基于改进模拟退火算法的片上网络任务调度优化研究

Research of Task Scheduling Optimization for Network on Chip Based on Improved Simulated Annealing Algorithm

随着半导体技术的不断发展, 传统的总线通信方式渐渐地不能满足片上系统的通信要求, 因此出现了将通信机制与计算资源分离的片上网络 （Network-on-Chip, NoC）.本文主要对片上网络任务调度的优化问题进行研究, 基于改进的模拟退火算法对调度过程中的关键问题进行优化, 利用编程实现优化算法, 并通过测试用例进行验证.

With the continuous development of semiconductor technology, the traditional bus communication gradual- ly can＇t meet the communication requirements of on-chip system, resulting in a network Network-on- a-chip com- munication mechanism and computing resources separation chip. This paper mainly studied the optimization of task scheduling in NoC. Based on the improved simulated annealing algorithm, the key problems in the scheduling pro- cess were optimized. The optimization algorithm was realized by programming and verified by test cases.