Hardware/Software Partitioning Algorithm Based on Hopfield Neural Networks and Genetic Algorithm
Oriented to hardware/software partitioning problem in embedded system and SOC, we compare the properties of Genetic Algorithm (GA) with Hopfield Neural Networks (HNN). A hybrid algorithm (GA_HNN) is puts forward based on HNN and GA innovatively. Directed Acyclic Graph (DAG) produced by the TGFF tool are used in hardware/software bi-partitioning as the mathematical model. Combined the fast search capacity of HNN with excellent global search capacity of GA, the algorithm can effectively avoid trapping in undesirable local minima. The experiments for DAG show that the algorithm has better performance than the genetic algorithm and HNN, and save more energy.
Genetic Algorithm Hopfield Neural Networks hardware/software partitioning embedded system system on chip
Xiaofeng Chen Bing Guo Yuansheng Wu Jihe Wang Yan Shen Yunben Liu
School of Computer Science & Engineering Sichuan University Chengdu, China School of Control Engineering Chengdu University of Information Technology Chengdu, China
国际会议
成都
英文
1631-1635
2010-12-17（万方平台首次上网日期，不代表论文的发表时间）