Exact Best-Case End-to-End Response Time Analysis for Hard Real-Time Distributed Systems.

The prediction of best-case response time is quite necessary to guarantee control accuracy and stability of distributed hard real-time systems. Published time analysis techniques can precisely dealt with the precedence relations between tasks on different processors, but usually ignored or inaccurately estimated the effects from the precedence relations between tasks on the same processor. This leads to a pessimistic analysis. This paper presents an algorithm to analyze the best-case end-to-end response time for hard real-time distributed systems. This algorithm extends the traditional analysis technique by exploiting the precedence relations between tasks on both different processors and the same processor and achieves a more accurate analysis. Simulation results have shown that this algorithm can achieve more accurate results than existing methods. [ABSTRACT FROM AUTHOR]