一种故障节点再利用NoC容错路由算法
发布时间：2019-08-03 13:25
【摘要】：构建一种微细粒度故障模型,有效地区分链路故障和通道故障,并在该模型的基础上提出一种基于故障节点再利用的容错路由算法.本算法由一个动态暂存机制(dynamic pseudo-receiving mechanism,DPR)和动态转向控制(dynamic turn management,DTM)构成,分别用于暂存数据包和避免死锁.最后实验结果表明,本文算法较先前的容错算法,时间延迟降低27%,吞吐量提高21%.尤其在数据包注入率增大时,本文方案有着明显的优势,保证了网络的可靠性
[Abstract]:A fine granularity fault model is constructed to effectively divide link fault and channel fault, and a fault-tolerant routing algorithm based on the reuse of fault nodes is proposed on the basis of the model. This algorithm is composed of a dynamic temporary storage mechanism (dynamic pseudo-receiving mechanism,DPR) and a dynamic steering control (dynamic turn management,DTM), which are used to temporarily store data packets and avoid deadlocks, respectively. Finally, the experimental results show that compared with the previous fault-tolerant algorithm, the time delay of the proposed algorithm is reduced by 27%, and the throughput is increased by 21%. Especially when the packet injection rate increases, this scheme has obvious advantages and ensures the reliability of the network.
【作者单位】： 国家数字交换系统工程技术研究中心;
【基金】：国家自然科学基金项目(61572520)资助
【分类号】：TN47
[Abstract]:A fine granularity fault model is constructed to effectively divide link fault and channel fault, and a fault-tolerant routing algorithm based on the reuse of fault nodes is proposed on the basis of the model. This algorithm is composed of a dynamic temporary storage mechanism (dynamic pseudo-receiving mechanism,DPR) and a dynamic steering control (dynamic turn management,DTM), which are used to temporarily store data packets and avoid deadlocks, respectively. Finally, the experimental results show that compared with the previous fault-tolerant algorithm, the time delay of the proposed algorithm is reduced by 27%, and the throughput is increased by 21%. Especially when the packet injection rate increases, this scheme has obvious advantages and ensures the reliability of the network.
【作者单位】： 国家数字交换系统工程技术研究中心;
【基金】：国家自然科学基金项目(61572520)资助
【分类号】：TN47
【参考文献】
相关期刊论文 前3条
1 张士鉴;韩国栋;沈剑良;柯t，
本文编号：2522588
本文链接：https://www.wllwen.com/kejilunwen/dianzigongchenglunwen/2522588.html
