异构认知网络环境下的动态分级资源管理方法.pdfVIP

Vol.33 No .l 第 33 卷第 1 期 通信学报 2012 年 l 月 Joumal 00 Commuoicatioos January 2012 异构认知网络环境下的动态分级资源管理方法 文娟，盛敏，张攻 (西安电子科技大学 ISN 国家重点实验室信息科学研究所，陕西西安 71∞71) 摘 要z 针对异构认知网络中的资源管理问题，提出了基于认知的动态分级资源管理方法 (DHRM). 根据不同 时间尺度，引入小波神经网络、基于维纳过程的预测方法和增强学习算法获得业务分布变化、切换呼叫资源需求 量以及用户喜好等信息，从而动态调配异构多网络各级可用资源。在资源合理分配摹础上，根据各网络实时状态 以及用户喜好，通过多属性决策算法动态地将业务流分配到最佳接入网络中。仿真结果表明， DHRM 相对于网间 静态资源管理方法系统容量提高了约20%. 关键词g 认知网络:异构无线网络:无线资源管理:智能学习 中固分类号: TP393 文献标识码:A 文章编号: l 000-436X(2012)01-0107-07 Dynamic hierarchy resource management for heterogeneous cognitive network WEN Juan , SHENG Min , ZHANG Yan (The State Key Lab of ISN InfOl田酬。0 Scieoce Institu旬， Xidian Univcrsity,Xian 71∞71， China) Abstract: A dynamic hierarchy resow四 management approach-DHRM based on intelligent prediction was proposed for heterogeneous cognitive network. In DHRM , a四or，也ng to different time scale, the me由od of wavelet neural network, wiener prediction and reinforcement leaming were brought ωget the variation of t:r毡ffic distribution，由，e resource re- quirement of the handover calls, and the information of users preferen臼s， and available hierarchical resources of all networks were allocated flexibly. Multi-attribute decision making method, bωed on network status and user preferenω was used to make decision to dynamically assign network traffic flow to 也e most appropriate network. Simulation results show 由at， the system capacity is improved about 20