sample time Sample Time 1h Cand 224 Oxel ir F Th 4 As a result the statistics columns get filled for the specified sample time ShortName bus_1 Bit Rate 250 kbits s CAN FD Bit Rate Periodic Load 51 0 BusSimConfig A bus_1 Stuffing 10 HPF SOA25 RandomComOffset0 RandomConstantDrifts 1000ppm BusEvalConfig Sample Time 1d X Canld Can ShortNa Sender Recei Paylo Tx Period TxOff Min Average Q2 5 065 2 0A Frame_5 Ecu_7 6 bytes P 50 ms 25ms 0 400 ms 0 501 ms 0 867 ms 10 Oxa 2 0A Frame_a Ecu_10 6 bytes P 100 ms 50 ms 0 400 ms 0 498 ms 0 872 ms 11 Oxb 2 0A Frame_b Ecu_2 2 bytes P 100 ms 50ms 0 260ms 0 338ms 0 730 ms 13 Oxd 2 0A Frame_d Ecu_l0 Ecu_l 4bytes P 100 ms 50ms 0 744ms 0 847 ms 1 225 ms 19 0413 20A Frame_13 Ecu_9 6 bytes P 100 ms 50 ms 0 400 ms 0 504 ms 0 882 ms 21 045 20A Frame_15 Ecu_3 5 bytes P 100 ms 50ms 0 364ms 0 459ms 0 848 ms 23 047 20A Frame_17 Ecu_10 8 bytes P 50 ms 25ms 0 472ms 0 571ms 0 951 ms 28 0dc 20A Frame ic Ecu_6 Ecu_1 1bytes P 50 ms 25ms 0 224ms 0 331ms 0 833 ms 29 0xld 2 0A Frame ld Ecu_2 8 bytes P 200 ms 150 ms 0 744ms 0 838ms 1 456 ms 33 021 20A Frame_21 Ecu_3 Ecu_1 4bytes P 100 ms 50 ms 0 708 ms 0 822 ms 1 472 ms 37 0x25 20A Frame_25 Ecu_9 8 bytes P 100 ms 50 ms 0
73. W Sim V1 4 7 User Manuat RTaW Sim v1 47 Pro C Users jomn syn pegate netcar syn NetcarSimi docs help mw LESTE Pro C Users jorm svn pegase net carSim docs he File Samples Simulation Worst Case Optimization What If Plot Tools E A Response times for dialog Dialog after 1d ii Architecture Architectures Architecture Buses CANI CANA Ecus ECU1 ECU ECUS GateWay ComPatterns CANI Dialog OccurrenceModels Response times for dialog CAN1 Dialog 0 300 7 0 275 0 250 0 225 4 0 200 0 175 0 150 7 Probabilities 0 125 4 0 100 4 Dialog TimeDatas A Response times for di CANL SeqmentedTx 0 075 0 050 4 0 025 4 0 000 r 20 2959 30 395 40 45 30 309 60 65 74 a Response times in ms Specific Models Mir Mar Results and Graphics Histogram Observed Unfavorable Scenario 4 7 5 6 Second run statistics If you have performed a second run simulation see Section 4 6 4 then the corresponding node appears under the SecondRunSimulations node in the Results and Graphics section D A1 Second Statistics for System Simulation 32 Architecture aes Mnemonic Al ReferenceSimulation A System Simulation ShortName Second Statistics for System Simt Results and Gr
74. anpinaneeccervnernouernrtre von eeanrindcrecusirs ane onovennaien 173 Da INO MIG CAN or E 173 5 1 1 Instantiation Of periodic FraAMeS ccccccecececeeeeeeeeeeeeeeeeeeeeeeaees 173 5 1 2 Downward traversal Of COM sStaCkK sssssssssssrrssrrrerrrrsrrrnreserreo 174 5 1 3 Arbitration ANd tranSMISSION cccccecceceeeeeeeeeeeeeaeeeeeeaeeseaeeaneaes 176 5 1 4 Time drift of local ClOCKS c cc ccecccceceseeeeeeeeeeeeeeeeeeeesneatseeaeeneaes 177 5 1 5 Transmission delay StatiStiCS ccccccesesceceeeseeeeaeeeeaeeeeatsneseaeees 177 JA GLOWA VO sprin TEIN EE AAA EEE OAP PPE ESEE 177 Te AMO GATE WOY eer aA EE EEO 178 5 2 2 Transmission delay StatiStiCS ccccceceseeeeceseseeeeeeeeeeseeaeseeaeeneas 178 5 3 Event triggered transmissions Of fraMeS ssssesssrrrssrrrrrrrsserer 178 5 3 1 OVOIVICW cccccccececeececeeeeeeeeseeaeateeeaeseeaeseeeeateneatentaeanentaeentaneneataneneas 178 5 3 2 Simulation of event triggered tranSMiISSIONS cccceceeeeeeeee 179 2009 2014 RTaW 3 182 RTaW Sim V1 4 7 User Manuat 5 3 3 Transmission delay StatiStiCS ccccceccseeeeceeeeeeeeeeeeeeeseeeeseeaeanees 179 5 4 Transmission CV OMS vamrnceeiweniietaectdwhdenrinensunie bide ccoundoneenedesdenanneed 179 Deas OO SWE WY sa stemaceneaannomnacntsersutssannciienovecsmnnenseueonsponieatineitmontestesctdinaes 179 5 4 2 Simulation Of transmission ELLOS ccccecceceeeceeeeeeeeeeeeeueeeeueeeanenes 180 5 4 3 Transmission delay Stat
75. ansmission buffer meanwhile Furthermore the other ECUsS may then have frames with a higher priority to send and thus it may take some time until the faulty frame will actually be transmitted Transmission errors can also occurs as bursts where several consecutive attempts to transmit a frame end in a transmission errors 5 4 2 Simulation of transmission errors One can identify the following random delays related to the transmission errors e delay between two consecutive transmission error detections e delay between a detection and the consecutive resumption of the nominal communication It would be possible to infer the parameters of a probability distribution for the inter error detection times but when trying to generate these error detection times during a simulation we would have a problem with the fact an error can only occur if there is actually a frame transmission For this reason we have decided to model the inter error detection times in terms of number of consecutive correctly transmitted frames between two transmission error bursts To make this approach work we furthermore add probability distributions for the following aspects e number of consecutive faulty frames error burst e position of the faulty bit in the transmitted frame e time until resumption of normal communication The distribution of the number of consecutive faulty frames burst could for example be specified by a histogram 2009 2014
76. aphics 4 Simulations RandomSourceSeed 709490156681136600 WithBusOfflnducedDelays A System Simulation GateWayDelayDistributions body gt powertrain Uniform 469 SystemStatistics 12d 4 SecondRunSimulations powertrain gt body Uniform 100 4 A 1 Second Statistics for System Simulation 4 BusSimulations 4 A1 body Stuffing 10 Interface BusStatistics l 4 A1 powertrain Stuffing 10 Int s ut 4 gt BusStatistics BusSimulations SystemStatistics P E S Bus BitStuffin BusInterfacesConfig OffsetCo ComOffsetsConfig ClockDriftConfiguration a Graphics powertrain 10 powertrain InterfacesConfig SOAS RandomComOffset0 RandomConstantDrifts 2 Statistics for A Bus body Stuffing 10 body 10 body InterfacesConfig SOA25 RandomComOffset0 RandomConstantDrifts 2 Statistics for A Bus powertrain Stuffing npm ETA t Notice that the generated mnemonic of the second run simulation A 1 in the above screen shot starts with the mnemonic of the reference simulation from which the quantiles are taken A in the above screen shot 2009 2014 RTaW 167 182 RTaW Sim V1 4 7 User Manuat The second run statistics are located below the BusStatistics and SystemStatistics nodes of the SecondRunSimulations The following Screen shot shows the details pane for the response time of frame4 on the bus body Architecture y
77. ated through the context menu of the relation category nodes those with gray labels in the relation trees To create a new bus right click on the Buses node and select the Add New menu item File Samples Simulation Worst Case C Architecture x 4 Architectures Add New D ComPatterns 2009 2014 RTaw 86 182 RTaW Sim V1 4 7 User Manuat This brings up a dialog where some required parameters must be set such as the ShortName and the speed New Bus ShortName Low Speed Cancel If a value is wrong or missing then the field is displayed in yellow The tool tip of the field describes then the problem Speed alue required If you click on the Create Button before all errors are corrected and missing values are provided then the tool pops up a dialog with an error message like this one Error a Values missing ar wrong For some attributes Speed The tool tips explain the error Notice that the tool tips of the labels might provide some information about the required value such as the time unit kbits s in the example of the bus speed If you are not able to provide correct values you can abort the creation by clicking on the Cancel button After having successfully clicked on the Create button the corresponding details pane opens in the tabbed folder on the left of the user interface as shown below for the bus 2009 2014 RTaw 87 182
78. ateways Bus Interfaces Configuration Default Offset Configuration DOAS Inter ECU COM Offset Configuration RandomComOffsett Com Offsets Generator Seed 30092011 Clock Drift Configuration RandomDrift 1ppm r Drift Mode 7 Drift Bound ppm 200 Drift Generator Seed 1102007 Event Occurrences Model Frame Error Model NonBurstyTransmissionErrorModel Communication Pattern Models Pause Resume Stop 2009 2014 RTaWw 43 182 RTaW Sim V1 4 7 User Manuat Make sure to select not only the same bus offset configuration and Sample times as in the previous simulation of see Section 3 1 but also the same inter ECU offset and clock drift configuration to ensure that the results will be comparable Then select the transmission error model described above and run the simulation This time we will not use one of the predefined plots but we will create a plot by hand that will allow us to compare the maxima of the frame response times with and without transmission errors For this purpose click on the Results and Graphics tab on the left side then right click on the Graphics node and finally chose the Add New entry 4 Graphics Stati Add New B S Mono Bus CAN LS 5t Li ng 10 Mone Bus CAN LS Stuffing 10 Effect of event triggered transmissio Enter Effect of errors on maxima as name New Graphic in ShortName rrors on maxima Sco
79. ation A Statistic Scope AndKindofTimes Sample Time hi Then left click on the BusSimulation field and select the simulation that corresponds to the case with drift but without event triggered transmissions ng 10 Offsets DOAS Com Offset RandomComOffset0 Drift No drift ng 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift 1ppm ng 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift lppm Events InFrequentEve Sel Cone To finish change the Statistic field to Maximum the Scope And Kind of Times field to Bus Response Times the Sample Time field to 1d and click Create 2009 2014 RTaWw 38 182 RTaW Sim V1 4 7 User Manuat FrameStatClurve ShortWame Wo events DisplayColor green BusSimulation Mone Bus CAN LS Stuffing 10 Offsets DOAS Com Offset Re Statistic Maximum Scope And Kind of Tirmes Bus Response Ti Sample Time iM r refte Cancel Then create a second curve but this time enter Infrequent events as name and select red as color As entry for the BusSimulation field select the simulation that corresponds to the infrequent events case Select BusSimulation s 10 Offsets DOAS Com Offset RandomComOffset0 Drift No drift 10 Offsets DOAS Com Offset RandomComOffset0
80. ation entry from the simulation menu Simulation Simulation Worst Case Optimization i Initialize from existing Second Run Simulation This brings up a dialog for selecting the simulation from which the reference quantiles Qn will be taken 2009 2014 RTaW 133 182 RTaW Sim V1 4 7 User Manuat Select a simulation Simulation IAB eeu T ven Cancel After having chosen the simulation a simulation configuration dialog is shown in order to start and monitor the simulation The configuration dialog has the title Second run simulation It also allows to change configuration parameters before starting the second run simulation In Section 4 7 5 6 is explained where the results of the second run simulation can be found and how to read them 4 6 5 Command line execution of a simulation It is possible to run a simulation from the command line prompt by providing the simulation configuration in a properties file rtaw sim sim cfg properties where 1 rtaw sim is the name of the command 2 sim cfg properties is name or path to a properties file containing the configuration of the simulation During the simulation the statistics are stored 1 in a RTlaW Sim model which is then saved in the same folder as the configuration file The name of the resulting model file starts with the name of the configuration file followed by a time stamp time when the simu
81. b Framel P E T 100 20 and under the relation category node Frames Architecture x Buses El Low Speed B Frames Framel BusConnections OffsetContigs 4 3 1 2 Creation by duplication Many composed data entities but not all can be duplicated For instance it is possible to duplicate e Architecture e Frame e CANBusInterfaceConfig e OffsetConfig e ComOffsetConfig e ClockDriftConfig e OccurrenceModel of communication patterns e ProbabilityLaw e EventOccurrenceModel e TxErrorModel e Graphic e Curve 2009 2014 RTaw 89 182 RTaW Sim V1 4 7 User Manuat When a data entity consists of sub entities Sub nodes in the exploration tree then also all sub entities are duplicated for example if an Architecture entity is duplicated then also all its ECUs Buses Frames OffsetConfigs etc are duplicated If an entity can be duplicated then a Create Copy menu entry is available in its context menu in the exploration tree Architecture F 4 Architectures 4 Architecture 7 Buses a CANLS 4 Frames E frame framel Rename frame Delete frames Create Copy framed r r and in the table view Canld Canlype ShortName Sender Receivers Paylo TxMode 742 Ox2e6 20A frame Ecu_3 i y Rhees P 198 0x6 70A framel
82. bed how an RTaW Sim system description and which alternative formats such as for example CSV are Supported 4 6 Performing a simulation In this section we describe how to prepare and run a simulation and obtain statistics about the frame response times 4 6 1 Statistical concepts Except in specific cases for instance small networks and no clock drifts or jitters worst case response times WCRT cannot be found with the help of simulations At best the WCRT can be reproduced when the corresponding worst case scenario is known and used to configure the initial conditions of a simulation The reason is that a worst case scenario is one out of a huge number of possible scenarios The maximal response time observed during a simulation is a Statistic that approaches the closest the worst case response time see the following figure but it is a statistic that converges very Slowly and with a speed that cannot be foreseen 2009 2014 RTaw 123 182 RTaW Sim V1 4 7 User Manuat Typical distribution of response times Latency constraint Computed bound WCRT Max observed by simulation Unfavorable scenario Probability 1 10 quantile i i i Response time Statistics that converge within a reachable time horizon are quantiles Formally for a random variable X a p quantile is a value x such that P X lt sx p or equivalently P X gt x lt 1 p In other words it is a threshold L such that for an
83. bus_2 Ecu_4 Gateway bus 2 In Bus Off is Min 410 935 ms Average 425 586 ms Q2 434 411 ms Q3 437 649 ms g4 437 649 ms g5 Q6 Max 437 649 ms Entries Value Count Width 412 485 ms 37 4 749 ms 417 261 ms 296 4 804 ms 422 093 ms 805 4859 ms 476 980 ms 560 4916 ms 431 925 ms 126 4973 ms 436 926 ms 13 5 030 ms 4 7 5 3 Busy period lengths Under the BusStatistics Node of some bus double click on the time horizon for which you wish to see statistics on busy period length on that bus 2009 2014 RTaWw 162 182 RTaW Sim V1 4 7 User Manuat Architecture x A bus_1 Stuffing 10 HPF SOA25 RandomComOffset0 RandomConstantDrifts 1000ppm 1d amp Specific Models kal TimeSpan 1d Results and Graphics 4 Simulations 4 A System Simulation 4 BusSimulations 4 A bus_1 Stuffing 4 BusStatistics gt 12h ih te gt lm gt 22h ObservedBusLoad 0 5096023 BusyPeriodLengths Busy period lengths for bus bus_1 In order to visualize a statistics click into the corresponding field A Bus CAN Stuffing 10 Offsets Zero Com Offset RandomCor TimeSpan 1d ObservedBusLoad BusyPeriodLengths Busy period lengths for FrameRespTimes Frame VARIG MD Awe MTAA ak and select the Show detail entry 2009 2014 RTaW 163 182 RTaW Sim V1 4 7 User Manuat Busy period lengths for bus CANL es Min 9 616 ms Av
84. bytes Ecu 10 198 0xc6 2 0A 8 bytes ShortName C Bit Rate 1 CAN FD Bit Rate Periodic Load 3 Bit Stuffing 10 OffsetConfig BusEvalConfig Scope and Kind of Statistics End to End ShortNa Receivers Period Mini Dea frame32 frame26 frame40 frame35 frame22 framel Ecu_11 Ecu_12 Ecu_13 Ecu_14 Ecu_15 Ecu_16 a gt 201 Oxc9 208 Oxd0 209 Oxd1 215 Oxd7 235 Oxeb 241 Oxfl 271 Ox10f 2 0A 2 0A 2 0A 2 0A 2 0A 2 0A 2 0A frame91 frame62 frame52 frame30 frame42 frame79 frame45 7 bytes 8 bytes 3 bytes 6 bytes 8 bytes 1 bytes 8 bytes 274 0x112 2 0A frame96 NA Fenn AlN 5 bytes O lL aan Specific Models y Results and Graphics Before doing a simulation let us first have a look at the proposed model for event triggered transmissions for this purpose click on the Specific Models tab on left Specific Models Results and Graphics and then click on the InFrequentEvents node EventOQccurrenceModels FrequentEvents TnFrequentEvents TxErrorModels in order to see the parameters of that model InFrequentEvents 2 5 ShortName InFrequentEvents terOccurrencelime Exp lambda 1 100000 FrameDistribution Probability distribution of inter event occurrence times in microseconds The model consists in a probability distribution for the delays in microseconds between successive occurrence
