基于NEH算法的三值光学计算机任务调度优化研究  

Task scheduling optimization of ternary optical computer based on NEH algorithm

作　　者：庞昆 张杰 刘德方 张雨臣 王先超 PANG Kun;ZHANG Jie;LIU Defang;ZHANG Yuchen;WANG Xianchao(School of Mathematics and Statistics,Fuyang Normal University,Fuyang Anhui 236037,China;School of Computer and Information Engineering,Fuyang Normal University,Fuyang Anhui 236037,China)

机构地区：[1]阜阳师范大学数学与统计学院,安徽阜阳236037 [2]阜阳师范大学计算机与信息工程学院,安徽阜阳236037

出　　处：《阜阳师范学院学报（自然科学版）》2019年第4期44-48,共5页Journal of Fuyang Normal University(Natural Science)

基　　金：国家自然科学基金项目(61672006);安徽省质量工程项目(2018jyxm0507);阜阳师范学院创新团队项目(XDHXTD201703,XDHXTD201709);阜阳师范学院横向项目(XDHX2016021)资助

摘　　要：任务调度问题一直是计算机领域的一个研究热点,对新兴的三值光学计算机尤为重要。本文将加法器分解成五个独立的逻辑运算部件,分三步完成MSD加法,从而构成加法流水线,以充分发挥其巨位性与并行性的优势。选取响应时间作为优化指标,基于NEH算法对给定任务集进行调度优化,结果表明对于给定任务集采用短作业优先策略调度任务会使的平均响应时间最短。Task Scheduling is always a hot research topic in computer field.This is especially true for ternary optical computers.In order to give full play to its superpotential and parallelism,this paper decomposes the adder into five independent logic operation units and completes MSD addition in three steps,thus constituting the adding pipeline.The response time is chosen as the optimization index,and the given task set is optimized based on NEHalgorithm.The results show that the average response time is the shortest for a given task set when the task is scheduled under the short job-first policy.

关 键 词：流水作业调度 响应时间 单任务多数据 

分 类 号：TP315[自动化与计算机技术—计算机软件与理论]