Combined event & process simulation model of a distributed data collection system

Event and process driven simulation techniques are developed to assess loading and throughput in a multiprocessor-based distributed data collection system. Aspects of both simulation techniques are utilized to facilitate rapid model development while retaining flexible selection of configuration parameters. Statistics obtained for expansion capability and latency times for normal and priority players indicate bus traffic itself is not a limiting parameter unless process size exceeds local memory capacity. However, channel loading can cause significant delays unless properly balanced.