85. cated by the tool tip of the Drift Bound label the unit of the Drift Bound is ppm parts per million and means that the clocks are at most 1 us slower or faster than the nominal clock every 1 second lus being 1 millions part of a second When starting the Simulation RTaW Sim generates a constant clock drift factor for each ECU in the range 1 lppm a Simulate Length amp Sample Times Simulate the traffic on bus CAN LS W Buses CAN LS Bit Stuffing 10 Gateways Bus Interfaces Confiquration Default Offset Configuration DOAS Inter ECU COM Offset Configuration RandomComOffsett Com Offsets Generator Seed Clock Drift Configuration Drift Mode CONSTANTDRIFT Drift Bound ppm Drift Generator Seed Event Occurrences Model Frame Error Model Communication Pattern Models Now make sure to select the Inter ECU Offset Configuration called RandomComOffsetO which has been generated for the first Simulation This will make the comparison meaningful We furthermore choose longer sample times 5s 1h 6h 1d to better see the effects of clock drifts The simulation requires more time than the first one without drift but you are informed of the progress 2009 2014 RTaw 27 182 RTaW Sim V1 4 7 User Manuat Simulation oot ditties td E Pause Stop The speed factor is this time much higher because the GUI overhead is much lower in proportion than in the fi
86. cated in the hardware buffers 3 The number of available hardware buffers has the expected impact on the length of the software queue and the simulation allows to derive statistic on the number of waiting frames In particular if there are as many hardware buffers as frames to send then the length of the software queue will always remain zero Furthermore offsets may allow a zero length software queue even if there are less hardware buffers than frames to send 5 1 3 Arbitration and transmission The bus starts in the idle state Then if at some point in time there is at least one ECU with at least one frame instance to be sent then the waiting instance with the highest priority is chosen for transmission in CAN terminology one says that this frame wins the arbitration phase and the transmission end is scheduled based on e the size of the frame e the bus speed e the bit stuffing model An end of arbitration event is also scheduled at the time where the last bit of the identifier has been transmitted Based on this event the CAN controller start to honor transmission cancellation request that might have been issued during the arbitration phase When the transmission end occurs the frame instance is considered to have arrived at the receivers and its transmission delay ends see Section 5 1 5 At the same moment the subsequent inter frame period end is scheduled with a duration of 3 bits When the inter frame period
87. cerned field and explained by the tool tip Remember that the tool tip is shown when the mouse pointer is positioned on top of the field New Frame ee So ShortName f1 Canlype 2 04 CAN id Dec 4 CAN id Hex Payload amp bytes TxMode Period Period must be set for periodic and mixed frames Deadline Notice that the value for TxMode disables all checks and can be used as temporary value Property Meaning Constraint ShortName Name of the frame Must always be provided CANId Dec CAN identifier in decimal Must always be format provided either in 2009 2014 RTaWw 103 182 RTaW Sim V1 4 7 User Manuat decimal or hexadecimal format CANId Hex CAN identifier in hexdecimal Must always be format provided either in decimal or hexadecimal format CANType e CAN 2 0A 11 bit Must always be identifiers provided e CAN 2 0B 29 bit identifiers Payload Payload of the frame in bytes Must be provided if TxMode is P E P E or Dig Period Period of the recurrent Must be provided if transmission expressed in milliseconds TxMode is P or PE MinimumDel Minimal delay between Must be provided if ay successive transmissions TxMode is E or PE Deadline Latency constraint on the response time of the frame expressed in milliseconds see also Section 2 Role FF First Frame Must be provided if FC Flow Control
88. cifies for which statistic the convergence graphic should be created 4 7 4 2 Creation of custom graphics Custom graphics can be created in the Results and Graphics section under the Graphics node We base the following explanations on the example of Section 3 6 In order to create a graphic right click on the Graphics node lt lt Architecture lt lt Specific Models gt gt Results and Graphics 4 Simulations A System Simulation SecondRunsSimulations Evaluation graphie Add New and then select the Add New entry 2009 2014 RTaW 145 182 RTaW Sim V1 4 7 User Manuat New Graphic Graphic ShortName SY ScopeAndKind Bus Respons K Bus Response Times End to End Response Times Inter Transmission times In the creation dialog a name for the graphic ShortName has to be provided and also the default scope and kind The scope and kind Specifies two aspects e Scope the perimeter of the statistic that could be local to a bus or spanning several buses with crossing of gateways e Kind the statistic to display such as the frame response times or the frame inter transmission times There are currently three options Bus Response Time Displayed frame response times are local to a bus t e they correspond to the delay from the instantiation of the frame in the sending ECU until the transmission end End to end Response Time
89. cus so that the zooming with Ctri Mouse Wheel does work without having to first click on to the panel Version 1 4 1 e Introduction of response times curves with network scope FrameNetworkStatCurve This allows to summarize in a unique curve all response times of a CAN communication architecture e Combobox editor for frame payloads e Correction of some problems and small improvements Version 1 4 0 e Introduction of an Architecture entity that groups buses and ECUs this allows to describe analyze and compare several architectures in the same file e Introduction of a graphical viewer and editor of the topology of an Architecture e Added where appropriate a Create copy menu entry in context menus of tree nodes and table lines e Added support for CAN FD e Added possibility to initialize the simulation configuration dialog according to an already performed simulation 2009 2014 RTaw 10 182 RTaW Sim V1 4 7 User Manual e Introduction of an option for load scaling which allows to preserve original frame identifiers Version 1 3 19 e Simulation configuration dialogs keep now the parameters from previous invocation until the software is quit Version 1 3 18 e Optional receiver column added to csv import format see Section 4 2 1 1 4 e For more comfort the default CAN bus interface configuration is now directly visible in the CANBusInterfacesConfig pane see Section 4 4 4 e Ad
90. d End to End Response Times Inter Transmission times Change the name of the curve to End 2 end bus 2 the color to green the ScopeAndKind property to End to End Response Time and then turn back to the graphics tab 2009 2014 RTaWw 152 182 RTaW Sim V1 4 7 User Manuat Statistics 14 ms 13 ms 12 ms 11 ms A 10 ms 7 Ooms 4 8 ms 7 ms 6 ms 5 ms Response Times in 4 ms 3 Ms 7 12 39 106 128 179 242 282 359 432 465 580 642 700 722 750 869 Frames in increasing order of identifiers Local bus_2 End 2 end bus_ 2 The two curves are superposed for the frames the are transmitted only locally on bus 2 For those that are forwarded by the gateway from bus 1 two bus 2 the green curve shows the end to end response time Let us also illustrate the FrameNetworkStatCurve For this purpose add a new curve New Curve FrameBusStatCurve FrameNetworkStatCuve FrameWertCurve FrameDeadlineCure Statistic gt ScopeAndKind Bus Response Tir Sample Time O O7 Create Cancel E and make sure to select the appropriate type from the combo box Then enter the needed information 2009 2014 RTaW 153 182 RTaW Sim V1 4 7 User Manuat FrameNetworkStatCurve Short ame Network DisplayColor r
91. d Graphics 4 2 1 4 3 Selective This functionality allows to import parameter sets with automatic mapping of related entities Parameter sets that may be imported are e CANBusInterfaceConfig e OffsetConfig e ComOffsetConfig e ClockDriftConfig e OccurrenceModel of communication patterns e ProbabilityLaw e EventOccurrenceModel e TxErrorModel Once you have selected the model file the following selection dialog appears 2009 2014 RTaWw 69 182 RTaW Sim V1 4 7 User Manuat a Architectures 4 Architecture 4 Buses a CAN LS a CanBuslnterfacesConfigs Default Offset ontigs DOAS SOAS Shaping5 Zero Ts ClockDrittConfiqurations No drift RandomDrift 1ppm ComOffsetsConfigs RandomComOffset0 yr Select Cancel Notice that with Ctrl left mouse button you can select several individual parameter sets for import It allows for example to import an OffsetConfig from an other model file that contains the description of the same or similar communication configuration The mapping of the frames concerned by the offsets is established based on the name of the frame the name of the bus to which the frame belongs and the name of architectures to which the bus belongs If a correspondence can not be established automatically then the tool asks the user to identify explicitly the corresponding entity The later may for example happen if an entity has been renamed
92. dapt column widths to contents Copy to Clipboard in CSV Format Show selected CANInterfaceSpeciicConfig Delete selected CANInterfaceSpecificContig Add New CANInterfaceSpecificContig Double click on the line or choose the Show selected menu entry from the context menu in order to edit the parameters 2009 2014 RTaWw 109 182 RTaW Sim V1 4 7 User Manuat CAN LS Default CAN LS InterfacesConfig CAN LS Defa CAN LS InterfacesConfig CAN LS Defa 23 BusConnection Ecu_ CAN LS Queuing HPF TxBufferCount 3 UseHwCancellation SingleShot I Order Queue ButferDelay Since we are looking at a specific bus interface configuration an additional category is available for the instantiation order SpecificlxOrder It allows to define a custom instantiation order by adding the frames sent by the concerned ECU in the corresponding order CAN LS Default CAN LS InterfacesConfig CAN LS Default Ecu_2 fs BusConnection Ecu 2 CAN LS Queuing HPF TxBufferCour UseHwCancellation SingleShot Order Queve2BufterDelay After having clicked on the Create Button an empty list appears in the field In order to modify it select the Show details entry from the context menu UseHwCancellation SingleShot E keQrder Show details QueuetBufferDelay b Delete Create New Then the details pane appears where you can add the frames in t
93. ded a needed constraint on transmission offsets of frames offset lt period e Java 7 is now required Version 1 3 17 e Glossary section e Shortcut Ctri S has been added for saving the current model to its file e Corrected problem that was hindering the simulation without communication patterns Version 1 3 16 e Moved Pdf of user manual and sample csv import files to the Samples menu inside the tool e Correction of rounding error in histogram buckets has no effect on values of quantiles e Improvements of the frame panel with respect to consistency constraints e New option for frame dialogs that allows to decide when the delay exactly starts instantiation or transmission start of request frame see Section 4 4 6 2 e More and better default values for graph attributes e Description of second run simulations added Version 1 3 15 2009 2014 RTaw 11 182 RTaW Versi Versi Main S m V1 4 7 User Manuac Gantt charts added for every bus concerned by an end to end unfavorable scenario Correction of hexadecimal CAN id in frame details pane on 1 3 14 Completion of check for missing frame receivers Automatic update of receivers column added Sample size to histogram added to data panes on 1 3 13 Time units have been added to response time statistics in the frames table of the bus details pane and the possibility to export as CSV with or without units Frame error rate F
94. drives the transmission are assumed to not drift apart provided periodic transmissions then the response times statistics converges very soon basically two Icm of the frames periods is enough and it is not needed to simulate for a longer duration Of course as soon the microcontroller clocks may have drifts as it occurs in practice then it is less obvious to know how long is enough but RTaW Sim helps you in that regard with the possibility to visualy check the convergence of statistics see Section 4 7 To begin with the tutorial open the sample file that corresponds to its start File Simulation Worst Case Optimization What If Plot Tools Arch Case Studies A Tutorials Can Bus Tutorial CAN landmark 0 sim hs Architecture Gateways gt Tutorial CAN landmark 1 sim Buses Tutorial CAN landmark 2 sim Ecus Tutorial CAN landmark 3 sim ComPatterns Tutorial CAN landmark 4 sim Double click on the node representing the bus called CAN LS LS stands for Low Speed this is a 125kbit s network 2009 2014 RTaw 14 182 RTaW Sim V1 4 7 User Manuat File Samples Simulation _Worst Architecture 4 Architectures 4 Architecture 4 Buses gt CAN LS 4 Ecus R gt Ecu 0 to get a detailed view of the bus RTaW Sim v1 4 2 Pro Model not yet saved ee Xx File Samples Simulation Worst Case Optimization What If Plot Tools Architecture amp Architecture CAN LS X
95. e Ad Frame A3 Frame Ad Frame Ao Frame A110 Frame A11 Bus Name Speed Frames Name Frame Bl Frame B2 Frame Bs Bridge Ecu Name Source Bus Target Bus Delay Frame Mappings Source Frame Frame A11 Bridge Ecu Name Source Bus Target Bus Delay Frame Mappings Source Frame Frame Bl Frame B2 CAN_A 125 Sending ECU Receiving ECL Identifier CANType Payload TxType Period MinDelay Offset Offset ECUI 0x583 FD STD 32 P ECUI ECU 0x17640000 FD_EXT 12 P ECU ECU1 ECU4 0x107C0000 FD EXT 64 P E ECU ECU1 0x66C FD STD B ECU ECU1 Ox6F80000 FD_EXT B ECUS ECU1 0x224 STD 5P ECUS ECU Ox3F16357 EXT 7P CAN B 50 100 100 1000 20 Off1 Off2 0 0 0 0 20 0 oo Sending ECU Receiving ECL Identifier CANType Payload TxType Period MinDelay Offset Offset ECU3 ECU2 61341696 EXT AP ECU3 ECUS i 120 STD 8P ECU2 ECU3 183500800 EXT B ECU2 CAN_A CAN B Target Frame Frame B3 ECU2 CAN B CAN_A Target Frame Frame A4 Frame Ao Table 1 Sample CSV import file shown as table Table 1 shows the contents of the sample file Sample Csv Import File which is accessible through the Samples menu 2009 2014 RTaw 100 10 Off1 offs 0 0 64 182 0 5 RTaW Sim V1 4 7 User Manuat Simulation Worst Case Optimization What If Plot Tool Case Studies an Tutorials Csv Import gt SampleCsvForlmportxls SampleCsvForlmport_CAN FD xls SampleCsvForImport_CAN20B xls Notice tha
96. e Value Se Curves ShortName Display Hidden ScopeAn Mn Stati Adapt column widths to contents Copy to Clipboard in CSV Format Show selected Curve Create Copy Delete selected Curve Add Curve 4 nm Data Graphic The following dialog will appear FrameBusStatlurve ShortNarme SY DisplayColor black BusSimulation Statisti ScopeAndKind Bus Response Tir SampleTime i Notice the default value for the ScopeAndKind property which is equal to the one of the graphic The type of curve to draw is Specified by the combo box selection four options are available FrameBusStatCuve FrameNetworkStatlurve Frame WertCurve FrameDeadlineCurve FrameBusStatCurve Displays frame response time 2009 2014 RTaWw 148 182 RTaW Sim V1 4 7 User Manuat Statistics for a specified bus FrameNetworkStatCurve Displays frame response time Statistics for all buses of a Specified architecture This kind of curve provides a comprehensive overview of all relevant frames response time of the network of buses that is described by an architecture entity If a frame is produced ona first bus and forwarded to a second bus through a gateway then the curve will contain two points one for the local response time on the first bus and one for the end to end response time that spans both buses FrameWcrtCurve Displays frame worst case response time bounds Fram
