A NEW METHOD FOR RELIABILITY ANALYSIS OF EMBRYONIC ELECTRONIC ARRAY

Most existing reliability model of embryonic electronic array (EA) can only compute the reliability in hot standby. The model of series system with standby parts (SSSP) is proposed, which can compute the reliability in cold, warm and hot standby by Markov chain. We further design the reliability models of EA by SSSP with two self-repair reconfiguration strategy in three standby modes, such as single cell elimination strategy or column elimination strategy. Experiments are conducted on EAs with random configuration. The reliability of EA is varied with both the failure rates of cells and the spare columns of array. The performance of SSSP is compared with that of Monte-Carlo method (MC). The results show that, SSSP is similar to that of MC for EA with single cell elimination strategy or column elimination strategy. All the results of SSSP are in the confidence interval of the results of MC, where the confidence level is 95%. The time complexity of SSSP is O(N), which is lower than that of MC. Compared with traditional reliability model of EA, SSSP is applicable to EA in cold, warm and hot standby.