基于自相似排队模型的片上网络缓存分配算法  

Buffer Allocation Algorithm for Network-on-Chip Based on Self-Similar Queuing Model

作　　者：韩国栋[1] 刘亮亮[1] 陈庆强[1] 

机构地区：[1]国家数字交换系统工程技术研究中心SoC实验室,郑州450002

出　　处：《计算机辅助设计与图形学学报》2012年第9期1226-1231,共6页Journal of Computer-Aided Design & Computer Graphics

基　　金：国家"八六三"高技术研究发展计划(2009AA012201)

摘　　要：为提升片上网络在自相似通信特征下的网络性能,依据片上网络数据流的自相似特性,通过建立虚通道自相似排队模型,提出一种基于自相似排队模型的片上网络缓存分配算法.算法首先利用已建立的排队模型计算虚通道缓存溢出概率,然后依次为溢出概率最大的虚通道缓存队列递增缓存深度,每次增加一个微片,直至达到总缓存配额阈值.实验结果表明,与现有的基于排队模型的算法相比,文中算法在相同缓存开销下可获得更小的平均包延迟,在同等平均包延迟下可节约25%的缓存开销.To improve the network performance of network-on-chip（NoC） under self-similar traffic process, by taking into account the self-similar traffic properties of NoC, a buffer allocation algorithm based on self-similar queuing model is proposed through the queuing modeling of the virtual channels under self-similar traffic. In the algorithm, the overflow probability of each virtual channel is first calculated through the queuing model. And then one flit buffer depth is added each time to virtual channels with the largest overflow probability till the whole buffer size reaches the threshold. The experimental results indicated that when compared to existing algorithms based on queuing model, the proposed algorithm can attain lower average packet latency with the same buffer consumption and save 25 % of buffer consumption as the average packet latency is the same.

关 键 词：片上网络 自相似排队模型 缓存分配 平均包延迟 

分 类 号：TP302[自动化与计算机技术—计算机系统结构]