97. e and in csv files in the indicated folder Click the OK button and then close the simulation dialog by clicking on Es 3 2 Visualizing statistics Now let us look at the results First of all the obtained statistics can be visualized in the CAN bus tab by choosing the corresponding bus Simulation configuration BusSimConfig 2009 2014 RTaw 20 182 RTaW Sim V1 4 7 User Manuat CAN LF 2 ShortName CAN BitRate 125k CAN FD Bit Rate Periodic Load 31 25 Bit Stuffing 10 OffsetConfig BusSimConfig A CAN LS Stuffing 10 Default DOAY RandomComOffset0 RandomConstantDrifts 200ppm bi BusEvalConfig Sample Time 15 Scope and Kind of Statistics End to End Resp Canld Canlype ShortName Sender Receivers Paylo TxMode Role Period Mini Dead TxOfft 36 0x24 20A frame Ecu 12 5 bytes P E 200 ms 50 ms 165 ms and a sample time Sample Time 15m As a result the columns Min Average are filled you probably need to scroll or enlarge the window to see them well To get more easily an insight on how these statistics relate one to each other let us draw a graphic For this purpose one can use one of the predefined graphics pull down the Plot menu Tools Create graphic with all statistics Create graphic showing convergence for a statistic Worst Case Scenario Observed Defavorable Scenario and select the Create graph
98. e average throughput is described in Section 4 7 5 4 4 4 6 2 Frame Dialogs Frame dialogs are based on two kinds of frames e Request Frame e Response Frame These are the first entities that need to be defined in RlaW Sim See Section 4 4 2 on how to create frames and define their properties These frames must have the TxMode set to Dlg with the respective roles REQ and RESP The following table describes the frames that need to be defined if the dialog is local to a bus Purpose of Payload TxMode Role frame Request Frame Payload of request Dlg REQ frame Response Frame Payload of Dig RESP response frame 2009 2014 RTaw 118 182 RTaW Sim V1 4 7 User Manuat If the frame dialog is crossing a gateway then you have to create the corresponding frames also on the other buses You must also make sure that the frames which are sent by the gateway have the TxMode set to B and no Role specified the role and also the payload are inherited in that case through the gateway frame mapping that will be defined later The following table describes the frames that need to be defined if the dialog is crossing one frame gateway Purpose of Bus Payload TxMode Role frame Request Frame busl Payload of request Dig REQ frame Copy of the bus2 Must not be B Must not be Request Frame provided provided on the second since it is Since it is bus inherited inherited Re
99. e on the last bus For this reason statistics on end to end response times are collected for each frame that is forwarded by a gateway on the considered bus see Section 3 6 for how to visualize them 5 3 Event triggered transmissions of frames In this section we describe how event driven transmissions of frames can be simulated 5 3 1 Overview Event triggered transmissions may occur with purely event driven frames type E or with mixed frames type P E When an event occurs the induced the transmission of a frame then this frame is instantiated and queued at the first time in the future which is compatible with the MinimumDelay of the frame The later specifies a minimal time between successive instantiations of a same frame 2009 2014 RTaW 178 182 RTaW Sim V1 4 7 User Manuat 5 3 2 Simulation of event triggered transmissions Event triggered transmissions are often due to events whose frequency cannot be characterized in a deterministic manner e g driver s action in a vehicle In such a case a solution can be the usage of an inter arrival time probability distributions for the events that induce frame transmissions This probability distribution is used to program event occurrences When such an event occurs a second probability distribution is used to determine which frame is to be actually instantiated This is the approach used by RTlaW Sim The parameters of that model are specified by the attribut
100. eDeadlineCurve Displays the deadlines of the frames Let us create a FrameBusStatCurve As for all curves a name ShortName and a color DisplayColor are required for every curve Enter Local bus 2 and keep the proposed default value black To select the BusSimulation left click on bus simulation field a panel with all available simulations will appear 2009 2014 RTaw 149 182 RTaW Sim V1 4 7 User Manuat Select BusSimulation zs 4 Simulations 4 A System Simulation 4 BusSimulations A bus_1 Stuffing 10 HPF SOA25 RandomComOffset RandomConstantDrifts 1000ppm A bus_ Stuffing 10 HPF SOAT RandomComOffset RandomConstantDrifts 1000ppm j Choose the one for bus_2 and then click on Select ShortName DisplayColor black BusSimulation A bus 2 Stuffing 10 HPF SOAS f RandomComOffsei Statistic Q 1 10 6 ScopeAndKind Bus Response Tir Sample Time 1d z When a bus simulation has been selected the dialog proposes the default values for the Statistics and the Sample Time property The entire simulation horizon is proposed as Sample Time and the highest order quantile for which robust estimations are available see Section 4 6 1 In order to visualize the corresponding graph click on Create and then on the Graphic tab 4
101. ect Disconnect from bus1 archl Show details Disconnect from bus iS r Delete e right click on the corresponding node in the exploration tree on the left and select Delete 4 Ecus 4ecul 4 BusConnections 4 busl SentFrames ReceivedFrames GateWays ecu 4 BusConnections bus GateW Delete N ComPatterns 4 4 1 3 Entity details Double clicking on a node in the architecture graph or right clicking and selecting Show details of allows to open the corresponding details panel Let us take the example of an ECU node 2009 2014 RTaw 99 182 RTaW Sim V1 4 7 User Manuat Ecu_18 Ecu_19 Ecu 20 gt Show details of Ecu_1 Disconnect Ecu 1 from d Delete Ecu_1 Ecu_11 Ecu_12 Ecu_13 Ecu 14 Ceo 1 Selecting Show details or Ecul opens the details panel of Ecul Architecture Ecu 1 23 ShortNarme Ecu_1 BusConnections Bus SentFrames bus 1 Frame_ Frarne_62 Frame_ab Frame_168 Frar bus 2 Frame_d Frame_1c Frame_21 Frame_4d Frame 4 4 1 4 Automatic layout The context menu of the architecture panel contains an entry for laying out the architecture graph right click on the panel and select Layout as circles 2009 2014 RTaw 100 182 RTaW Sim V1 4 7 User Manuat Architecture i Ecu 5 Add new ECU Add new bus Layout as circles Save as Image As a r
102. ed SystemSimulation A System Simulation Sample Time 1d Statistic Q 1 10 6 The resulting curve is hard to read Statistics m EER Response Times in a a D e M W fe n om co A aR eo a 4 74 i44 240 323 392 465 5602 623 722 836 9i 986 Frames in increasing order of identifiers Local bus_ End 2 end bus_2 Network Data Graphic Go back to the Data tab and hide the first two curves 2009 2014 RTaWw 154 182 RTaW Sim V1 4 7 User Manuat Statistics 25 End 2 end bus ShortName Statistics Style Absolute Value r Scope Curves ShortName Display Hidden ScopeAndkir Local bus_ black show v Bus Responsi End 2 end bus_ green Network red End to End F show Network The resulting graphic may seem a bit chaotic but is has the advantage of showing all relevant response times Statistics 15 ms 14 ms 7 13 ms 7 12 ms ss ms 7 m i Ea Ln 9 ms 7 8 ms 1 7 mE 6 ms 5 ms 4 Response Times in 4 ms 3 ms 1 2 mE 1 ms 7 0 ms i 4 f4 144 240 323 392 465 562 623 722 36 9i0 986 Frames in increasing order of identifiers Furthermore if response times are shown as percentage of the deadlines this kind of graphic allows to easily identify critical frames see Section 4 7 4 3 2009 2014 RTaW 155 182 RTaW Sim V1 4 7 User Manuat 4 7 4 3 Usage of periods and deadlines in response time graphics
103. ed to transmit the bits of the frame since COM layer Operation are supposed to have zero delay 5 2 Gateways In this section we describe how the functioning of gateways is Simulated 2009 2014 RTaW 177 182 RTaW Sim V1 4 7 User Manuat 5 2 1 Frame gateway The basic functioning of a frame gateway is the following when a frame to be forwarded arrives on the reception side it is taken into account by the frame gateway which will instantiate the corresponding frame in the target bus interface after a certain amount of time has elapsed From that point on the frame is handled as any other frame sent by the gateway ECU on the target bus waiting in frame queue or in transmission buffers RTaW Sim simulates this behavior in the following way as soon as a frame to be forwarded arrives on the reception side a delay timer is armed When this timer expires the corresponding target frame is queued in the target bus interface together with the other frames sent by the gateway ECU The delay time is randomly chosen according to a specified probability distribution see Section 3 6 5 2 2 Transmission delay statistics When a frame gateway is used the actual source of a frame is located on some other bus This implies that the total delay for the multiplexed communication is the sum of bus delays and gatewaying delay starting with the instantiation of the first frame on the first bus and ending with the transmission end of the last fram
104. edition 2009 2014 RTaw 6 182 RTaW Sim V1 4 7 User Manuat For more information please license realtimeatwork com contact US at RTaW Sim editions Main Feature Starter Pro Simulation of CAN buses 7 CAN2 0A CAN2 0B ARINC 825 CAN FD Gateways Communication patterns Event driven transmissions Frame dialogs e g diagnostics Segmented transmissions Transmission errors Worst case analysis of CAN buses CAN20 A CAN20 B CAN FD Rare events statistics gt Q6 Optimizations CAN identifiers Transmission offsets What If analysis Scale Load Migration to CAN2 0B Migration to CAN FD Productivity features Selective import Import of typical tx error models Duplication of entities Curves with network perimeter 2009 2014 RTaw KX KK mK XIX mK LK KK KKK KK KR EK LN KL KN lt lt XXX lt A a Pg ag ey a es a a OO 7 182 RTaW Sim V1 4 7 User Manuat 1 3 Installation Installer based distributions are available for Windows and Linux More information and the download links for the free version are available at the following address http www realtimeatwork com downloads 1 4 Support This help is integrated in the tool as html pages and as pdf document File Samples Simulation Worst Case Optim
105. ee Section 4 6 1 for more details Some graphs may not be visible because for a same frame many Statistics may have the same value Let us therefore hide the two quantiles click on the Data tab at left bottom asap dB O25 0 00 26 135 209 274 3l4 348 6460 t Fr Minin Graphic Eales This will bring up a table with the different graphs Now left click on the Hidden cell of the line that corresponds to the 1 10 7 quantile and mark the graph as hidden Curves ShortName DisplayC Hidden Minimum aqua Average black Q 1 10 2 blue Q 1 10 3 fuchsia Maximum green Do the same with the 1 10 quantile Then click in the Graphic tab on the left bottom 4 SERN apata in order to see the result 2009 2014 RTaw 23 182 RTaW Sim V1 4 7 User Manuat Statistics for A CAN LS Stuffing 10 Default DOAS RandomComOffset0O NoDrift 15m 4ms 3 75 ms 3 o ms 3 25 ms 7 3 ms 7 E 2 75 ms 2 9 M5 7 z 2 25 MS 2 ms 1 75 ms7 End Response Times in ms 1 5 ms 7 to 1 25 ms End 35 122 208 2 71 308 334 458 483 517 560 586 618 697 743 782 8 0 907 940 972 1045 1398 Frames in increasing order of identifiers Minimum Average Maximum Now all graphs are well visible at least for most frames For the frames where only the minimum is visible the average and maximum have the same value as the minimum meaning that the response time of
106. eld before s and click Add s stands for second and m for minute see 2 Do the same for 1m and 15m The result should look like EE Simnttion of Actes O File Len gth amp 5 ample Times Length of Simulation 4 Buses CAN LS Gateways Intermediate Statistics 55 1m 15m Include bus off blocked Generate Trace Pause Resume Stop Now we are ready to start a simulation the largest Intermediate Statistics time being used as simulation length 2009 2014 RTaWw 19 182 RTaW Sim V1 4 7 User Manuat Simulation When the simulation ends a second dialog pops up Simulation Report 28 Simulated 15m in 2s 117ms P Simulation speed 739 3 kilo events second Speedup 425 times faster than real world system Statistics are also stored in the sub folder De Soft RTaW Sim sim_ results Tutorial CAN landmark 0 002 and gives some information about the simulation e simulating the CAN bus communication during 15 minutes has taken 2 seconds and 117 milliseconds e the speed of the simulation was 739 3 kilo events second depending on the CPU power a speed of up to 2 mega events second is achievable e running the same experiment on real hardware would have taken 425 times longer depending on the network complexity and the CPU power a speedup between 100 and 1500 is achievable e the statistics are stored both in the RTaW Sim input fil
107. elected Frarne Add New Frame Copy to Clipboard in CSV Format And finally they can also be accessed through the context menu of a reference field in a details pane right click on the field as shown below for the BurstLength reference in the details pane of a TransmissionErrorModel BurstyTransmissionErrorModel 25 ShortName BurstyTransmissionErrorModel Burstlength BurstLength Remove Ser Parameters that are displayed in a text field can be edited as expected simple text editor If an input is illegal letter when a number is expected then an error message pops up to inform the user about the problem If an entered value is invalid negative number when a positive number is required or required but deleted then the back ground of the text fleld changes to yellow and the tool tip provides more information about the error 2009 2014 RTaw 91 182 RTaW Sim V1 4 7 User Manuat References to other data entities may be either displayed in a combo box DisplayColor green tomComorrseto Sample Time or in a text field BusSimulation Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset Drift No drift In the later case they can be changed by clicking on the field which brings up a dialog that allows to select a new reference 4 Simulations Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOftset0 Drift Mono Bu
108. erage 1 468 ms Q2 5 407 ms Q3 7 463 ms Q4 9 176 ms Q5 9 176 ms 06 9176 ms Max 9 176 ms UnfavorableScenarioConfig Entries Value Count Width 0 618 ms 1403038 0 007 ms 0 751 ms 149744 0 008 ms 0 824 ms 643652 0 009 ms 0 893 ms 839075 0 011 ms 0 968 ms 21232869 0 011 ms 1 367 ms 4 0 015 ms 1 448 ms 36332 0 017 ms 1 516 ms 76601 0 017 ms 1 569 ms 7879 0 018 ms 1 587 ms 139172 0 018 ms 1 643 ms 10803 0 019 ms 1 TIN rar AJATRN AAW rar Data dicaj The lower part shows the table view of the histogram which can also be visualized as a graphic by selecting the Histogram tab 4 7 5 4 Segmented transmission related response time Segmented transmission response times statistics are accessible under the node SegmentedTxTimeDatas besides the OccurrenceModels node 2009 2014 RTaw 164 182 RTaW Sim V1 4 7 User Manuat RTaW Sim v1 4 7 Pro C Users jorn svn pegase netcar svn NetcarSiml docs help odt material DigAndSegTx DigAndSegTx sim Le File Samples Simulation Worst Case Optimization What If Plot Tools Architecte Architecture 4 Response times for Segmented Transmission Segmented x after 1d 4 Architectures SegmentedTx CANI SegmentedTx Throughput 3 061 kbytes s 4 Architecture a Buses Min 31 992 ms Average 36 586 ms t CANI b CAN2 Qz a Ecus Q4 41 967 ms g5 42 698 ms p ECU1 40 078 ms g3 41 248 ms p ECUZ Q 43 692 ms Max 43
109. es and the Scheduling policy of the CAN communication stack Frame response time Delay between the instantiation time and the transmission end of a frame instance Frame Deadline Latency constraint on the response time of the frame 3 Quick start The goal of this section is to allow you to get quickly an idea about the kind of investigations RlaW Sim allows to conduct on CAN based communication Systems In Section 3 1 it is shown how to obtain statistics about the response times of CAN frames i e the time between a frame is ready to be sent and the time it is received by all stations and Section 3 2 is dedicated to the exploration of the simulation results In Section 3 3 we show the effects of clock drifts which are the main driver for the 2009 2014 RTaw 13 182 RTaW Sim V1 4 7 User Manuat variability int the response times of periodic frames In Sections 3 4 and 3 5 we show how the effects of event driven transmissions and transmission errors can be analyzed for response time maxima of CAN frames And finally in Section 3 6 the simulation with gateways is illustrated 3 1 Evaluating frame response times This is the most basic feature of a CAN simulator and probably the most useful because having a precise idea of the frame response times on a CAN bus is difficult without a tool aS soon as there are more than a few frames It should be pointed out that if the microcontroller clocks that
