An Evaluation of Latency Time by Simulation for Distributed Multiplex System 910717
Currently both the SAE and ISO are working on the standardization of distributed multiplex transmission protocols, and the Japanese Society of Automotive Engineers (JSAE) is in the process of evaluating the performance and reliability of a number of communication protocols with a view to contributing to this work of standardization.(1)
This paper summarizes that aspect of the JSAE's communication protocol evaluation program that concerns response.
CSMA /CD with non-destructive arbitration systems that are used in motor vehicles resemble ordinary CSMA/CD systems in that transmission delay times vary in accordance with traffic. Thus when there is a stringent need for response, it is essential to make judgments on the basis of the probability in terms of delay time. It is, however, impossible to determine delay probabilities experimentally due to the problem of the number of test runs, and this creates a need for methods using simulation and analysis. It may also be noted that none of the simulation and analysis techniques developed for CSMA/CD heretofore takes up the issue of delay time probability.
The authors have therefore developed a method of simulation that establishes delay time probabilities for frames of any given level of priority under any given level of traffic. In order to allow measurements to be made rapidly, actual measurements of delay times under heavy traffic were made and compared with the results obtained by simulation, thus verifying the validity of the simulation model. Comparatively good agreement between simulation and bench tests was obtained, demonstrating the validity of the simulation model.
In addition a model using a simplified delay-analysis method (S-DAM) was developed and the results obtained were close enough to those from the simulation that it has been possible to use S-DAM as a guide in the design of ordinary in-vehicle LAN systems.