一种基于网络编码的应用层多播算法  

An algorithm of application layer multicast based on network coding

作　　者：史玉琢[1] 郝琨[2] 

机构地区：[1]天津财贸管理干部学院,天津300170 [2]天津城市建设学院电子与信息工程系,天津300384

出　　处：《信息化纵横》2009年第17期42-44,48,共4页

摘　　要：研究了对多播网络进行网络编码的方法,提出了一种基于网络编码的应用层多播算法。该算法在计算网络拓扑时考虑了链路的花费。源端和中间节点使用随机线性编码方法进行编码,在目的端进行解码操作使得目的端能从乱序的信息和部分丢失的信息中恢复出原始数据,提高了网络的可靠性。通过对ns-2的扩展并进行仿真实验,结果证明了基于网络编码的应用层多播算法是可以提高网络的吞吐量,并且和网络中最大的吞吐量比较接近。在信息块不是很大的情况下,编码延迟率的增长是在一定的范围内的。This paper proposes an algorithm of application layer muhicast based on network coding. A minimum cost algorithm is used when the network topology is established. Source and intermediate nodes code based on random linear mechanism, sink node decodes form the disordered and losing message to restore original date, which raises the credibility of the network. From the simulation results analysis, the throughput algorithm on random network coding is very close to the theoretically most throughput. In the case of code-delay ratio increase is not very significant, the message is not very large.

关 键 词：多播 网络编码 吞吐量 

分 类 号：TN911.22[电子电信—通信与信息系统]