110. es of an EventOccurrenceModel e InterOccurenceTime probability distribution of the inter occurrence times of the events that induce frame transmissions e FrameDistribution probability distribution of the index of the frame to be chosen when an event occurs The range of indexes is given by the listing all event driven frames in the order of increasing CAN Id 5 3 3 Transmission delay statistics From the transmission delay statistics point of view event driven transmissions are considered as any other kind of transmission 5 4 Transmission errors In this section we describe an advanced feature of the simulator which is the modeling of transmission errors 5 4 1 Overview There exists several causes for a CAN frame to end up being considered as faulty and to be retransmitted e incorrect bit stuffing e fixed value fields having the wrong value e CRC error detection e an ECU is sending a dominant bit but sees a recessive one 2009 2014 RTaw 179 182 RTaW Sim V1 4 7 User Manuat When a frame is detected as faulty by the sending ECU or a receiving ECU then the current transmission is ended through the sending of an error frame The sending of the error frame can lead other ECUs to also send error frames Communication returns back to nominal State after a delay that spans from 6 to 29 bits The faulty frame enters then again the arbitration phase unless a frame with a higher priority has arrived into a tr
111. esult the buses and their ECUS are laid out as circles 2009 2014 RTaw 101 182 RTaW Sim V1 4 7 User Manuat Architecture x Ecu_s Ecu19 Ecu_18 Ecu_s Ecu_ Ecu 20 Ecu_l Cm O A a a Pi Ecu 10 Ecu_13 Ecu_14 Ecu_9 4 4 2 Frames Frames can be created in the context of a bus through the Add New Frame entry of the context menu of the frames table in the bus pane Architecture CANI CAN es Buses ShortName CAN Speed 500 bal BusSimConfig CAN Ecus BusEvalConfig Sample Time Scope and Kind of Statistics Recenver Sender Canld CanType ShortName Sender ComPatterns 50 0x32 20A CAN2 Frame 1 Adapt column widths to contents Copy to Clipboard in CSV Format Show Timing Chain Show Histogram Show Unfavorable Scenario Show selected Frame Delete selected Frame Add New Frarne 2009 2014 RTaWw 102 182 RTaW Sim V1 4 7 User Manuat The following screenshot shows the Frame creation dialog New Frame lt e eo ShortName A Canlype 20A i CANid Dec E i CANid He Payload TxMode Role Period MinimumDelay Deadline Create Cancel The different properties are explained in the following table Notice that transmission modes TxMode imply certain constraints on the values of other properties which are enforced by the GUI A violation of these constraints is notified by the yellow color of the con
112. et saved File Samples Simulation Worst Case Optimization What If Plot Tools Architecture Architecture CAN LS frame5 4 Architectures 4 Architecture J 4 Buses CAN id Dec 601 CAN id Hex 4 CAN LS 4 E ra m Cc ShortName frame5 CanType 2 0A Payload frameO framel frame2 frame3 TxMode Period Deadline Role 500 ms MinimumDelay frame4 frame5 frame6 frame frames frame9 frame10 framell framel2 Specific Models Results and Graphics A containment tree consists of two kinds of nodes e Data entity nodes displayed in black e Relation category nodes displayed in gray A data entity node represents a data entity such as the description of a CAN bus and is typically labeled by the name of the entity it represents A relation category node represents a certain type of relation with other data entities like the set of frames on a bus 2009 2014 RTaWw 59 182 RTaW Sim V1 4 7 User Manuat A double click on a data entity node opens the corresponding details pane as can be seen on the left side in the figure above for the example of a frame The Architecture tree covers system defining entities like buses ECUs frames and various configuration parameter definitions like frame offset configurations The Specific Models tree covers data entities like models for transmission errors event driven frames transmission and the related probability
113. etween the transmission start and resorption of the error This effect is even stronger with bursty errors Performing a similar simulation based on the BurstyTransmissionErrorModel and adding a similar graph to the previous plot gives the following graph 2009 2014 RTaW 47 182 RTaW Sim V1 4 7 User Manuat Effect of errors on maxima 11 ms 10 5 ms 7 10 ms 9 5 ms 7 9 ms 5 8 2 ms 8 ms 7 J3 ms 7 7 ms 7 6 5 Ms 7 6 ms j 5 9 MS 7 3 ms 7 4 9 ms Response Times in ms 4 ms 7 3 0 ms 7 3 ms 7 2 0 MS 7 2 ms j 1 5 MS 1 1 ms 7 0 5 ms 0 ms E T T T T T T T T T T T T T T T T T T T T 36 135 209 274 314 348 460 487 524 568 590 627 709 756 786 876 915 942 983 1211 1445 Frames in increasing order of identifiers No Errors Non bursty errorsNon bursty errors Bursty Errors As expected the bursts induce even longer response times 3 6 Simulation with gateways Samples Simulation Worst Case Optimization What If Plot Tools Case Studies O Can Bus Gateways Tutorials a Architecture Tutorial Gateway landmark O sim Tutorial Gateway landmark 1sim In this section we will see how to configure a simulation with gatewaying and how to visualize the results To start with let us open an example with gateways The sample system contains a CAN bus at 250 kbit s a CAN bus at 500 kbit s and one gateway that connects them see below The onl
114. gateway crossing only the frames on the first bus need to be identified since the frames on other buses can be found by following the frame mappings of the gateways Frames that may be selected for the intended role appear in bold in the selection dialog 2009 2014 RTaWw 115 182 RTaW Sim V1 4 7 User Manuat a a a Frames CANI Pil CANI P2F CANI PIF CANI FT CANI CF CAN FF CANI RESP CANI REQ CANT Framel CANI Frame l If for some reason the frame can not be selected an explanation is provided by the tool tip just position the mouse on top of the frame you would like to select e a a Frames CANI Pit CANI P27 CANI PIF CANI FC CANI CF CANI FF CANI REI CANI REL The Frame is already used as FirstFrame by some other SegmentedTx f CANI Framel CANI Frame Finally you need to define at least one occurrence model which specifies the values of all the parameters that determine the actual behavior during a simulation 2009 2014 RTaW 116 182 RTaW Sim V1 4 7 User Manuat ECU GateWay 4 ComPatterns a CANL Seqg Tx OccurrenceMag Add New The parameters have the following meaning Parameter Meaning ShortName Name of the occurrence model STmin Minimal distance between the transmission acknowledgment of a Consecutive Frame and the instantiation of the next Consecutive Frame The granularity is 1 microsecond BS Block size i e the number
115. hat no priority inversions occurs This is different from simulation where different non ideal communication stack behaviors may be taken into account see Sections 4 4 4 and 4 6 3 e if no gateway is present then a single computation step is sufficient furthermore the results are worst case response times that may effectively occur e if gateways are present then thenumber of iterative computation steps that is required cannot be foreseen beforehand furthermore the results of the computation are upper bound on the worst case response times When the computation is done a SystemEvaluation entity appears under the Evaluations node in the results and graphics section Architecture lt lt Specitic Models lt lt gt Results and Graphics 4 Simulations 50 4 System Simulation SLA System Simulation 3 4 System Simulation SecondRunSimulations 4 Evaluations a Worst Case see Analysis a Graphics On how to efficiently visualize the computed worst case response time bounds see Sections 4 7 1 and 4 7 4 2 4 2 5 Optimization The optimization menu offers functionalities related to the optimization of offsets and priorities Optimization What If Plot Tools Rate Monotonic CAN Identifiers Random Offsets SOA Offsets 2009 2014 RTaw 75 182 RTaW Sim V1 4 7 User Manuat 4 2 5 1 Rate Monotonic CAN Identifiers This functionality allows to modify CAN identifiers so that a frame wi
116. he wished order 2009 2014 RTaW 110 182 RTaW Sim V1 4 7 User Manuat CAN LS Default CAN LS InterfacesConfig CAN LS Default Ecu_2 CAN LS InterfacesConfig CAN LS Default Ecu_2 IkOrder 24 FramelnixOrders frarnel2 ane Show details of selected Frame Remove selected Frame reference Add Frame reference Meaning Queuing Queuing policy of the software queue which is used when the tx buffers are full e HPF highest priority first e FIFO first come first serve TxBufferCount Number of hardware buffers dedicated to transmission UseHwCancellation Specifies if cancellation of ongoing transmission requests is used when a higher priority frame appears at the head of the software queue SingleShot e True checked when a transmission error occurs the frame ts not automatically resent e False not checked when a transmission error occurs the frame is automatically resent IxOrder Determines in which order frames are instantiated when their creation is programmed at the same logical moment in time There are two categories of instantiation order e GenericlxOrders e Highest Priority First Order ideal case e Inverse Order of Highest Priority First opposite of the ideal case e Random Order mixture of all possible cases e SpecificlxOrders a specific and fixed instantiation order see above Queue2BufferDelay Delay in us for copying a frame from the queue 2009 2014 RTaw
117. he delay distribution meDelay 4 7 Exploring performance evaluation results The statistics that have been obtained by means of a simulation can be explored and visualized in several ways 4 7 1 Table view of frame response times The details pane of a bus allows to visualize frame related statistics and worst case response times bounds under the form of a table To use this feature the user must open a bus details pane and select a bus simulation configuration pus 1 es ShortName bus 1 Speed 500 Periodic Load 45 4 Bit Stuffing 10 BussimContig as BusEvalContig A Bus bus 1 Stuffing 10 Offsets SOA 5 Com Offset RandomComOffset0 Drift RandomDrift 1000ppm Sample Time Mono Bus bus 1 Stuffing 10 Offsets SOA 5 Com Offset RandomComOffset0 Drift RandomDrift 1000ppm Note that the names of results which have been obtained in the context of a multi bus simulation are prefixed by letter like A for example while names of the results obtained through a mono bus simulation start with Mono bus The configuration parameters that were used in the simulation that lead to a certain Simulation result can be viewed in the corresponding details pane accessible through the Results and Graphics tree Architecture a Specific Models Te Se 3 z 4 Simulations 4 A System Sim ay BusSimulation System Statistics Mono Bus bus_1
118. he quantile is almost always immediately followed by a response time below the quantile 4 1 12d Stay times above Q4 for framed 2 5 Min 50 002 ms Average 50 029 ms Q2 30 455 ms Q3 530 455 ms Q4 100 005 ms g5 Q6 Max 100 005 ms Entries Value Count Width 50 166 ms 1861 0 578 ms 100 094 ms 1 1 152 ms 2009 2014 RTaw 168 182 RTaW Sim V1 4 7 User Manuat The following is the histogram of the intervals below the quantile Q4 As can be seen it happens that one response time below the quantile is immediately followed by a value that exceeds the quantile Q4 A 1 12d Stay times below Q4 for frame4 2 5 Min 50 002 ms Q2 11827692000 ms Q4 21098518 000 ms Q Entries Walue 50 166 ms 149 764 ms 251 424 ms 351 081 ms 452 280 ms 550 057 ms 653 744 ms 750 598 ms 804 280 ms 851 937 ms O44 946 ms 1048 111 ms 1245 682 ms 1350228 ms 1397678 ms 1550 270 ms 1642179 ms 1750 572 mec Count 739 fo He 2 RP RPP PW Be Width 0 578 ms 1 725 ms 2 095 ms 4 042 ms 5 207 ms 6 332 ms 7 526 ms 8 642 ms 9 759 ms 9 808 ms 10 879 ms 12 067 ms 14 341 ms 15 545 ms 16 091 ms 17 848 ms 18 906 ms MAR mc Average 556501 519 ms Q3 21052942000 ms g5 Max 21098518 000 ms mW The histogram of the intervals below the quantile Q6 shows that at this level at least three consecutive response times are below the quantile Q6 2009 2014 RTaw 169 182 RTaW Sim V1 4 7 User Ma
119. he tool allows to perform Section 4 provides a reference manual that describes all the functionalities offered by RlaW Sim And finally Section 5 explains the basic functioning scheme of the simulation model underlying RTaW Sim 1 1 License of the software RTaW Sim is copyrighted by RTaW In this License the Product means the software product RlaW Sim The attached software product is provided as is without warranty of any kind either express or implied including but not limited to the implied warranties of title non infringement merchantability and fitness for a particular purpose No oral or written information or advice given by REALTIME AT WORK its agents or employees shall create a warranty and user may not rely on such information or advice 2009 2014 RTaw 5 182 RTaW Sim V1 4 7 User Manuat You may NOT resell charge for sub license rent lease loan or distribute the Product without our prior written consent We reserve the right to withdraw any such consent or part thereof for any reason and without notice and to demand that you immediately cease any activity in respect of which permission is withdrawn Software developers SDK are available for licensing in 3rd party software products You may NOT repackage translate adapt vary modify alter create derivative works based upon or integrate any other computer programs with the Product in whole or in part You may NOT use the Product to engage in or all
120. iStiCS ccccceceseececeseeeeeeeeeeeeeseeaeseeaeaneas 181 O RET SCC CS a ren cunideaansidatnienutcnenssmaeeerecouiseunieuss celts suneesecne nuts 181 2009 2014 RTaW 4 182 RTaW Sim V1 4 7 User Manuat 1 Introduction RTaW Sim is a timing accurate simulator of Controller Area Networks CAN that provides frame response time distributions and statistics about the frame buffer usage at the microcontroller and communication controller level RTaW Sim is able to simulate and predict the performances of CAN 2 0A CAN2 0B ARINC825 and CAN FD networks possibly interconnected through gateways with a very accurate modeling of the communication stack and communication controller RTaW Sim helps the designer compare the impact of different design alternatives choose the right communication stacks e g waiting queue policy and communication controllers e g number of buffers and configure them RTaW Sim enables the designer to also perform Simulation Based Fault Injection SBFI for instance analyzing the effects of clock drifts or the impact of transmission errors on transmission latencies Additional information about typical industrial use cases of RTaW Sim and how it relates to other temporal verification tools such as RlaW Tracelnspector trace analysis can be found in references 6 and 7 freely available from RTaW web site Section 3 offers six tutorials that allow to quickly get an insight into the kind of analysis that t
121. ialogOverCAN type New CommunicationPattern Sm DialogOverlAN ShortNamne Dialog StartBus CANL RequestFrame CANT REQ ResponseFrame CANI RESP Create Cancel A frame dialog is best defined by specifying first the bus where the Original sender is located followed by the identification of the two kinds of frames Notice that e only frames with the correct inherited role may be selected 2009 2014 RTaWw 120 182 RTaW Sim V1 4 7 User Manuat e In case of gateway crossing only the frames on the first bus need to be identified since the frames on other buses can be found by following the frame mappings of the gateways Frames that may be selected for the intended role appear in bold in the selection dialog Select RequestFrame a Frames CANI Pil CANI P2F CANI PIF CANI FC CANI CF CANI FF CANI RESP CANL REQ CANI Framel CANI Frame If for some reason the frame can not be selected an explanation is provided by the tool tip just position the mouse on top of the frame you would like to select Select Requeste jae S Frames TANI Pil TANI PAY TANI PIF TANI FC TANI CF TANI FF TANI RESP The Frame is already used as RequestFrame by some other DialogOverCAN 2009 2014 RTaW 121 182 RTaW Sim V1 4 7 User Manuat Finally you need to define at least one occurrence model which Specifies the
