基于网络编码的多播网络码字构造  

Construction of Codes in Multicast Network Based on Network Coding

作　　者：王静[1] 刘景美[1] 刘向阳[2] 王新梅[1] 

机构地区：[1]西安电子科技大学综合业务网国家重点实验室,西安710071 [2]西安通信学院军事综合信息网教研室,西安710106

出　　处：《北京邮电大学学报》2008年第4期98-101,共4页Journal of Beijing University of Posts and Telecommunications

基　　金：国家“863”计划”项目(2007AA01Z215);国家自然科学基金项目(60502046,90604009,60573034)

摘　　要：针对网络编码技术,提出了一种存在链路故障多播网络的码字构造方法.该方法通过对接收节点搜索大于信源发送速率k的n条离散路径,确保接收节点任意k条输入链路的全局编码向量线性无关,接收节点收到k个正确符号,避免链路故障,恢复信源信息.利用最大距离可分码已有的成果,给出不同发送速率下信源所需的最小有限域,构造最大距离可分码(n,k)生成矩阵,将其列向量作为接收节点输入链路的全局编码向量,设计网络码字,实现网络编码.Based on network coding, a method for failure is proposed. Specifically, n disjoint paths （ cast transmissions can work well even at most n-k constructing codes in multicast networks with link n is larger than the multicast rate k so that multi- link failures occur） are selected for each receiver, and the global coding vectors of any k input links should be linearly independent. Then the receivers can obtain k right symbols, and avoid link failure. By using some results available of maximum distance separable（MDS）codes, the minimal finite field is obtained, and is enough to implement network coding for different multicast rate. To design codes of multicast network and implement network coding, a generator matrix of a （n, k ） MDS code is constructed based on the obtained minimal finite field, and its column vectors are allocated to input links of receivers as their coding vectors.

关 键 词：网络编码 链路失败 最大距离可分码 有限域 

分 类 号：TN911.2[电子电信—通信与信息系统]