多数据源联合协作的Ad Hoc网络分布式最优资源分配算法  

Distributed Optimal Resource Allocation Algorithm Based on Mission of Cooperative Multi-Flows in Ad Hoc Network

作　　者：陈怡[1] 胡瑞敏[2] 高戈[2] 

机构地区：[1]华中师范大学计算机科学系,武汉430079 [2]武汉大学计算机学院,武汉430072

出　　处：《计算机学报》2011年第8期1478-1487,共10页Chinese Journal of Computers

基　　金：国家"九七三"重点基础研究发展规划项目"面向智能监控的视频编码方法"(2009CB320906);国家自然科学基金重点项目"移动音频技术理论与关键技术"(60832002);国家自然科学基金(60970160;61070080;61003184);国家科技重大专项"新型移动多媒体音视频编解码关键技术研发"(2010ZX03004-003-03);湖北省自然科学基金"基于关注度的分级音频编码技术研究"(2010CDB08602)资助~~

摘　　要：近年来基于Ad Hoc网络的智能化视频监控等新型网络应用逐渐兴起和普及,该类应用需要考虑多个数据源的联合协作.但是现有Ad Hoc网络的资源分配机制主要针对每个独立的数据流,在实现资源分配过程中没有考虑不同数据之间的依赖关系,导致针对联合任务的资源利用效率不高,成为制约网络服务质量提高的重要因素之一.文中首次把基于任务的评价参数纳入到无线Ad Hoc网络最优分配模型中,使得资源分配模型能够描述构成任务的各数据流之间的依赖关系.提出了基于任务利用率的动态资源调节算法(Mission utility based ResourceDynamic algorithm,MRS).证明了该算法满足非线性问题最优约束,所求数据源发送速率满足系统全局最优.仿真结果表明,提出的MRD算法能够提高多数据源协作Ad Hoc网络利用效率.Flexibility of Ad Hoc network flourishes many network applications, such as video surveillance and so on, with multiple flows and joint collaboration characteristics. However, existing algorithms rely on an independent data stream utility function, the result of competition for resources often does not meet the mission requirements of the joint distribution of bandwidth, causing the joint task inefficient. It is an important constraint to improve the quality of network services. We are the first to quantify the perception of mission-resource relationship, improving the precision of original model. A mission-based dynamic resource utilization adjustment algo- rithm （Based-on mission utility resource dynamic algorithm MRD） is proposed. The paper proved the global optimum of the algorithm. Simulation shows that this method can improve the Ad Hoc network efficiency.

关 键 词：协作 任务 效用 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]