云进化算法在NoC互连测试中研究

Research on NoC Interconnect Test in Cloud Based Evolutionary Algorithm

提出了一种基于云进化算法的NoC互连测试方案。该方案利用云模型对物种进化统一建模,重点解决云模型对进化的表示和对进化的控制两个问题,结合NoC互连测试问题,在功耗限制条件下,建立基于云进化算法的互连测试模型,以获取最优测试矢量集;实验结果表明:该算法取得了较好的测试效果,有效提高了测试效率。

This paper presents a solution using Cloud Based Evolutionary Algorithm （CBEA） to solve the Ne- twork-on-Chip （NoC） intereonneet test. With the eloud model, inheritance and mutation of the species can be modeled naturally and uniformly, the algorithm focuses on solving the two problems using the eould model to express the evolution and to eontrol the evolution. Considering of the NoC intereonnect test aetual problems and under the power constraints, this paper establishes the test mode of the CBEA, in order to find out the optimal test veetors set. Experimental results show that the algorithm aehieved good test results and improved the test efficiency.