基于网络编码的组播率与编码节点数的平衡

Tradeoff between multicast rate and number of coding nodes based on network coding

为探究单源组播网络编码的组播率与最少编码节点数之间的关系,利用线性网络编码的导出与扩展技术,对两者间的关系进行了理论分析和推导,得出了"最少编码节点数随组播率单调递增"的结论。构造了一个多目标优化模型用于精确地描述两者间的数量关系。为求解这个多目标优化模型,设计出能搜索所有可行编码方案的策略。运用该策略,并结合NSGA-II,提出了求解该模型的算法。在需要兼顾两者平衡的情况下,模型的解为确定编码方案提供了选择依据。所提算法不仅能搜索出整个Pareto集,而且能在指定可行组播率区域的前提下,以较小的运算代价得出相应的部分Pareto集。仿真结果验证了理论分析的结论,表明了所提算法的可行性和有效性。

Based on single-source multicast network coding, in order to explore the relationship between multicast rate and the number of minimal needed coding nodes, by employing the technique of generation and extension of linear network coding, theoretical analysis and formula derivation of the relationship were given. It is concluded that the number of the minimal needed coding nodes monotonously increases with the increasing of multicast rate. A multi-objective optimization model was constructed, which accurately described the quantitative relationship between them. For the sake of solving this model, a search strategy was derived to search all feasible coding schemes. By combining the search strategy with NSGA-II,an algorithm for solving this model was presented. In the case of being required to consider the tradeoff between them, the solution of the model is the basis of choice for determining network coding scheme. The proposed algorithm not only can search whole Pareto set, but also search part Pareto set related with certain feasible multicast rate region given by user with less search cost. The simulation results verify the conclusion of theoretical analysis, and indicate that the proposed algorithm is feasible and efficient.