Cloud-based Evolutionary Algorithm for Network on Chip (NoC) Test Scheduling using NoC as TAMs

fNetwork on Chip is a novel system-on-chip paradigm. Developing an efficient test strategy is essential for future NoC development. We propose to use cloud theory to solve core test scheduling problem for NoC using NoC communication network as test access mechanisms (TAMs). A cloud model is build and used by our evolutionary algorithm to optimize NoC testing time under power consumption constraint. The goal is to use cloud evolutionary algorithm to schedule testing of NoC cores as much in parallel as possible. Experiments show that the proposed algorithm converges much faster than the tradition evolutionary algorithm and can find testing schedules that reduce NoC testing time.