122. ible delay distributions must be defined at the level of the gateway 2009 2014 RTaw 132 182 RTaW Sim V1 4 7 User Manuat i LL RTaW Sim v1 4 2 Pro CA Users tiziana RTAW Sample Tutorial Gateway landrark O SiM lili NNN aE x File Samples Simulation Worst Case Optimization What If Plot Tools Architecture 2 Architecture bus_2 gt bus_1 amp Architectures ai Architecture Buses bus_1 bus_2 Ecus Ecu_1 Bus Gatel b FrameMappings Source Target b Frame_168 Frame_168 Frame_365 Frame_365 Frame_397 Frame_397 Frame_2de Frame_2de Frame_1d1 Frame_1d1 Name Source Ecu_1 bus_2 Target Ecu_1 bus_1 Show details of selected ProbabilityLaw Delete selected ProbabilityLaw Add new ProbabilityLaw i Specific Models if Results and Graphi amp s 4 6 4 Configuring a second run simulation Remember from Section 4 6 1 that the probability 10 6 to exceed the quantile Q6 means that 1 out of 1 million response times exceed Q6 but that this is true in the average In other words it is not impossible that for example 3 consecutive response times are larger than Q6 The purpose of a second run simulation is to collect statistics about the time intervals during which consecutive response times of a frame remain above or below the quantiles Since the quantiles are themselves obtained by simulation you first have to perform a normal simulation Then select the Second Run Simul
123. ic with all statistics entry Then choose the statistics taken after 15 minutes of simulation 2009 2014 RTaWw 21 182 RTaW Sim V1 4 7 User Manuat Som 4 Simulations 4 A System Simulation a BusSimulations a A CAN LS Stuffing 10 Default DOAS RandomComOffset RandomConstantDnfts 200ppm a BusStatistics 1m im 35 After a short delay the following graphic is displayed Statistics for A CAN LS Stuffing 10 Default DOAS5 RandomComOffsetO NoDrift 15m 4 ms 3 5 ms 1 25 ms 1 ms 0 75 ms 0 5 ms 0 25 ms ms 35 122 208 271 308 334 458 483 51 7 S60 586 618 697 743 782 870 907 940 972 1045 1398 Frames in increasing order of identifiers Minimum Average Q 1 10 2 Q 1 10 3 Q 1 10 4 4 Q 1 10 5 Q 1 10 6 Maximum The x axis represents the frames in increasing order of their identifiers displayed in decimal format and the y axis represents their response times in milliseconds Each point on a curve is the value of a certain statistic for a certain frame Here the five graphs are those corresponding to the statistics Minimum Average 2009 2014 RTaW 22 182 RTaW Sim V1 4 7 User Manuat Maximum and two quantiles 1 107 quantile and 1 10 quantile A 1 10 quantile is a threshold such that the probability that a response time of the frame is larger than that threshold is lower than 10 s
124. ight nevertheless be acceptable if the quantile is not close to the deadline For this reason quantile estimations based on more than 10 outcomes above the quantile are shown in black in order to indicate that they are robust those based on 1 to 10 are shown in dark gray see the illustrative table below If the sample size is smaller than 10 then the maximum is the only possible estimation of the quantile Qa This is a conservative estimation since no outcome is larger than the maximum the probability to be larger than the max is O in the sample which is smaller than 10 as required for a quantile In that case the estimation of the quantile is Shown in light gray It can only serve as a rough estimation of the quantile 2009 2014 RTaw 126 182 RTaW Sim V1 4 7 User Manuat Period Mini Dead TxOff Min Average Q2 Q3 Q4 Q5 g bla 10 0 0 148 0 247 1 031 1 456 1 750 1 811 1 823 10 5 0 218 0 318 1 130 1 560 1 854 1 919 2 013 0 572 0 779 1 986 2 280 2 500 2 648 2 662 0 500 0 671 1 524 1 832 2 056 2228 2 299 20 0 0 720 0 971 2 056 2 307 2 500 2524 2 524 20 10 0 702 0 901 2 009 2 200 2 500 2 588 297 20 10 0 236 0 375 1 524 2 104 2 529 2 679 2 658 20 0 0 962 1 310 2 529 2 005 2 971 3 006 3 026 20 10 0 720 1 002 2 178 2 558 2 141 2 805 2 033 40 0 0 112 0 291 1 750 2 443 2 741 2 689 10 5 0 166 0 261 1 143 1 578 1 897 2 080 2 218 40 0 0 166 0 346 1 730 2 360 2 838 2 961 20 0 1 168 1 591 2 838 3111 3 296 3 451 3 499 40 0 0 236 0
125. iguration or ComOffsetConfig determines the starting offsets of the ECUS communication tasks These parameters capture the fact that the nodes on a CAN bus will not start running exactly at the same time If no such configuration is specified then a random offset will be generated for each ECU with a random value between O and 1 second The Com Offset Generator Seed field displays the random generator seed used to generate the random offsets and allows to change it if needed e g to enable the user to reproduce similar experiments The resulting offsets can be viewed in the details pane of the generated ComOffsetConfig under the bus entity node The name of the generated configuration starts by RandomComOffset Notice that the unit of the communication task offsets is us A clock drift configuration specifies how the frequencies of the ECU clocks deviate from an ideal reference clock Notice that these clocks drive the periodic instantiation of frames and induce frame periods that differ to a certain degree from their nominal value Clock drift configurations can be defined as a ClockDriftConfiguration entity below a bus node in the Architecture tree If not provided they are generated before the Simulation based on the Drift Mode e NODRIFT the clocks behave perfectly without any drift 2009 2014 RTaw 130 182 RTaW Sim V1 4 7 User Manuat e CONSTANTDRIFT for each ECU a constant drift factor is ge
126. ing worst case response time bounds 2009 2014 RTaW 73 182 RTaW Sim V1 4 7 User Manuat Optimization What If Plot Tools Worst case response times bounds As for the simulation it contains a sub menu with an entry for every existing architecture Selecting for example S1 opens the configuration dialog for the architecture S1 Worst case response time calculus for 51 E a Buses Analyze bus busl busl busz 4 Gateways pee Offset Configuration Zero Bit Stuffing 10 ke For each bus you have to select the bit stuffing and the transmission offsets to be used in the computation ae Worst case response time calculus for S1 4 Buses Account for gateway traffic through busl gt bus2 z busl busz Frame delay distribution a Gateways bus gt bus niform L00 ps 300 ps i es at To For each gateway you have to select a Frame delay distribution The worst case analysis will use the maximum of this distribution When you have clicked on Yes to start the computation the following progress bar appears Le response times Computations of worst case 2009 2014 RTaW 74 182 RTaW Sim V1 4 7 User Manuat Notice e The worst case analysis assumes that communication stacks behave perfectly i e t
127. ingleShot kOrder QueuetBufferDelay The properties are explained in the table at the end of this section 2009 2014 RTaW 108 182 RTaW Sim V1 4 7 User Manuat In the SpecificConfigs table the default property set may be overwritten for specific bus interfaces In order to define a bus interface specific property set you have to right click in the table and choose the Add New menu entry CAN LS Default 25 CAN LS InterfacesConfig CAN L5 Default default ShortName Default DefaultConfig Default Interface Config SpecificConfigs BysConnection Queuing TxBufferCount UseHwCancellation SingleShot Order Queue ButferDelay Adapt column widths to contents Copy to Clipboard in CSV Format Show selected CANInterfaceSpecificContig Delete selected CANInterfaceSpecificConfig us New CANInterfaceSpeciticContig Then the following creation dialog appears New CANInterfaceSpecificConfig Sr BusConnection Click in the BusConnection field in order to specify the bus interface and then click Create As a result a new line appears in the table which corresponds to the newly created specific configuration which is initialized with the current values of the default configuration CAN LS Default amp CAN LS InterfacesConfig CAN LS Default default ShortName Default DefaultContig Default Interface Contic SpecificConfigs BusConnection Queuing TxBufferCount UseHwCancellation SingleShot Orde A
128. ion parameters such as frame offset strategies or the period of the middleware communication task when used It is very complementary to RIlaW Sim in the sense that it provides worst case results where RTaW sim provides average or distribution quantiles on the performance metrics For more details please consult the following web page http www realtimeatwork com software netcar analyzer IMPORTANT NOTE NETCAR Analyzer has been discontinued as a Stand alone product it is now exclusively distributed as a RlaW plugin that is available in the Professional edition 2009 2014 RTaw 62 182 RTaW Sim V1 4 7 User Manuat RTaW Sim is able to read legacy NETCAR Analyzer xml files choose NETCAR Analyzer xml file format in the import windows see Section 4 5 and to import the contained CAN bus ECUs frames offset configurations and the computed worst case response times the later can then be compared with the response times obtained by Simulation see ref 4 for a study showing the differences one can expect between simulation and analysis results IMPORTANT NOTE until the version 1 4 1 of NETCAR Analyzer the worst case response times must have been computed in the order of appearance of the corresponding offsets otherwise their correspondence could be incorrect in RlaW Sim 4 2 1 1 3 NETCARBENCH file NETCARBENCH is a GPL licensed software that generates automotive message sets according to a set of user defined parameters NET
129. ization What If Plot Tools Architecture A Help Html Architectures Help Pdf Architecture Help WCRI Pdf Buses About FCUS LomPatterns If this help does not provide the answer you are looking for or if you encounter a problem please refer to the FAQ or use the Helpdesk forum http www realtimeatwork com forum viewforum php f 6 If you need professional support or customized extensions please contact us at license realtimeatwork com 1 5 New releases and updates To be informed about new releases and update you should subscribe to our eNewsletter here http www realtimeatwork com subscribe 2009 2014 RTaWw 8 182 RTaW Sim V1 4 7 User Manuat 1 6 Changelog Version 1 4 7 Two options are provided for sorting frames in the unfavorable scenario Gantt charts by CAN id or grouped by sender first Added close button to simulation dialog Data dependency checks added on Frame attributes to avoid that analysis or simulation results get outdated with respect to the input data on which they are based Added Role column to CSV import so that frames used in frame dialogs and segmented transmissions can be imported not yet documented Quicker access to statistics on frame dialog and segmented tx delays added see Section 4 7 5 4 and Section 4 7 5 5 SOA offset heuristic replaced by SOPA heuristic which better handles priorities see Section 4 2 5 3 Possibility added to incrementally modify off
130. l E i NAY n AN s t AMA I 2 15 22 43 85 1031141161 361671781912 052192362873 3235837888 124284444846 1 06426626867 48858885 1383 11377 Frames in increasing order of identifiers 0 25 ms 0 ms Minimum Average Q 1 10 2 Q 1 10 3 Q 1 10 4 Q 1 10 5 Q 1 10 6 Maximum WCRT Statistics are estimations of parameters of probability distributions Randomness makes these statistics be more or less close to the actual value of the parameters they estimate but the higher the Sample size the closer are the estimates to the actual value For a periodic frame with a period P the sample size in a simulation of duration T is T P For example if T 1h and P 10ms then the size of the sample is 3600 s 0 01 s 360000 2009 2014 RTaWw 125 182 RTaW Sim V1 4 7 User Manuat In order to obtain robust estimations of a 1 10 quantile there Should be at least 10 outcomes greater than the estimation of the quantile several tens ideally This means that the sample should consist of at least 10 outcomes Quantile Recommended minimal sample Size Q2 10 t 10 1000 Q3 10 104 10 000 Qa 107 4 10 100 000 Qs 10 10 1 000 000 Qe 10 10 10 000 000 For the example of the periodic frame with period 10ms and 1h of Simulation time it means that the quantiles Q2 Q3 Q4 will be robust estimates Qs would imply only 3 outcomes above the quantile This m
131. l functionalities that allow to modify communication architectures in order to analyse the effects of certain evolutions Plot Tools Scale load Change identifier types to CAM2 0B Change frame types to CAN FD Group frames with same period CAN FD Scale payload CAN FD Consider modifying a copy of the original architecture see Section 4 3 1 2 for how to create such a copy this way you will be able to analyze and compare the modified communication architecture with the original one 4 2 6 1 Scale load This function allows to increase the periodic bus load of an existing bus by duplicating subset of frames while not changing the number of ECUs connected to the bus Consider applying this functionality to a copy of the original architecture see Section 4 3 1 2 for how to create such a copy This way you will be able to analyze and compare the higher loaded communication architecture with the original one When you select the menu entry the following dialog appears 2009 2014 RTaw 79 182 RTaW Sim V1 4 7 User Manuat Parameters for load scale Architecture Architecture Bus sige Bit Stuffing 10 Target Load 75 Keep identifiers Yes Cancel The configuration parameters have the following meanings e Architecture communication architecture where bus loads are scaled e Bus the bus where the resulting load should reach the Specified Target Load e Bit stuffing the bit
132. lation was started It has xml as Suffix For example sim cfg 2010 10 04 14 52 47 xmlL 2 in csv files see Section 4 8 3 Here however the sim_results folder is located in the same folder as the configuration file and not the original model file Furthermore the simulation specific sub folder has the same name as the resulting model file described in the previous item The simplest and less error prone way to create a configuration properties file is to use the File menu in the simulation configuration dialog after having specified the different simulation options see Section 4 6 3 2009 2014 RTaw 134 182 RTaW Sim V1 4 7 User Manuat Save Simulation Configuration ar onhe her bust pasting 0 bus 2 tomas Busintetaces Configuation HPF bus 1 gt bus 2 iiaii Offset Configuration Inter ECU COM Offset Configuration RandomComOftsett Com Offsets Generator Seed 30092011 Clock Drift Configuration RandomConstantDrifts 1000ppm Drift Mode CONSTANTDRIFT Drift Bound ppm Drift Generator Seed 1102007 Evert OccurencesModel SS O E Communication Pattern Models The resulting simulation configuration file can be stored anywhere because it contains the absolute path to the model file 2009 2014 RTaWw 135 182 RTaW Sim V1 4 7 User Manuat If the filename is provided as absolute path in the ModelFile property then configuration file can be
133. lows to change the type of all frames of a bus or of all buses of an architecture to CAN FD This is useful for e transforming CAN2 0A B configuration files into CAN FD configuration files e studying the effects of a migration to CAN FD on bus loads and frame response times When you select this menu entry then the following dialog is displayed Migrate to CAN FD Architecture Architecture Bus je Yes Cancel A First you have to select an architecture If you do not select a bus then the change is applied to all buses of the architecture otherwise only to the selected bus Notice e the CAN FD speed property must first be set for the bus es to migrate e this functionality does not increase the payload of the frames 4 2 6 4 Group frames with same period CAN FD This function allows to exploit the higher payload of CAN FD frames by grouping frames that 1 are sent by the same ECU 2 have the same period 3 either e none is forwarded by a bridge e all are also forwarded by a bridge When you select this menu entry the following dialog appears Groupe frames CAN FD Architecture Eiaa Aai a Wes Cancel 2009 2014 RTaw 82 182 RTaW Sim V1 4 7 User Manuat You have to select an architecture Notice e all frames of the selected architecture must have the type CAN FD e the name of the grouped frames is a concatenation of the names of the original frames
