基于业务预测的混合时隙分配算法  

Hybrid slot allocation algorithm based on traffic prediction

作　　者：田维珍[1] 郭丽[1] 余江[1] 

机构地区：[1]西安通信学院通信指挥系,西安710106

出　　处：《计算机应用研究》2014年第11期3438-3441,3476,共5页Application Research of Computers

摘　　要：为了提高战术数据链系统的时隙利用率,提出一种基于业务预测的混合时隙分配方法(traffic prediction based hybrid slot allocation algorithm,TPHSAA)。该算法将固定时隙分配与动态时隙预约机制相结合,使用Kalman方法预测各节点在下一时帧的业务量并计算所需时隙数,时间基准节点(network time reference,NTR)收集各节点时隙请求信息并结合其优先级进行动态时隙的预约分配。OPNET仿真结果表明,TPHSAA算法在时隙利用率、吞吐量、时延等各种性能方面明显优于传统的固定时隙分配方法,实现了时隙资源的动态调整和充分利用。To improve the performance of tactical data link （TDL）system,This paper proposed a new slot allocation algorithm called TPHSAA.This method added dynamic slot reservation （DSR）mechanism to the traditional static slot assignment,while priorities were taken into consideration.In the DSR part,each node employed Kalman prediction method according to its traf-fic model to compute the slot number that it would need in the next frame stage,then,sent the value to the NTR which execu-ted the slots assignment algorithm by summarizing all nodes’requirements and priority levels.The simulation results by OP-NET show that TPHSAA performs better in terms of slot utilization,throughput and end to end delay,rational allocation of net-work bandwidth resources is realized.

关 键 词：业务预测 动态时隙预约 服务质量 战术数据链 优先级 

分 类 号：TP393.04[自动化与计算机技术—计算机应用技术] TP301.6[自动化与计算机技术—计算机科学与技术]