基于FLS的多级队列CPU调度系统仿真研究  

A Multi-Stage Queue Scheduling Algorithm Based on Fuzzy Logic Symbols

作　　者：周思林[1] 谢从华[1] 

机构地区：[1]常熟理工学院计算机科学与工程学院,江苏常熟215500

出　　处：《计算机仿真》2015年第7期335-339,共5页Computer Simulation

基　　金：江苏省自然科学基金(BK2012209)

摘　　要：传统的多级队列调度方法为每个队列分配一个优先等级,并根据队列的等级将过程分配到不同的子队列中。高优先级队列中的过程通常会抢占低优先级队列过程的资源,使低优先级队列中的过程一直得不到服务。针对上述问题,利用模糊符号逻辑,提出了一种多级队列CPU调度算法。上述算法将就绪队列分为两个子队列,每个子队列包含I/O极限处理和CPU极限处理,且CPU时间动态地分配到每个队列中。设计了两种模糊干扰系统,一种系统动态地为两个子队列分配CPU处理时间,另一种系统计算时间量子来调度I/O极限过程。仿真结果表明,所提算法改善了过程饥饿问题,以及平均响应时间性能。In traditional multilevel queue scheduling approach,each queue is assigned a priority,and processes are assigned to different sub-queues according to their priority. Processes form the higher level queues which often seize the sources of processes from the lower level queues,which can cause the starvation of processes in lower queues. For this problem,a multilevel queue CPU scheduling algorithm was proposed based on fuzzy logic symbols.The algorithm divided the ready queue into two sub- queues,each of which contains I / O bound and CPU bound processes,and is assigned dynamically the CPU time. Two fuzzy interference systems were designed,one of which dynamically assigned CPU time to two sub-queues,and the other of which calculated the time quantum to schedule the I / O bound processes. Simulation results show that the proposed algorithm can improve the starvation problem and the average response time.

关 键 词：模糊干扰系统 调度 极限 模糊逻辑符号 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]