RELIABILITY-ANALYSIS OF COMPLEX-MODELS USING SURE BOUNDS

As computer and communication systems become more complex it becomes i ncreasingly more difficult to analyze their hardware reliability, beca use simple models can fail to adequately-capture subtle but important features. This paper describes several ways we have addressed this pro blem for analyses based upon White's SURE theorem. We show: how reliab ility analysis based on SURE mathematics can attack very large problem s by accepting recomputation in order to reduce memory usage how such analysis can be parallelized both on multiprocessors and on networks o f ordinary workstations, and obtain excellent performance gains by doi ng so how the SURE theorem supports efficient Monte Carlo based estima tion of reliability the advantages of the method. Empirical studies of large models solved using these methods show that they are effective in reducing the solution-time of large complex problems.