134. me of the ECU with the frame gateway function the bus where the frames are received the bus to which the frames are forwarded delay in microseconds between the transmission end of the source bus and the queueuing of the corresponding frame in the target bus interface name of the source frame name of the target frame DBC file DBC is a proprietary VECTOR Informatik Gmbh file format that is very widely used for describing CAN based communication systems RTaW Sim Professional edition includes a DBC file importer You need to select the Vector DBC file filter in the file selection dialog see Section 4 5 in order to import dbc files Then the following dialog is shown in order to provide additional information needed by the converter 2009 2014 RTaw 66 182 RTaW Sim V1 4 7 User Manuat Missing information For import Usage The Following informations need to be provided because they are not contained fin a standard way in the dbc File Bus speed in kbits s Name of the custom attribute that contains the Frame periods in milliseconds Select a predefined value or edit the field Name of the custom attribute that contains the Frame offsets in milliseconds IF there are no offsets in the file keep the default attribute name Bus speed ik bits s 125 Custom attribute for period SenMsgcycleTime Custom attribute For offset GenMsgStartDelayTime Ok Cancel First of all the bus speed needs to
135. nerated based on the Drift Bound and the random number generator seed The drift factor defines the actual speed of the clock where 0 8 means 20 slower whereas 1 5 means 50 faster A drift bound of 1 ppm means at most 1us faster or Slower every second lus second being 1 millionth of 1 second In other words the generated clock drift factors are randomly located in the range 1 10 6 1 10 6 Readers can learn more about the effects of clock drifts on response time distributions in 5 The Event Occurrence Model is optional See Section 3 4 and Section 5 3 for an explanation The Frame error model is optional See Section 3 5 and Section 5 4 for an explanation of frame transmission error models Communication Pattern Models are optional See Section 4 4 6 1 for segmented transmissions and Section 4 4 6 2 for frame dialogs In order to add patterns click in the field and the following configuration dialog will pop up Communication Patterns Occurence Models Hs Add iter Remove selected item Yes Cancel N Here you can select the occurrence models of communication patterns that are initiated on the considered bus Select the Length of Simulation according to the robustness of the statistics you wish to achieve see Section 4 6 1 If no times for intermediate statistics are specified then snapshots of the statistics are only taken at the end of the simulation
136. nuat A 1 12d Stay times below Q6 for framed si Min 150 008 ms Average 46150912 327 ms Q2 207190288 000 ms 038 Q4 Q5 06 Max 207190288 000 ms Entries Walue Count Width 149 764 ms 3 1 725 ms 3238 972 ms 1 37 289 ms 8325 440 ms 1 95 849 ms 2663221 000 ms 1 30661 123 ms 56260492 500 ms 1 67079 234 ms 127894580 000 rms 1 148452 703 ms 18854160 000 ms 1 217064141 ms 20436532 000 ms 1 235281 656 ms 29881830 000 ms 1 344023 500 ms 36762680 000 rms 1 423241 344 ms 37188376 000 ms 1 428142250 ms 38941008 000 ms 1 448319 969 ms 50746632 000 ms 1 5842355 688 ms 6243199 000 ms 1 718766 938 ms 66897084000 ms 1 F70172 688 ms 146354880 000 ms 1 1684954125 ms a 186383392 000 ms 1 2145794 250 ms MATATA NNN rnr 1 PRNNRI NNN ene 4 8 Exporting data Data can be exported in spreadsheet format see Sections 4 8 1 and 4 8 3 and graphics as images see Section 4 8 2 4 8 1 Tabular data export Right click anywhere in a table to make the context menu appear and choose the Copy to Clipboard in CSV Format 2009 2014 RTaWw 170 182 RTaW Sim V1 4 7 User Manuat TAN LS 25 ShortName Cah LS BusConfig OFFsetsDOa CrFsetlonfig OrrsetsD04 FrameErrorModel Sample Times Ss Frames anid 36 O24 59 Ox3b 73 Ox49 12 0x7a 135 0x87 196 Oxc4 196 Oxc 201 Oxc9 ane ft odin ShorkName Framess Framegs Frames Frame Framess framese Frannie 1 Framed Framen S
137. occur at a same discrete instant but a deterministic order that determines in which order these simultaneous events are actually executed must be provided In order keep the complexity of the simulation model low and thus also the time needed to run a simulation one needs to keep the event rate i e the number of event occurrences by time unit as low as possible This can be achieved by modeling certain actions that consists of several events and necessarily take some non zero time in the real system by only one simulation event which occurs instantaneously it takes zero simulation time The validity of such Zero simulation time simplification must be checked carefully to make sure that the behavior of the simulation model is sufficiently close to the real world system regarding the intended use of the Simulation results In Section 5 1 we explain how the nominal behavior of a CAN bus is simulated and in Section 5 4 we describe how transmission errors are taken into account 5 1 Nominal CAN In this section we describe the basic feature of the simulator which is the sending of periodic frames over a CAN bus while taking into account clock drift effects 5 1 1 Instantiation of periodic Frames Based on the local clock of the sending ECU see Section 5 1 4 clock drift a periodic alarm is used to schedule the instantiation of the Successive instances of a periodic frame The very first expiration of the alarm happen
138. og Backlog ResidualBacklog Residual Backlog LengthOfStaylnErrorWarning Length of Stay in Error Warning Results and Graphics Samulation InterErrorWarning Times Inter Error Warning Times LengthOtStaylnErrorPassive Length of Stay in Error Passive HTT ations D 4 A System Simulation J InterErrorPassiveTimes Inter Error Passive Times LengthOfStayInBusOtf Length of Stay in Bus Off 4 BusSimulations gt A Bus bus_1 Stuffing 20 0 Off a A Bus bus 2 Stuffing 20 Off 4 BusStatistics ald gt FrameRespTimes InterBusOffTimes Inter Bus Off Times gt FramelnterTransTimes a QueveStatistics Ecu_15 bus_2 Ecu_16 bus_2 Ecu_17 bus_2 Ecu_18 bus_2 Ecu_19 bus_2 Ecu_20 bus_2 Ecu_21 bus _2 Ecu_4 Gateway bus m segmentedTxRespTimeses DialogRespTimeses gt Th gt Oh gt SystemStatistics E Evaluations Graphics 5 4 nl t For the three states Error Warning Error Passive and Bus Off RTaW Sim gathers two kinds of statistics 2009 2014 RTaW 161 182 RTaW Sim V1 4 7 User Manuat e Length of stay e Time between successive entering in the state An empty field means that the state has never been entered In order to visualize a statistics click into a field LengthOfStayInBusOt Length of Staw in Rus OFF how details Delete Create New and select the Show details entry Ecu_4 Gateway
139. ommunicated for confidentiality reasons Potentially confidential names are those of e Architectures e Buses e Frames e ECUS e Communication Patterns 4 2 8 2 Generation of Deadlines This function allows to generate deadlines as percentage of frame periods This is useful when no individual frame deadlines are known but only some general rule about the latency constraints on the frame response times There are two variants of the functions e generation of frame deadlines that applies to the transmission on the specified bus e generation of timing chain deadlines that applies to end to end transmission through gateways from the source ECU until the reception by ECUs on the specified bus 2009 2014 RTaw 84 182 RTaW Sim V1 4 7 User Manuat Generation of Frame Deadlines Bus Percentage of period 100 Both kinds of deadlines can be visualized in the bus pane in the Deadlines column by selecting Bus Response Times or End to End Response Times Notice that for a frame that is not sent by a gateway the local deadline is always displayed whatever scope is chosen Sample Time Scope and Kind of Statistics End to End Response Tim ShortName Sender Recelvers Deadline 1 Frame 18e Ecu 3 5 Frame 194 Ecu 3 Ecu 4 P 5 Frame 19a Ecu 12 Ecu 3 P 10 5 Frame 19d Ecu 8 7 P 10 5 Frame _1bc Ecu_4 Gateway bus_2 Ecu_15 S 6 10 Frame _1be Ecu_4 Gateway bus_2 Ecu_15 6 E
140. ow others to engage in any illegal activity You may NOT transfer or assign your rights or obligations under this License to any person or authorize all or any part of the Product to be copied on to another user s computer You may NOT decompile disassemble reverse engineer or otherwise attempt to discover the source code of the Product except to the extent that you may be expressly permitted to reverse engineer or decompile under applicable law REALTIME AT WORK and any third party software vendor or provider Shall have no liability to users or any customers of users for any claim loss or damage of any kind or nature whatsoever arising out of or in connection with a the deficiency or inadequacy of the product for any purpose whether or not Known or disclosed to the user b the use or performance of the product c any interruption or loss of service or use of the product or d any loss of business or other consequential loss or damage whether or not resulting from any of the foregoing In no event shall REALTIME AT WORK or any third party software vendor or provider be liable to users or any customers of user for any special indirect incidental or consequential damages even if the user has been advised of the possibility of such damages 1 2 Advanced features RTaW Sim offers two categories of features e basic features available in all editions e advanced features marked with a only available in the Professionnal
141. p 69 times faster than real world system Statistics are also stored in the sub folder De Sott RTaW Sirn sim_results Tutorial Gateway landmark 0_001 SSS es In order to illustrate the convergence of statistics let us draw the convergence plot for the 1 10 quantile recall that for each frame the 1 10 quantile is the threshold that is exceeded on average by only 10 1 o0 transmissions of the frame In order to draw the convergence plot choose the corresponding menu entry as shown below Tools Create graphic with all statistics Create graphic showing convergence for a statistic Minimum Worst Case Scenario Average Observed Unfavorable Scenario Q 1 10 2 Qitslo 3 Q 1 10 4 O 1 10 5 O 1 10 6 Maximum and select the simulation result for bus 2 2009 2014 RTaw 52 182 RTaW Sim V1 4 7 User Manuat 7 v hd Select a busSimuleon cee M 2 4 Simulations a A System Simulation a BusSimulations A bus_1 Stuffing 10 HPF SOA25 RandomComOftset0 RandomConstantDrifts 1000ppm A bus_2 Stuffing 10 HPF SOA5 RandomComOffset0 RandomConstantDrifts 1000ppm See Cane On the resulting graphic we can see that with the exception of the black curve which corresponds to the very short sample time of 1 minute all curves are quite close to each other A bus_2 Stuffing 10 HPF SOA5
142. peAndKind Bus Respons Create Cancel and click Create This will bring up an empty plot Click on the Data tab at the bottom on the left side 0 05 0 00 t sia a F Graphic iy In order to add graphs right click anywhere in the table and select the Add New Curve entry 2009 2014 RTaw 44 182 RTaW Sim V1 4 7 User Manuat Effect of errors on maxima 25 ShortName Effect of errors on maxima Curves ShortName DisplayC Hidden Scop Adapt column widths to contents Edit selected Curve Delete selected Curve Add New Curve Copy to Clipboard in CSV Format Enter No errors aS name and select green as color ow Ts FrameStatCurve ShortName No errors DisplayColor BusSimulation Statistic 838 8 ScopeAndKindofTimes Sample Time r Then left click on the BusSimulation field and select the simulation that corresponds to the case with drift but without transmission errors uffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift No drift uffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift ppm uffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift lppm Events InFrequent uffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift RandomDrift lppm Events Frequentk uffing 10
143. peed Sim Conic ComOrrsets Contig ClockDriftConfiguration r Type Sender thee PEETS P Ecu_iz 5 F Ecu 12 5 P Ecu_1 5 E A ee ed Edit selected Frame Delete selected Frame Add New Frame Copy to Clipboard in CS Format z Feo 15 As a result the entire table is copied to the clipboard in CSV format with TAB as separator You can paste then the data to the Spreadsheet program of your choice Notice that this feature is available for all tables in RTaW Sim 4 8 2 Exporting graphics You can export a graphic as an image in PNG format For this purpose right click on the border of the graphic and choose the Save as entry of the context menu 2009 2014 RTaw Waximurn 525 mOffsetO Drift istic Maximum zoom In d zoom Out d Suto Range Properties Print 171 182 RTaW Sim V1 4 7 User Manuat The same context menu also allows to print the graphic to any printer installed on your system including virtual printers such as a PDF writer 4 8 3 The sim_results folder With each simulation is created a folder containing several csv files with statistics drawn from the simulation The following table describes these files The folder that contains these files is named like the RlaW Sim file with a sequence number as post fix and is located in a sub folder called sim_results which is itself located in the same folder where the RTlaW Sim file is s
144. presents the frames in increasing order of their identifiers displayed in decimal format and the y axis represents the maximum of their response times in milliseconds Each graph links the maxima of all frames for a specific sample time The three graphs are those corresponding to the sample times 5 seconds 1 minute and 15 minutes The exact superposition of these graphs is due to the perfect periodicity of the system e all frames are periodic e the offsets between frames sent by a same station aka intra ECU offsets are fixed e the offsets between frames sent by different stations aka inter ECU offsets are random but also fixed because we have ignored possible clock drifts between ECUs 2009 2014 RTaw 25 182 RTaW Sim V1 4 7 User Manuat The period of the system also called hyper period is the least common multiple of the frame periods Icm for short which is here equal to 2 seconds It also explains the first graphic the periodicity of the system allows a reduced set of different possible response times for each frame and in some cases only one value Notice however that a real CAN bus based communication system is not strictly periodic because of clock drifts that makes the inter ECU offsets vary over time To get an insight into the effects of clock drifts the reader is invited to perform the tutorial on clock drifts in the next section 3 3 Analyzing the effects of clock drifts Let us introduce the effect
145. rcles Save as Image When a new bus is created then the following creation dialog appears where a name and a transmission speed must be provided New Bus Bus ShortName busl Speed 125 kbits s Create Cancel When a new ECU is created then the following creation dialog appears where a name must be provided ShortWame ecul Cancel k Once the minimal required properties are provided and Create has been clicked the rest depends on how the creation has be initiated If the creation has been initiated from the exploration tree on the left then the details pane of the object is opened on the right as for example for a bus 2009 2014 RTaWw 95 182 RTaW Sim V1 4 7 User Manuat Architecture 4 Architectures 4 archi 4 Buses bust 4 Ecus archl ecul busl ShortName bus BitRate 125 CAN FD Bit Rate Periodic Load 0 0 BusSimConfig BusEvalConfig Sample Time Canld Cant ShortNa Sender Receivers Payl TxMo Ro Bit Stuffing 10 v OffsetConfig vr Y Scope and Kind of Statistics End to End Re Peri Mini Dea TxO Min Avera If the creation has been initiated form architecture graph pane then the node corresponding to the newly created entity appears at the point where the context menu has been invoked arch o3 Add new ECU Add new bus Layout a5 ircles Save as Image
