基于GA-MMAS算法的片上网络低能耗映射
Low Energy Mapping for Network-on-Chip Using GA-MMAS Algorithm
-
摘要: 针对片上网络的低能耗IP映射问题,提出了一种基于该混合算法的映射算法——GA-MMAS算法.该算法首先利用优先映射通讯量大的核的方法代替启发因子来确定启发信息,以改善最大最小蚁群算法 (MMAS)的最优解,然后将MMAS与GA结合,利用GA的快速性得到MMAS初始信息素,来弥补MMAS的前期信息素匮乏的缺点,提高了最优解的精确性,从而得到通信能耗更低的映射结果.实验表明,该映射算法与初始映射相比,能耗可以节省36%~60%,与已有的GA, AS和MMAS相比分别能节省3%~25%, 10%~30%和3%~30%的能耗.Abstract: Aiming at the problem of low energy IP mapping for Network-on-Chip,an algorithm named GA-MMAS is proposed by studying the hybrid algorithm which combines the Genetic Algorithm (GA) with Ant System Algorithm (AS).Firstly in this algorithm,instead of using heuristics to obtain the elicitation information in MMAS,we implement this behavior via priority mapping of IP core with larger communication volume,which improves the optimal solution of MMAS.Then with the combination of MMAS and GA,the advantage of speed in GA makes compensation to the lack of pheromone in the early stage of MMAS,and in turn enhancing the accuracy of optimal solution,which leads to lower energy consumption.Experimental results show that the algorithm saves about 36%~60% of energy consumption compared to initial mapping.Compared to GA.Ant System Algorithm (AS) and MMAS,the algorithm saves about 3%~25%,10%~30% and 3%~30%.
下载:
