摘要
受制于频谱资源有限性及链路负载差异性,网络拥塞成为认知无线Mesh网络研究中亟待解决的关键性问题.针对该问题,通过量化节点通信功率等级,并综合考虑网络干扰、链路有效容量及流量守恒等因素,建模了联合功率控制与信道分配的拥塞避免模型.进一步,提出了基于嵌套优化的拥塞避免机制,包括基于遗传算法的功率控制与信道分配、基于遗传算法的路由调度以及基于链路需求的最优路由算法.分别设计了组合编码和序列编码规则及流量守恒的约束控制机制,以保证个体进化的有效性及算法的快速收敛.一系列仿真实验表明该算法能够有效提高网络吞吐量,满足数据传输的实时性需求.
Due to the limited radio spectrum resources and the differences of link loads, network congestion becomes one of the key issues in cognitive radio wireless mesh network. For this rea- son, by quantifying transmission power levels, the effect of network interference, link capacity and flow conservation are considered comprehensively. As a contribution, the congestion avoid- ance model joint with power control and channel assignment is presented. And then, a nested op- timization technique is proposed to avoid the network congestion, which includes a genetic ap- proach for joint power control and channel assignment, a genetic approach for route scheduling and an optimal routing algorithm. In order to guarantee the individual validity and fast conver- gence, the rules of combinatorial coding and sequence-based coding are designed respectively with appropriate constraint control mechanisms. Extensive simulation results show that the algorithm can effectively improve the network throughput, and meet the real-time data transfer require- ments.
出处
《计算机学报》
EI
CSCD
北大核心
2013年第5期915-925,共11页
Chinese Journal of Computers
基金
国家自然科学基金项目(60903159
61173153
61070162
71071028
70931001)
中国博士后科学基金项目(20110491508
2012T50248)
中央高校基本科研业务费专项资金(N110404014
N110318001)
高等学校博士学科点专项科研基金(20070145017)资助~~
关键词
认知无线Mesh网络
拥塞避免
功率控制
信道分配
遗传算法
cognitive radio mesh network
congestion avoidance
power control
channel assign ment
genetic algorithm
