面向认知无线网络的频谱分配策略、随机模型理论及系统优化方法的研究

项目名称： 面向认知无线网络的频谱分配策略、随机模型理论及系统优化方法的研究

项目编号： No.61472342

项目类型： 面上项目

立项/批准年度： 2015

项目学科： 自动化技术、计算机技术

项目作者： 金顺福

作者单位： 燕山大学

项目金额： 82万元

中文摘要： 随着无线通信业务的飞速发展，网络运行环境愈发复杂多变，无线频谱资源弥足珍贵，认知无线网络应运而生。本项目致力于认知无线网络频谱分配策略相关问题的研究。引入感知门限、接入阈值、信道绑定及信道关闭等机制，自适应地控制频谱接入、数据传输及频谱切换，面向不同的网络应用，提出新的频谱分配策略。考虑非理想信道感知结果，建立一类新型优先级随机模型，并给出模型解析的理论框架，实现频谱分配策略的跨层数学建模与跨层性能分析。基于分治思想和梯度变化理论进行单一策略参数的优化，改进万有引力算法，进行多策略参数的联合智能优化；融合优先级随机理论和博弈论，揭示认知用户的纳什均衡现象并研究其社会最优问题，形成针对频谱分配策略的完整系统优化方法。本项目的研究意义在于秉承认知无线网络提高频谱利用率的理念，打破理想感知与负载饱和等假设的局限，充实动态频谱分配策略、跨层性能建模与随机系统优化理论，具有重要的理论和应用价值。

中文关键词： 认知无线网络；网络资源管理；频谱分配；随机模型；系统优化

英文摘要： With the rapid development of wireless communication, network operation environment is more complicated, wireless spectrum becomes a precious resource than before, cognitive radio network has thus emerged. This project is devoted to the research on spectrum allocation strategy in cognitive radio networks. By introducing schemes of sensing threshold, access threshold, channel bonding, channel closing, et al., the system admission, data transmission and spectrum handoff are controlled dynamically, and then a set of novel spectrum allocation strategies are proposed for different applications. Taking into account the imperfect channel sensing results for the licensed spectrum, a kind of novel priority queueing model is established, and a general framework for model analysis is constructed. The spectrum allocation strategy is evaluated by building a cross-layer mathematical model. Based on the divide-conquer method and gradient variation theory, the optimziation for single strategy parameter is given; With an improved gravitation algorithm, the joint intellegent optimation for multiple strategy parameters is carried out; Combining priority stochastic theory and Game theory, both the Nash Equilibrium phenomenon and the socially optimal strategy for cognitive users are investigated. Thus, a comprehensive system optimization method for the spectrum allocation strategy is developped. The significance of this project lies in adhering to the concept that the spectrum utility should be improved, breaking through the assumption restrictions of perfect spectrum sensing and saturated traffic load, enriching the dynamic spectrum allocation strategy, as well as the cross-layer system modelling and stochastic system optimization theory. This project has potential application in theoretical research and practical engineering.

英文关键词： Cognitive Radio Networks;Network Resource Management;Spectrum Allocation;Performance Modelling;Performance Optimization