146. reasonable default value 4 4 6 Communication pattern There are two kinds of communication pattern e Segmented transmission see Section 4 4 6 1 e Frame dialogs see Section 4 4 6 2 2009 2014 RTaWw 112 182 RTaW Sim V1 4 7 User Manuat 4 4 6 1 Segmented Transmission Since the maximal payload of a CAN frame is limited to 8 bytes the transmission of longer messages requires segmentation In order to analyze the response time of the transmission of long messages and also its impact on the response times of other frames RTaW Sim allows to simulate the repeated occurrence of segmented transmissions The behavior of the simulation of segmented transmissions is compliant with according to ISO 15765 2 The segmented transmission is based on 3 kinds of frames e First Frame e Flow Control Frame e Consecutive Frame These are the first entities that need to be defined in RlaW Sim See Section 4 4 2 on how to create frames and define their properties All these frames must have the TxMode set to Seg with the respective roles FF FC and CF The following table describes the frames that need to be defined if the segmented transmission is local to a bus Purpose of Payload TxMode Role frame First Frame Must not be Seg FF provided Flow Control Must not be Seg FC Frame provided Consecutive Must not be Seg CF Frame provided If the segmented transmission is crossing a ga
147. riod 2009 2014 RTaWw 158 182 RTaW Sim V1 4 7 User Manuat ShortName Relative Loads Style Relatively to Period ScopeAnd Curves R ShortName Jaci a a D Local bus 2 black hide Bus Response Times End 2 end bus 2 green hide End to End Response Times Network red show Network because we actually want to create a variant of the original graphic where the relative loads are shown instead of response times Statistics Relative Loads 25 Relative Loads D to T 4 74 144 240 323 392 465 J02 623 22 836 910 986 Frames in increasing order of identifiers Data Graphic 4 7 5 Viewing of other statistics In this section we show the statistics that are viewable in the Results and Graphics section of the GUI 2009 2014 RTaWw 159 182 RTaW Sim V1 4 7 User Manuat Specitic Models Results and Gra lms 4 7 5 1 Frame queue lengths Under the QueueStatistics node of a BusStatistics node double click on the node that corresponds to the ECU for which you wish to see statistics about the frame queue length p EETA Sox RTaW Sim v1 4 2 Pro C Users tiziana RTAW Sample Tutorial Gateway landmark 1 sim lE ae File Samples Simulation Worst Case Optimization What If Plot Tools Specific Models y BusInterface Ecu_2 bus_1 Backlog Backlog J Results and Graphics 2 ResidualBacklog Residual Backlog LengthOfStaylnErrorWarning 4 A bus_1 Stuffing 10 HPF SOA
148. rrence model described above and run the Simulation This time we will not use one of the predefined plots but we will create a plot by hand which will allow us to compare the maxima 2009 2014 RTaw 36 182 RTaW Sim V1 4 7 User Manuat of the frame response times with and without transmission errors For this purpose click on the Results and Graphics tab on the left Side then right click on the Graphics node and finally chose the Add New entry a Graphics I Stati Add New Mono Bus CAN LST Enter Effect of event triggered transmissions on maxima as name New Graphic ShorttName sions on maxima ScopeAndKind Bus Respons and click Create This will bring up an empty plot Click on the Data tab at the bottom on the left side 0 05 o 00 t Graphic iy In order to add graphs right click anywhere in the table and select the Add New Curve entry Effect of event triggered transmissions on maxima 2 5 ShortName Effect of event triggered transmissions on maxima Cures ShortName Display Hidden ScopeAn Bus Adapt column widths to contents Edit selected Curve Delete selected Curve Add New Curve Copy to Clipboard in CSV Format Enter No events as name and select green as color 2009 2014 RTaWw 37 182 RTaW Sim V1 4 7 User Manuat New Curve a FrameStatlurve ShortName No events DisplayColor green BusSirmul
149. rsrrrerrrrrrrsrrrrrrrrerre 140 4 7 3 Viewing the unfavorable scenarioO s sssssssssseerrrrrsserrrrrereerreno 142 4 7 4 Frame statistics visualized as graph S sssssssssrrrrsserrrrrrrssrren 143 4 7 4 1 Generation of frame response time QraphicCs ccceeee eee 144 4 7 4 2 Creation Of CUSTOM graph iCS ssssssesssrssrerserersrerrrrrresrrrrrrrenne 145 4 7 4 3 Usage of periods and deadlines in frame response time OAD CS eaae E E E A ates nee 156 4 7 4 4 Creating graphics DY COpy nNng sssssssssrsssrrrrrsserrrrrrrssrrrrrrrerne 157 4 7 5 Viewing of Other statiStiCS ssesssesseresrrrssrrrerrreerrrrrrsrrrrrrrrerne 159 4 7 5 1 Frame Queue lengthS ssssessssssrrrserresrrrsrrrrrrerrrrrrrrrererrrrreerne 160 4 7 5 2 Error Warning Error Passive and Bus Off related statistics 161 4A J 9 3 BUSY DeTIOG EM GUN Sinacrcccrsssransuenerpnnarenmtadeeevacunade escvrensane 162 4 7 5 4 Segmented transmission reSPONSE tIMES cccce cece esse ee ees 164 4 7 5 5 Frame dialog response tiMeS cccccccececeeeeeeeeeeeeeeeateseeaeeaneas 166 4 7 5 6 SECON run STATISTICS sssesserssrresrrrerrrrserrrsrrrerrrnerrrnerrrerrrrre 167 Ai EXPONO Gale eass Er EE E E E 170 4 31 Tapular Gala CX DOW vicecriendeisseannisnonmnecacaroninrvesnccrtrnmuntddaawniaenein 170 49 2 EXPONO Gi AICS taper rcissomawecinterresdadid A TEE 171 4 8 3 The SiIM_ results fOld EL reer nena nena n nea taees 172 SoMa TOO Casas vccsc
150. rst simulation When the Simulation is finished close the dialog as done before and turn back to the CAN LS tab In order to visualize the generated constant clock drift factors double click as shown below on the configuration called RandomDrift Lppm Architecture 4 Architectures 4 Architecture 4 Buses 4CANLS gt Frames gt BusConnections gt CanBusInterfacesConfigs OffsetConfigs 4 ClockDriftConfigurations No drift RandombDritt 1ppm to open the details pane 2009 2014 RTaw 28 182 RTaW Sim V1 4 7 User Manuat RandombDrift 1lppm 2 ShortName RandomDrift lppm DetaultDriftFactor 1 RandomGeneratorseed 1102007 DefaultDriftMode CONSTANTDRIFT ClockDrifts Ecu_l min of periods 50 ms 0 03 ppm Ecu_3 min of periods 100 ms 0 05 ppm Ecu_2 min of periods 100 ms 0 04 ppm Ecu_10 min of periods 100 ms 0 14 ppm Ecu_12 min of periods 100 ms 0 25 ppm Ecu_11 min of periods 100 ms 0 2 ppm Ecu_9 min of periods 200 ms 0 31 ppm Ecu_15 min of periods 200 ms 0 21 ppm Ecu_ min of periods 200 ms 0 34 ppm Ecu_0 min of periods 200 ms 0 24 ppm Ecu_4 min of periods 500 ms 0 8 ppm Ecu_13 min of periods 500 ms 0 27 ppm Ecu_ amp min of periods 500 ms 0 3 ppm Ecu_L min of periods 500 ms 0 33 ppm Ecu_14 min of periods 500 ms 0 57 ppm Ecu_ min of periods 500 ms 0 66 ppm Ecu_1
151. s CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOftset0 Drift Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOffset0 Drift Mono Bus CAN LS Stuffing 10 Offsets DOAS Com Offset RandomComOftset0 Drift 4 4 Definition of specific System Aspects In this section we describe how to define specific system aspects 4 4 1 Architecture An Architecture entity groups all information related to the description of a communication architecture An architecture is made up of buses ECUs and communication patterns which consists of frames and the communication stack related configuration parameter sets 2009 2014 RTaW 92 182 RTaW Sim V1 4 7 User Manuat Architecture a Architectures 4 Archi a Buses 4 Bus 1 Frames BusConnections CanBusinterfacesContigs OffsetConfigs ClockDriftContiqurations ComOftsetsContigs Bus 2 Ecus ComPatterns Bus_1 DialogOverlAN_1 Bus 1 5egmentedTx_1 Bus_liSegmentedTx_2 4 4 1 1 Duplication of Architecture Architecture entities may be duplicated see also Section 4 3 1 2 Architectures 4 Archi_1 a Bus Rename a Delete Create Copy iS
152. s at a time equal to the specified offset of the frame Subsequent expiration occur at a distance exactly equal to a multiple of the period of the frame in local time of the ECU 2009 2014 RTaw 173 182 RTaW Sim V1 4 7 User Manual When the alarm expires a frame instance is created and handed over to the communication stack COM stack for transmission see Section 5 1 2 The transmission delay starts from this instant see Section 5 1 5 for an overview Duration The instantiation takes zero simulation time 5 1 2 Downward traversal of COM stack Frames that are instantiated at the same time by the same ECU are Supposed to be handed over to the COM layer in HPF order i e the one with the highest priority first This assumption is irrelevant with a HPF software queue but important to know in case of a FIFO software queue When a frame instance is handed over to the COM stack for transmission the instance is stored in a hardware buffer or in the software queue e the software queue obeys the FIFO or HPF policy e the transmit buffers at the communication controller level obey the HPF policy when a new arbitration phase begins the controller always chooses the frame with the highest priority i e the lowest identifier The cancellation of transmission request at hardware buffer level is an optional feature of the CAN controller that may be configured as being used or not being used The algorithm for storing
153. s of clock drifts By clock drifts we understand the fact that the clocks of the ECUs which drive the periodic instantiations of the CAN frames do not exactly operate at the same frequency Due to production tolerances the oscillators are not exactly identical and their frequencies may also change over time because of environmental factors such as the temperature If you have performed the previous part of the tutorial just continue otherwise you can open the sample file that corresponds to the beginning of the second tutorial Simulation Worst Case Optimization What If Plot Tools Case Studies fe Tutorials Can Bus Tutorial CAN landmark 0 sim Csv Import d Gateways Tutorial CAN landmark 1 sim P Tutoral CAN landmark 2 sim Tutorial CAN landmark 3 sim Tutoral CAN landmark 4 sim Clock drifts can be modeled in sophisticated ways they depend essentially on the ambient temperature and the quality of the quartz but we will choose here a rather simple and widely applicable one based on fixed deviations of clock speeds positive or negative with respect to a nominal speed To avoid having to choose each clock speed individually RTaW Sim allows to generate them randomly in an user defined interval In the simulation configuration 2009 2014 RTaw 26 182 RTaW Sim V1 4 7 User Manuat dialog this can be configured by setting the Drift Mode to CONSTANT DRIFT and the Drift Bound to 1 for example As indi
154. s of the events that 2009 2014 RTaWw 35 182 RTaW Sim V1 4 7 User Manuat trigger transmissions The corresponding attribute is called InterOccurrenceTime and models the triggering events of all frames with type P E or E In case of the InFrequentEvents model such an event occurs every 100ms on average When an event occurs one frame is randomly chosen for transmission with equal probability among all mixed or event triggered frames if the FrameDistribution attribute is not set see Section 5 3 for more details about the modeling of event triggered frame transmissions a Simulate Length amp Sample Times Simulate the traffic on bus CAN LS Buses CAN LS Bit Stuffing 10 Gateways Bus Interfaces Configuration Default Offset Configuration DOS Inter ECU COM Offset Configuration RandomComOffsett Com Offsets Generator Seed 30092011 Clock Drift Configuration RandomDrift 1ppm Drift Mode CONSTANTDRIFT Drift Bound pprn 200 Drift Generator Seed 1102007 Event ccurrences Model InFrequentEwents Frame Error Model Start Pause Resume Let us now run a Simulation Make sure to select not only the same bus bit stuffing offset configuration and sample times as in the previous simulation see Section 3 1 but also the same inter ECU offset and clock drift configuration in order to ensure that the results will be comparable Then select the event occu
155. sed on their names If no corresponding entity can be found by name then the tool asks you to explicitly select one Finally the import is executed for all parameter sets that are available in the selected trace inspection 2009 2014 RTaw 71 182 RTaW Sim V1 4 7 User Manuat 4 2 1 5 Save Saves the currently opened model to its file If the model has never been saved to a file then a dialog allows to specify a file name The key combination Ctrl s is a short cut for this action 4 2 1 6 Save As Allows to save the model into a file with a possibly different name 4 2 1 7 Quit Shuts down the tool The key combination Ctril q is a short cut for this action If at the time of invocation the currently opened model has unsaved changes then the tool asks first if these should be saved 4 2 2 Samples This menu allows to open different sample files Simulation Wors Case Studies Tutorials b Csv Import The Case Studies are those of Reference 4 Simulation Worst Case Optimization What If Plot Case Studies d CAN HS WECS 2010 xml Tutorials CAN LS_WFCS 2010 xml Csv Import p B The Tutorials are used in Section 3 Samples Simulation Worst Case Optimization What If Plot Tools Case Studies B Tutorials Can Bus Tutorial CAN landmark 0 sim Csv Import Gateways Tutorid CAN landmark 1 sim Tutorial CAN landmark 2 sim Tutorial C4N landmark 3 si1m Tutorial CAN landmark 4 sim
156. set configurations and to generate offsets only for a certain ECU see Section 4 2 5 3 Round trip import of RtaW Tracelnspector estimated parameters see Section 4 2 1 4 4 Worst Case analysis extended to CAN FD Version 1 4 6 Enter in a dialog is like clicking on the Ok button DBC file import duplicated ENUM literal declarations in customer defined properties are now silently ignored Version 1 4 5 Estimation of remaining time displayed in simulation progress bar Short cut Ctrl C added to tables for copying of selected lines and Ctrl A for selection of all lines Menu merge import added import of typical Tx Error models with underlying probabilities laws see Section 4 2 1 4 2 Version 1 4 4 Improved topology layout in case of redundant buses 2009 2014 RTaw 9 182 RTaW Sim V1 4 7 User Manuat e Correction of a problem that occurs when only Intermediate Statistics are specified without an explicit Length of Simulation in the simulation dialog e Correction of a problem that hindered the visualization of unfavorable scenarios Version 1 4 3 e Correction of a file corruption problem when ECUs or buses are deleted Version 1 4 2 e Reorganization of user menus e Tabs in the tabbed panes can now be closed with the mouse second button or the mouse wheel e Tables can now be zoomed with Ctri Mouse Wheel e Panels that display topologies now automatically gain fo
157. sponse bus2 Payload of Dig RESP Frame response frame Copy of the busl Must not be B Must not be Response provided provided Frame on the since it is Since it is first bus inherited inherited Then you need to define at the level of the bus connections which ECUs send and receive these frames ECUL CANL es Architecture A Buses Ecu ECUL kaiak SentFrames ShortName Canld CanTyp Payload CAND ShortNam an Canlype Payload Ecus ECUT Adapt column widths to contents BusConnections Copy to Clipboard in CSV Format CANI Show selected Frame ies Remove selected Frame RecemvedFt PETs Add Frame NS The source ECU of the dialog sends the Request Frame while it receives the Response Frame The destination ECU receives the 2009 2014 RTaw 119 182 RTaW Sim V1 4 7 User Manuat Request Frame and the Consecutive Frame while it sends the Response Frame In case of gateway crossing do not forget to specify the sending and receiving at the bus connections of the gateway and also to define the frame gateways mappings in the two directions see Section 4 4 3 Next you need to define the corresponding communication pattern Architecture a CANI es a Buses Short iisk BusSimCe gt OAR a Ecus BusEval c ay Sample gt ECL gt GateWay Danii ComPattern Add New 5 f Oh 100 0x64 In the creation dialog for CommunicationPatterns make sure to select the D
158. srrrrrresene 76 Ao RARCON OSE ara E re etnnaedhnat canis 76 A DOR OR SE a E E EEE E 71 AO ANN a E A TE EEA 79 AOL CNE O ae E O 79 4 2 6 2 Change identifier types to CAN2 0B ccccccceceec eee eeeeeeeeeeaes 81 4 2 6 3 Change frame types to CAN FD sssssssssssresrrrrsssrrrrrrrssrrrrrrssene 82 4 2 6 4 Group frames with same period CAN FD ccccceceeeeeeeeeees 82 4 2 6 5 Scale payload CAN FD cccccccsseeesseeeeesteeesteeeateneaeeneanesneass 83 Aa POr EE 83 AO TOOL aaa E A 83 4 2 8 1 Anonymization of NaMeS ssssssssrresrrrserrrsrrrsrrrrsrrnrrrrrnrserrrrrnn 84 4 2 8 2 Generation Of DeadlineS s ssssssssssrrssrrrsrrrsrrrrrsssrrrrrrrssrrrrrrrsene 84 Ma FICO a E E E 85 Ao Dad eN r E EE E E sieeniaas 86 Aal CGO a O ne EE E 86 4 3 1 1 Creation from SCratCh sssssssssssnnsennnsennasnnnsennnenesnsonnnnasseerennnenne 86 4 3 1 2 Creation by AUPLICATION cc ceceecc cee eeeeeeeeeeeeeaeeeeaeseeataeeataneneas 89 Ai Saxe MOOC aO aE A ap nentatedrimatielennnn 91 4 4 Definition of Specific System ASPECUS ccccccceceseeeeeeeeeeeeeeeeneeeeaeas 92 A4 L ACCC CU C crrr nenna n REESEN NERE 92 4 4 1 1 Duplication of Architecture ssssssserresserrrrrreserrrrrrrsrrrrrrresene 93 4 4 1 2 Creation from scratch ssssssserssrrrsrrrrrrrrsrrrrrrserrrrrrrrserrrrrrnsrne 93 4 4 1 3 Entity detalls cc cecccceceseeceeeeeeeeseeaeseeeeaeeeeaesteeeanestanentaneaneas 99 4 4 1 4 Automatic LAYOUL
