摘要
系统分析了多媒体通信卫星网络MAC协议中的资源分配流程及资源申请时序关系,在此基础上,考虑卫星网络中的长延时环境,提出了一种基于流量预测的资源申请算法,以弥补现有算法严重滞后于流量变化的不足.该算法引入记录因子,对资源的申请与分配进行记录,以提高带宽的分配效率,减少带宽申请信令的数量,从而减少对网络资源申请的响应时间;将最小均方算法用于网络流量的预测,以削弱卫星链路长延时对网络资源申请造成的负面影响.仿真结果表明,该算法的性能大大优于现有的资源申请算法,在高负载度、长峰值传输时间的条件下,算法具有良好的适应性.
The resource allocation process and time sequence of resource requests in MAC protocol for multimedia satcom are analyzed systematically. Then, considering the long transmission delay between GEO satellite and earth, a resource request estimation algorithm based on traffic prediction is proposed to make up for the traffic deficiency in existing algorithms which are lagging far behind the increasing traffic. Introducing the recording operator into the algorithm to record the resource requests and relevant distribution so as to improve the bandwidth allocation efficiency, the algorithm can decrease the number of request signaling messages to shorten the response time for the request for network resource. LMS (least mean square) prediction is applied to predicting the network traffic so as to weaken the adverse effect of time delay due to long satellite link on request for network resources. The simulation results showed that the algorithm proposed is greatly superior to other existing resource request algorithms in the adaptability under conditions of high loadability and long peak transmission time.
出处
《东北大学学报(自然科学版)》
EI
CAS
CSCD
北大核心
2008年第8期1091-1095,共5页
Journal of Northeastern University(Natural Science)
基金
国家自然科学基金青年基金资助项目(60602061)
国家高技术研究发展计划项目(2006AA01Z413)