基于上下文认知机制的Ad Hoc网络性能优化策略管理模型的研究

项目名称： 基于上下文认知机制的Ad Hoc网络性能优化策略管理模型的研究

项目编号： No.61272454

项目类型： 面上项目

立项/批准年度： 2013

项目学科： 自动化技术、计算机技术

项目作者： 张沪寅

作者单位： 武汉大学

项目金额： 75万元

中文摘要： Ad Hoc网络是一种多跳无中心自组织无线网络，具有灵活性高、拓扑结构多变、QoS控制策略复杂等特点。Ad Hoc网络智能管理领域是目前学术界研究的前沿。本项目针对Ad Hoc网络策略管理模型，引入上下文认知机制，提出策略管理智能决策的体系结构，使网络具备学习能力，能够智能地改善Ad Hoc网络的维护效率和QoS性能。项目通过设计上下文知识模型，提出了基于节点相关性与概率统计的动态自适应分簇组网算法，在交互信息中，添加节点的相关性、消耗的能量及成为簇首的概率等因素构成分簇优化机制；同时构建具备上下文认知机制的策略管理模型，将策略管理系统分布化，使网络具备上下文感知能力；为了进一步验证模型的适用性，建立了合适的形式化描述方法，形成了完整的Ad Hoc网络性能评价体系，精确评价了新模型的QoS性能指标。通过项目的研究，能够提高Ad Hoc网络管理的灵活性和传输效率，促进策略管理模型的实用性。

中文关键词： Ad Hoc网络；上下文认知；分簇；路由；跨层优化

英文摘要： Ad Hoc Network is a multi-hop, decentralized and self-organized wireless network which has the characteristics of high flexibility, variability of topology, complexity of QoS control strategy, and so on. The field of intelligent management of Ad Hoc Network is at the forefront of academic research. Aiming at strategic management model of Ad Hoc Network, this project introduces context aware mechanisms and proposes the architecture of intelligent decision-making of strategic management to enable the network to have the ability of learning and intelligently improve maintenance efficiency and QoS performance of Ad Hoc Network. Through the design of contextual knowledge model, the project proposes node correlation and probability & statistics-based dynamic adaptive clustering algorithm, which adds the factors of node correlation, energy consumption, the probability of being cluster head to the interactive information to constitute the clustering optimization mechanism. At the same time, the clustering algorithm constructs strategic management model with context aware mechanisms and distributes the strategic management system to enable the network to have the ability of context-aware. In order to further validate the applicability of model, we establish an appropriate method for formal description to form complete pe

英文关键词： Ad Hoc Network；Context-awareness；Clustering；Routing；Cross-layer Optimization