Stochastic modelling based environment for synthesis and comparison of bus arbitration policies

This paper discusses a methodology to automatically design bus arbitration policies, and presents a comparison of different arbitration policies. Stochastic policies like randomized stochastic policy, deterministic policies, piecewise deterministic policies as well as some heuristic policies have been explored. We also discuss the extraction of data from the simulation models which has been used while implementing some of these policies. Modification of the policy generation methods to incorporate issues like fairness, starvation and deadlock prevention has also been discussed. We also discuss arbiter generation for a network processor system.