基于动态缓存的无线网络编码性能分析模型

Performance Model of Dynamical Buffer-based Wireless Network Coding

目前绝大部分网络编码模型是在假设节点缓存无限大的基础上建立的。结合IEEE802.11DCF信道随机接入机制,提出了一种基于动态缓存的精确分析无线网络编码性能的模型。该模型主要研究无线节点缓存有限及可变的情况下,网络节点平等接入和多优先级接入时的网络编码性能。通过构建嵌入式马尔科夫链模型对动态缓存进行数学建模,及离散二维马尔科夫链模型对DCF信道随机接入机制进行建模分析,从而得出一系列与网络编码性能密切相关的重要参数,如节点信道接入概率、数据包成功发送概率、平均编码数目、有效吞吐量等。结果表明,该模型精确评估了动态缓存和随机接入机制对无线网络编码的影响,为深入剖析无线网络编码特性和更好地运用网络编码提供了有力的依据。

Most of existing network-coding models are based on unlimited nodes＇ cache. Considering IEEE 802.11 random channel access mechanism, an accurate dynamic buffer-based wireless network coding model was proposed. The main idea is to compare the network-coding performance of different random access mechanisms, when nodes＇ cache is limited and metabolic. A Markov chain model and discrete two- dimensional Markov chain model were developed to characterize dynamical buffer and DCF （Distributed Coordination Function） random channel access mechanism respectively, to obtain important parameters closely correlated with the network-coding performance. The mathematical results show that the model accurately analyzes the influence of the dynamical buffer and DCF random channel access mechanism for the performance of wireless network coding, providing a powerful basis for the application of network coding.