最优线性网络编码的分布式构造方法

Distributed construction of optimal linear network coding

针对单源组播网络,在合理的假设条件下,提出了最优线性网络编码(在吞吐率最大的前提下编码代价最小)的分布式构造方法。首先对线性网络编码的机理进行分析,得出了不同组播率下编码方案间的内在联系,导出了一个在线测试组播容量的策略;提出的方法以遗传算法为搜索引擎,结合随机线性网络编码,嵌入了在线测试组播容量的策略。通过合理地设计编码系数的生成规则和进化策略,在获得最大吞吐率前提下达到最小编码信道数的同时,构造出相应的编码方案,且编码系数保存在各节点中,从而可以采用确定性网络编码数据传输策略传输数据。仿真结果表明提出的方法是有效的。

Aiming at the single-source multicast network, based on some reasonable assumptions, a distributed network coding construction approach is proposed, which can obtain the optimal coding scheme with maximal throughput and minimal coding overhead. By studying the intrinsic mechanism of linear network coding, the mutual relationship between two coding schemes at different multicast rates is found out, and a strategy is contrived to test multicast capacity with distributed way. The proposed approach uses a genetic algorithm as a search engine, combines with random linear network coding and embeds the strategy of distribu ted test multicast capacity. With the effective evolution strategy and the rule of generating coding coefficients, it can not only achieve the minimal coding channels with maximal throughput, but also construct the correspond- ing coding scheme whose coding coefficients can be stored in respective nodes, so that the data can be transmitted by a deterministic transmission method. Simulation results indicate that the proposed method is effective.