Advanced Search
    Zhang Xiaoping, Duan Wuqing, Li Menghan, and Zhao Youjian. Constructing Direct Scalable Router with High Radix Router Node[J]. Journal of Computer Research and Development, 2013, 50(1): 218-224.
    Citation: Zhang Xiaoping, Duan Wuqing, Li Menghan, and Zhao Youjian. Constructing Direct Scalable Router with High Radix Router Node[J]. Journal of Computer Research and Development, 2013, 50(1): 218-224.

    Constructing Direct Scalable Router with High Radix Router Node

    • Scalable router is a very hot research spot for Internet core router, and the method which constructs scalable router with a direct network is a very important research direction in this area. However, most current researches concentrate on the method with low radix router node to realize the direct network. This method results in the direct network with high increasing diameter and low increasing bisection bandwidth. To solve this problem, we propose an idea of constructing direct network with high radix router node in scalable router. That is, on some packet latency condition, the radix of a router node in scalable router should be determined by the switch throughput of the router node. Then the direct network can get the lowest diameter and the highest bisection bandwidth, and router gets the highest switch throughput under this packet latency constraint. With high radix router node, the direct scalable router may get many different aspects on performance, such as switch throughput, packet latency, internal routing algorithm, load balancing, flow control and fault tolerance, etc. Here, we only study the switch throughput and packet latency, which are two very important factors on performance. We first give theory analysis on switch throughput and packet latency,comparing the change rule of these two factors. Then, we give the simulation results to verify the correctness of the theory analysis.
    • loading

    Catalog

      Turn off MathJax
      Article Contents

      /

      DownLoad:  Full-Size Img  PowerPoint
      Return
      Return