基于遗传算法的低功耗有限状态机状态分配  

Genetic Algorithm Based Low Power FSM State Assignment

作　　者：叶锡恩[1] 夏银水[1] 陶伟炯[1] 王伦耀[1] 

机构地区：[1]宁波大学电路与系统研究所,宁波315211

出　　处：《计算机辅助设计与图形学学报》2006年第12期1861-1866,共6页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家自然科学基金(60676017);浙江省自然科学基金人才专项(R105614);浙江省教育厅项目基金(20051732)

摘　　要：提出一种通过状态分配来实现有限状态机的功耗和面积同时优化的方法·在分析现有成本函数的基础上,提出了一个新的成本函数,并利用遗传算法能进行多目标优化的能力来实现功耗和面积的同时优化·该算法用C语言实现,并对17个MCNC有限状态机标准电路进行测试·测试结果表明,与已有的功耗优化算法相比,文中算法在功耗和面积方面有一定的优势·A state assignment based algorithm is presented for simultaneous power and area optimization of finite state machines. Based on an analysis of the existing cost functions, a new cost function is proposed, By using genetic algorithm, which has the capability of multiple objective optimization, simultaneous optimization of power and area is implemented, The proposed algorithm is applied to 17 MCNC benchmarks. The results show that the proposed algorithm is advantageous over the existing power optimization algorithms in terms of simultaneous power and area optimization capability.

关 键 词：有限状态机 低功耗 遗传算法 状态分配 

分 类 号：O157.4[理学—数学]