北京邮电大学学报

  • EI核心期刊

北京邮电大学学报 ›› 2011, Vol. 34 ›› Issue (5): 33-37.doi: 10.13190/jbupt.201105.33.jixd

• 论文 • 上一篇    下一篇

基于二分图最大赋权匹配的网络编码中继选择

纪晓东,谢信乾,彭木根,王文博   

  1. 北京邮电大学 泛网无线通信教育部重点实验室, 北京 100876
  • 收稿日期:2010-10-27 修回日期:2011-04-18 出版日期:2011-10-28 发布日期:2011-08-26
  • 通讯作者: 谢信乾 E-mail:xxmbupt@gmail.com
  • 基金资助:

    国家科技重大专项项目(2010ZX0300300301)

Network Coded Relay Selection Based on  Maximum Weight Matching

  • Received:2010-10-27 Revised:2011-04-18 Online:2011-10-28 Published:2011-08-26

摘要:

针对多用户多中继场景,为了进一步提升系统的吞吐量,需要为用户选择合适的中继协助其传输. 考虑到多址网络编码中继的中继选择问题是一个复杂的优化问题,为了降低其求解复杂度,将中继网络建模为带权二分图,中继选择最优解即转化为图论中求二分图最大赋权匹配问题. 分别将Kuhn和Munkres(KM)算法和贪婪算法应用于多址接入中继网络的中继选择,蒙特卡洛仿真结果表明,KM算法求解的遍历容量略高于贪婪算法.

关键词: 网络编码, 中继选择, 最大赋权匹配

Abstract:

In wireless multiple access relay networks, network coded relay can help two users transmit messages in the same time frequency resource without interference. Base station recovers the origin messages utilizing joint detection, and the achievable rate increases. In multiplerelay scenario, proper relay selection methods can further improve the overall throughput. Since relay selection is a complex optimization problem, to reduce the complexity, the network is modeled as a weighted bipartite graph and relay selection problem turns into the maximum weight matching problem. The Kuhn and Munkres (KM) algorithm and the greedy algorithm are utilized to solve the relay selection for multiaccess network, and Monte Carlo. Simulations show that the KM algorithm achieves a bit higher ergodic capacity than that of greedy algorithm. 

Key words: network coding, relay selection, maximum weight matching