159. stuffing load to consider when computing the resulting periodic bus load e Target Load the periodic bus load that should be achieved on the target bus e Keep identifiers e if not checked then each frame copy will have an identifier one unit smaller than the original frame i e a higher priority than the original frame and to achieve this all frame identifiers are redefined e if checked then the algorithm does not modify any existing identifier and tries to assign an identifier one unit smaller than the original frame but if this is not possible because this chosen identifier is already used then the first free higher identifier is chosen Based on these parameters the algorithms applies a homothetic increase of the frame sets i e it rescales the frame set while keeping its characteristics similar to the original For example the proportion of frames with a period of 100ms will remain the same after rescaling If the original set contains 6 frames with period 100ms and if the load is increased by 50 then the generated set will contain 9 frames with a period of 100ms The algorithm proceeds as follows 2009 2014 RTaw 80 182 RTaW Sim V1 4 7 User Manuat 1 On the specified bus selection of a subset of frames uniformly distributed over the existing range of identifiers the number of selected frames is proportional to the load increase Bridging is treated as follows if a duplicated frame is the retransmission
160. t these are Excel files which need to be saved as CSV file with aS separator before the actual import but we have chosen the Excel format so that sections and keywords may be highlighted and the overall structure can be understood more easily The CSV format foresees two types of sections e required Bus sections for the description of a bus and the frames e optional Bridge sections for the description of frame gateways The meaning of the different parameters is as follows 2009 2014 RTaw 65 182 RTaW Sim V1 4 7 User Manuat Bus Name Speed Frames Name Sending ECU Receiving ECUs Identifier CANType Payload TxType Period MinDelay Offset Bridge Ecu Name Source Bus Target Bus Delay Frame Mappings Source Frame Target Frame 4 2 1 1 5 name of the bus speed of the bus in kilo bits per seconds name of the frame name of the Ecu that sends the frame comma separated list of names of ECUs that receive the frame on the same bus where they are produced trame identifier in decimal or hexa deciaml format STD CAN2 0A EXT CAN2 0B FD_STD CAN FD with standard identfiers FD_EXT CAN FD with extended identifiers number of data bytes P perodic E event driven P E mixed period in miliseconds minimal delay in miliseconds between two consecutive instances of the frame value of the transmission offset for the offset configuration identified by the header of the column na
161. t url http www realtimeatwork com page id 5 2009 2014 RTaWw 182 182
162. table gives an overview of the file formats and the ability to save back modifications File format Extension Open Import Save More details RTaW Sim sim xml Yes Yes Section 4 2 1 1 1 NETCAR Analyze xml Yes No Section r 4 2 1 1 2 NETCARBENCH xml Yes No Section 4 2 1 1 3 2009 2014 RTaw 61 182 RTaW Sim V1 4 7 User Manuat RTaW CSV CSV Yes No Section 4 2 1 1 4 DBC dbc Yes No Section 4 2 1 1 5 FIBEX Xml xml Yes No Section 4 2 1 1 6 AUTOSAR Xml xml Yes No Section 4 2 1 1 7 4 2 1 1 1 RTaW Sim file RTaW Sim files can be opened imported and saved Two variants are Supported e zipped xml file with the extension sim the default format e plain xml file with the extension xml Notice that the file selection dialog only shows files with the chosen extension Furthermore if you select for example an xml file to be opened or imported which is not a RlaW Sim file the tool will inform you and ask if other known formats such as the NETCAR Analyzer xml format should be tried RTaW Sim files can also be opened through the Recent Files entry of the File menu This sub menu shows all previously opened files except the one that is currently opened 4 2 1 1 2 NETCAR Analyzer file NETCAR Analyzer is a software tool that allows to compute worst case response times of CAN frames and configure communicat
163. teway then you have to create the corresponding frames also on the other buses You must also make sure that the frames which are sent by the gateway have the TxMode set to B and no Role specified the role is inherited in that case by the gateway frame mapping that will be defined later The following table describes the frames that need to be defined if the segmented transmission is crossing one frame gateway Purpose of Bus Payload TxMode Role 2009 2014 RTaw 113 182 RTaW Sim V1 4 7 User Manuat frame First Frame bus Must not be Seg FF 1 provided Copy of the bus Must not beB Must not be First Frame on 2 provided provided the second bus since it IS inherited Flow Control bus Must not be Seg FC Frame 2 provided Copy of the bus Must not beB Must not be Flow Control 1 provided provided Frame on the since it IS first bus inherited Consecutive bus Must not be Seg CF Frame 1 provided Copy of the bus Must not beB Must not be Consecutive 2 provided provided Frame on the since it IS second bus inherited Then you need to define at the level of the bus connections which ECUs send and receive these frames Architecture A ECUL CANI 2s Buses Ecu ECUT aie sentFrames ShortName Canld CanTyp Payload CAN nortNarm an CanType Payloa Ecus ECUT Adapt column widths to contents BusConnections Copy to Clipboard in CSV Format CA
164. th a smaller period has a smaller identifier and thus a higher priority this is called the Rate Monotonic priority assignment scheme Though Rate Monotonic is not necessarily the best policy with regard to meeting time constraints in practice it provides a sound basis to start with Notice that since the functionality modifies the priorities it might be useful to apply it on a copy of a reference Architecture see Section 4 3 1 2 for how to create such a copy This way you will be able to analyze and compare the reference architecture with the one with rate monotonic CAN identifiers When the menu entry is selected the following menu appears Rate Monotonic Identifier Assignment Architecture Architecture Bus x Wes Cancel If no specific bus is selected than all buses of the architecture are modified otherwise only the specified bus is touched 4 2 5 2 Random Offsets This functionality allows to generate random transmission offsets for a specified bus Offset generation configuration Architecture Architecture Bus bus 1 Offset Granularity 25 Wes Cancel o The granularity must be a common divider of the periods of all frames and all offsets will be a multiple of the granularity The smaller the granularity the smaller will be the resulting response times but more sophisticated algorithms like SOA produce much better offsets that lead to much smaller response times
165. tion In the simulation configuration dialog that shows up choose the Bus CAN LS and the Offset Configuration DOA5 DOA stands for Dissimilar Offset Assignment which is a simple offset assignment algorithm described in 2 Furthermore make sure to select NODRIFT as Drift Mode this will be explained in the next section 2009 2014 RTaw 17 182 RTaW Sim V1 4 7 User Manuat 3 Siruation of Architects O File Length amp Sample Times Simulate the traffic on bus CAN LS Z Buses cans Bit Stuffing 10 Gateways Bus Interfaces Configuration Default Offset Configuration DOAS Inter ECU COM Offset Configuration Com Offsets Generator Seed Clock Drift Configuration Drift Mode Drift Bound ppm Drift Generator Seed 1102007 Event Occurrences Model Frame Error Model T Communication Pattern Models Pause Resume Stop Select length amp Sample Times to specify the intermediate Statistic times File Length amp Sample Tin Buses R CAN LS Then enter Intermediate Statistics times i e time instance where snapshots of the evolving response time statistics will be taken 5s lm 15m 2009 2014 RTaW 18 182 RTaW Sim V1 4 7 User Manuat Length of Simulation d h m 5 ms Us Intermediate Statistics d h im 5 ms us 55 Delete Include bus off blocked Generate Trace Enter the value 5 in the fi
166. tored File Description Simconfig properties Contains all configuration parameters used for the simulation FrameRespTimes csv Contains for each frame a line with the characteristics of the frame and all Statistics FrameRespTimeDistrib Contains for each frame the histogram CSV data of the occurred response times with a certain bucket width BackLog csv Contains for each ECU a line with the reached maximum of the backlog and residual backlog of frames to send The backlog is the maximal number of frames waiting to be sent whereas the residual backlog is the maximal number of frames waiting to be sent just before the periodic communication task is instantiating new frames If the residual backlog is always zero and frames that are instantiated at the same time are instantiated in the order of their priorities then no inner priority inversion occurs see Section 5 1 3 even if the queue is a FIFO queue and no hardware cancellation is used 2009 2014 RTaw 172 182 RTaW Sim V1 4 7 User Manuat 5 Simulation model Discrete event simulation is based on the idea that changes in a system can be modeled by events that occur only at discrete moments in time i e moments whose number is finite in any interval of finite length The occurrence of an event modifies the State of the system and thus the simulation is driven by the Successive occurrence of these events Several events may
167. values of all the parameters that determine the actual behavior during a simulation 4 ComPatterns 4 CANI Dialog OccurrenceMode gt CANT Segmented Tx l ki Mew The parameters have the following meaning Parameter Meaning ShortName Name of the occurrence model DelayStart Specifies when exactly the measurement of the response time of an exchange Is started e FIRST FRAME_IX instantiation time of the request frame in the requesting ECU e FIRST FRAME_TX transmission start of the request frame sent by the requesting ECU ex OBD2 Period Repetition period of the dialog However the next exchange is only initiated after the reception of the response ResponseDelay Delay between the reception of a request frame and the instantiation of the expected response frame The following diagram illustrates the ResponseDelay of the ECUs Diagnostic Tool ECU Requ est Frame Response Delay Response Time First Frame of Response The property DelayStart specifies when the response time starts whereas it always ends with the transmission end of the response frame on the bus of the request sender 2009 2014 RTaw 122 182 RTaW Sim V1 4 7 User Manuat How to integrate a frame dialog into a simulation is explained in Section 4 6 3 How to visualize the response time statistics is described in Section 4 7 5 5 4 5 Open or import an existing model In Section 4 2 1 1 is descri
168. xists some Original frame produced by some ECU on Some other bus which has become the considered frame through the forwarding by frame gateways The End to End Response Time IS the delay between the instantiation queuing of the original frame until its transmission end on the current bus including the delays on the traversed buses and gateways For the ECU that receives the frame on the considered bus the end to end Response Time can be seen as the delay induced by the communication architecture network of buses If the frame is not sent by a gateway on the considered bus then the End to End Response Times is equal to the Bus Response Time it is a particular case of the general definition of End to End Response Time Inter Transmission Times Delay between successive transmission ends of the frame 4 7 2 Histogram view of statistics In the table view are displayed some statistics about the delays such as Min Average etc see Section 4 7 1 The corresponding histogram can also be visualized For this purpose right click on the line that contains the frame for which the histogram shall be displayed and choose Show Histogram 2009 2014 RTaw 140 182 RTaW Sim V1 4 7 User Manuat Canld ShortName Type Sender Receivers Paylo Period Mini 224 0xe0 Frame P Ecu5 Ecu 4
169. y type of gateways that is currently Supported is frame gateway 2009 2014 RTaWw 48 182 RTaW Sim V1 4 7 User Manuat which serves to forward complete frames and not signals across Sub networks the exact networking terminology would be bridges If you expand the node Ecu 1 then the node Gateways and finally double click on the gateway bus 1 gt bus 2 you can see the mapping of the frames from the source network to the target network It can be seen that the forwarded frames have the same names and identifiers on both buses but this is not necessary The attribute FrameDelayDistributions allows to define probability distributions for the gatewaying delay in microseconds it is the delay between the reception of the frame by the gateway on the source bus and the time where the corresponding frame instance is queued for being sent on the target bus It should be noticed that this delay does not include queuing and arbitration delays In our example we have chosen an uniform distribution between 100 and 200 microseconds ey RTaW Sim v1 4 2 Pro Model not yet saved File Samples Simulation Worst Case Optimization What If Plot Tools Architecture bus_2 gt bus_1 amp bus_1 bus_2 Ecus 4 Ecul FrameDelayDistributions Uniform 100 us 200 us BusConnections Name Source Ecu_1 bus_2 Target Ecu_1 bus_1 4 GateWays bus_1 gt bus 2 bus_2 gt bus_1 Ecu_2
170. y response time e the probability to be smaller than L is larger than p e the probability to be larger than L is smaller than 1 p For example the probability that a response time is larger than the 1 10 quantile is lower than 10 RTaW Sim allows to estimate the following quantiles Quantile Short Probability to be hand exceeded 1 107 quantil Q gt lt 102 1 e 1 10 quantil Q3 lt 10 0 1 e 1 104 quantil Q4 lt 10 0 01 e 1 10 quantil Qs lt 105 0 001 e 1 10 quantil Qs lt 10 0 0001 2009 2014 RTaw 124 182 RTaW Sim V1 4 7 User Manuat Notice that a probability of 10 to exceed the quantile Q means that 1 out of 10 1000 response times exceed Q but only in the average In other words it is not impossible that for example 5 consecutive response times are larger than Q3 In Section 4 6 4 we Show how RTaW Sim allows to collect statistics about consecutive response times above quantiles The rationale of this feature is to evaluate the extent to which successive large response times can occur possibly deadline misses which for instance may jeopardize the stability of control laws The following graphic shows how the statistics and the WCRT are Situated with respect to each other Min lt s Average lt Q2 lt Q lt Q Qs S Qs S Max lt WCRT Q2 Q3 Q4 Q5 Q6 Max WCRT Nm Nm vI 3 15 ms Py i
171. you then modify progressively 1 You may create for example a new configuration in the exploration tree 2009 2014 RTaw 77 182 RTaW Sim V1 4 7 User Manuat Architecture 4 Architectures 4 Architecture a Buses a CAN LS gt Frames gt BusConnections gt Can ysInterfacesConfigs a OffsetContig l DOAS Add New SOAS Shaping Aero ClockDriftConfigurations am mh with a default offset 0 in the example below New OffsetConfig k me ShortName Mixed Offsets Granularity 1 DefaultOffset 4 Or you may start to generate an offset configuration with identical granularity for all ECUs with the help of the Random Offsets or SOPA Offsets algorithm in order to obtain a basis to modify in the sequel 2 Then use again the the SOPA Offsets dialog in order to particularize the transmission offset for some ECUs Offset generation configuration Architecture Architecture Bus CAN LS Ecu_3 CAN LS Basis Mixed Offsets Modify Basis Offset Granularity 10 2009 2014 RTaWw 78 182 RTaW Sim V1 4 7 User Manuat Recall that if you check Modify Basis then the offset configuration will be modified each time but often it is also interesting to keep the intermediate configurations in order to analyses which step produced the highest impact 4 2 6 What lif This menu provides severa

RTaW-Sim User Manual - RealTime-at-Work

Contents

Download Pdf Manuals

Related Search

Related Contents