多单播模型下的网络编码研究

Survey of Network Coding for Multiple Unicast Model

每个信源-信宿对之间只有一个商品流(Commodity)唯一地表示从信源到信宿的流量,该模型称为多单播模型.由于无线网络、P2P等应用均可看作基于多单播模型,所以如何在多单播模型下提升网络的性能成为研究的重点.网络编码近年来作为能有效提升网络性能的方法之一,其应用于多单播模型下的各种科学问题成为研究热点.本文研究多单播模型下的网络编码关键理论,主要内容包括网络容量区域、编码构造算法和联合编码等方面,最后讨论基于多单播模型下网络编码的研究前景.

Multiple unicast is a model which means a source-sink pair uniquely identifies a commodity to be communicated from the source to the sink.There are many network applications based on multiple unicast,such as wireless networks,peer-to-peer networks,etc.One important researching direction focuses on how to improve network performances under multiple unicast models.Presently,network coding under multiple unicast models,as one of good methods to improve network performance,becoms a hot research point.This paper summarized the key theoretical problems including capacity region of network,coding construction algorithms,and joint coding and so on.Finally,the future directions on network coding under multiple unicast model are discussed.