基于量子粒子群算法的SOC测试调度优化研究

Test scheduling optimization of SOC based on quantum-behaved particle swarm optimization algorithm

在基于IP核复用技术的SOC(system-on-ch ip,SOC)芯片中,测试资源的稀缺性限制了IP核并行测试的能力,导致了SOC测试耗时过长的局面。同时SOC测试时必须满足一定的功耗约束,否则会造成测试芯片的损坏。针对SOC测试时间与测试功耗协同优化这一难题,本文采用群智能优化算法-量子粒子群(quantum-behaved partic le swarm optim ization,QPSO)算法来实现这一目标。结合QPSO算法和测试调度问题,设计算法的适应度计算法则并建立测试时间与测试功耗的协同优化数学模型。通过实验确定算法中参数的最佳取值。最后利用算法搜索最优解确定IP核在TAM(test access m echan ism)上的分配,实现SOC功耗与时间的协同优化。经过国际标准SOC电路验证表明在解决功耗约束下的SOC测试调度优化问题上量子粒子群算法与已有算法相比,不仅能够更好的达到缩短SOC测试时间的目的,而且算法收敛速度快,需要调整的参数少,实现简单。

In SOC（system-on-chip） chips based on IP core-reusing technology,the parallel test capability of IP cores is limited by the deficiency of test resources,and the time cost is also considerable.Meanwhile,in order to avoid the damage of device under test,certain power constraints should also be taken into account.To solve the problem of collaborative optimization for time cost and power consumption in SOC test,a QPSO（quantum-behaved particle swarm optimization） algorithm is introduced in this paper.With the consideration of test scheduling,the fitness calculation rule for QPSO algorithm is designed,and the mathematical model for collaborative optimization is also established as well.Then,the best parameter values in the algorithm are decided with experiments.Finally,the assignment of IP cores on TAM（test access mechanism） is determined by searching the optimal solution with the algorithm,so as to achieve the collaborative optimization of power consumption and time cost.The test results with SOC test benchmarks show that compared with existing algorithms,the proposed QPSO algorithm can decrease time cost in terms of test scheduling optimization with power constraints,and also features higher convergence rate,less adjustable parameters and simpler realization.