摘要
在认知无线Mesh网络中,由于信道状态变化导致的链路负载差异,网络拥塞成为影响认知无线Mesh网络系统性能的重要因素。针对这一问题,提出了基于最大最小公平策略的拥塞反馈算法。该算法通过综合分析基于随机搜索-遗传算法的多速率编解码调制、多重数据流的信道分配机制,以及优化的路由选择三种机制的网络资源分配约束条件,来构建跨层模型,计算网络拥塞。同时,通过拥塞值反馈,实现对物理层、链路层和网络层的联合跨层优化,最大程度避免网络拥塞。仿真结果表明,该算法在网络发生拥塞时收敛更快,能够有效避免拥塞,均衡负载,并能提升网络吞吐量。
Due to the difference of link load caused by the variation of channel status, congestion is an important factor for system performance in cognitive radio wireless mesh network. To solve this prob- lem, a congestion feedback algorithm based on max-rain fairness strategy is proposed. Through the com- prehensive analysis of the resource allocation constraints of multi-rate encoding/decoding modulation based on the random search-genetic algorithm, the channel allocation mechanism of multi-data-flow and the optimized routing mechanism, we build a cross-layer model and calculate the network congestion. Meanwhile, through congestion feedback this algorithm realizes the associated cross-layer optimization in physical layer,link layer and network layer respectively,and avoids congestion to the greatest extent. The simulation results show that this algorithm has better convergence when congestion occurs, and it can avoid congestion and improve the throughput.
出处
《计算机工程与科学》
CSCD
北大核心
2015年第5期925-929,共5页
Computer Engineering & Science
关键词
认知无线Mesh网络
拥塞反馈
跨层设计
随机搜索-遗传算法
最大最小公平策略
cognitive radio wireless Mesh network
congestion feedback
cross-layer design
randomsearch-genetic algorithm
max-rain fairness strategy.
