论文部分内容阅读
针对互连测试难题的分析,提出一种基于遗传算法的NoC互连测试方案。该方案采用NoC重用测试机制的方法,在功耗限制条件下,选取合适的测试端口和最短测试路径,同时根据互连测试中实际存在的问题,对算法进行适当改进,建立基于遗传算法的NoC互连测试模型,旨在获取最优矢量集的同时,测试代价更小。当NoC的规模达到一定程度时,采用划分测试方法,缩短测试路径,降低测试时间,提高测试效率。以SoCIN结构电路为仿真平台,分别对不同规模的NoC进行实验仿真。实验结果表明,遗传算法能快速有效地收敛到最优解,在测试运行代数及测试生成时间上取得了良好的测试效果。
Aiming at the problems of interconnection testing, a NoC interconnection testing scheme based on genetic algorithm is proposed. In this scheme, NoC reuse test mechanism is adopted to select the appropriate test port and the shortest test path under the limit of power consumption. At the same time, according to the actual problems existing in the interconnection test, the algorithm is improved properly and a NoC Interconnection test model, designed to obtain the optimal vector set at the same time, the test cost less. When the size of NoC to a certain extent, the use of divided test methods, shorten the test path, reduce test time and improve test efficiency. SoCIN structure circuit as a simulation platform, respectively, for different sizes of NoC experimental simulation. The experimental results show that the genetic algorithm can converge to the optimal solution quickly and effectively, and has achieved good test results in test operation algebra and test generation time.