泛在末梢环境下均衡多业务的终端聚合机制  

Device Composition Mechanism of Multi-service Equilibrium in Ubiquitous Stub Environments

作　　者：穆楠[1] 芮兰兰[1] 郭少勇[1] 熊翱[1] 

机构地区：[1]北京邮电大学网络与交换技术国家重点实验室,北京100876

出　　处：《电子与信息学报》2014年第9期2145-2151,共7页Journal of Electronics & Information Technology

基　　金：国家自然科学基金创新研究群体科学基金(61121061);国家自然科学基金(61302078;61271187);国家863计划项目(2011AA 01A102);教育部博士点基金(20110005110011);北京高等学校青年英才计划(YETP0476)资助课题

摘　　要：为了对异构终端进行有效聚合，解决泛在末梢环境中网络资源和终端能力有限性带来的多用户效益冲突分配问题，该文提出一种均衡多业务的终端聚合机制（DCMME）。首先以相对熵为均衡性指标设计了均衡服务质量效益函数，建立了面向多业务的终端聚合问题模型（MODC）；然后，该机制对问题模型进行降维简化，提出了基于均衡的终端聚合算法（EDC）；最后，通过C＋＋和MATLAB仿真表明，在该机制下，用户均衡性改善了0．5％～20％，由此证明该方法很好地权衡和保证了多泛在业务请求下各用户效益。In ubiquitous stub environments, the limited network resources and device capabilities are always causing the conflict of multi-user utilities allocation. To deal with this problem, a Device Composition Mechanism of Multi-service Equilibrium （DCMME） is proposed to make effective composition with heterogeneous devices. Firstly, based on the equilibrium index of relative entropy, the equilibrium service quality utility function is designed, and furthermore, the model of the Multiservice-Oriented Device Composition （MODC） problem is established. Then, this paper proposes to use dimensionality reduction of MODC in order to produce an Equilibrium-based Device Composition （EDC） algorithm. Finally, the simulation is implemented with C＋＋ programming and MATLAB. The results show that the DCMME improves the equilibrium of multi-user by 0.5%～20% and proves to perform well in balancing and ensuring each user＇s utility under multi-service circumstance.

关 键 词：泛在网络 泛在末梢环境 均衡多业务的终端聚合机制 信息熵 服务质量效益 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]