基于改进量子进化算法的NoC测试规划研究

Research on Test Scheduling of NoC Based on Improved Quantum-inspired Evolutionary Algorithm

首先利用量子旋转门动态调整策略和并行跳跃基因对量子进化算法进行改进.在功耗约束条件下,采用重用NoC测试访问机制和XY路由算法,利用改进量子进化算法将测试数据分配到不同的TAM上寻找最优测试方案.最后以ITC′02基准电路作为实验对象对算法进行仿真验证.实验结果表明,经过改进后的量子进化算法对NoC测试规划优化明显,能快速收敛得到最优解,大大降低了系统的测试时间.

This paper proposes a method of test scheduling of NoC（Network-on-Chip,NoC）based on Improved Quantum-Inspired Evolutionary Algorithm.Firstly,improving the Quantum-Inspired Evolution Algorithm with dynamic quantum rotation gate strategy and jumping genes.Then inorder to find the best test scheme under the constraint of system limited power consumption,this paper presented that employing the IQEA for allocating the test data to the TAMs,by reusing the NoC as TAM（Test Access Mechanism）and XY routing algorithm.Finally,ITC′02benchmark circuits as simulation object,the experimental results show that the proposed method of IQEA can effectively optimize the test scheduling of NoC,obtain the optimal solution by rapid convergence and reduce the total test time greatly.