认知无线传感器网络的关键组网协议设计与优化

With the fast proliferation of wireless technologies, the license-exempt 2.4GHz ISM band is becoming overcrowded. As a result, the transmission performance of the wireless sensor networks over 2.4 GHz ISM band cannot be guaranteed. Cognitive radio technology allows secondary users to access the licensed-bands while not disturbing the normal operation of primary users, which presents a promising direction to improve the transmission performance of wireless sensor networks. To address the inherent confliction between performance promotion via cognitive function and the requirement of low-overhead networking, this project will solve the following three key scientific problems: the design of channel rendezvous algorithms with the local information, the mode adaptation of the hybrid multimode networking protocols, and the cognitive-networking-modeling-based cross-layer optimization. To this end, this project will propose a computation over multi-access channel based distributed spectrum sensing protocol, spectrum-agile energy-aware hybrid MAC and routing protocols, and a mathematical-decomposition-based optimization method for protocol parameters. All of these contributions constitute the systematic design and optimization method of key networking protocols in cognitive radio sensor networks. Further, a GNUradio-based platform for physical experiments will be developed to demonstrate the efficiency of the proposed key networking protocols for cognitive radio sensor networks. This project will serve as a solid theoretical foundation for building self-organizing and energy efficient cognitive radio sensor networks with high transmission performance.

随着无线技术的快速发展，开放2.4 GHz ISM频段变得越来越拥挤，无线传感器网络的传输性能无法保证。认知无线电技术允许非授权用户在不影响授权用户正常通信的情况下接入授权频段，为提升无线传感器网络的传输性能提供了新的研究思路。针对认知性能提升和组网低开销之间的矛盾，本项目将解决基于局部信息的信道汇合算法设计、多体制混杂组网协议的体制调整机制设计和基于认知组网协议建模的跨层优化等关键科学问题，拟提出包括基于多址计算理论的分布式频谱感知协议、具有频谱和能量意识的多体制混杂MAC和路由协议以及基于数学分解的协议参数优化方法的认知无线传感器网络关键组网协议设计与优化方法。在此基础上，设计和开发基于GNUradio的物理实验平台，验证本项目所提出的认知无线传感器网络关键组网协议的有效性。本项目的成功开展将为设计自组织、低能耗、高性能的认知无线传感器网络提供坚实的理论基础。

认知无线电为提升传统无线传感器网络的传输性能提供了新的研究思路。考虑认知组网过程的低能耗、低开销、自组织等要求，本项目从“认知组网协议设计”、“认知组网优化”、“认知无线网络共存”三个层面系统地研究了认知无线传感器网络的协议设计与优化方法。. (1) 认知组网协议设计：针对受限带宽难以满足大规模、高并发无线传感器网络的实时、可靠接入问题，提出一套基于认知的无线传感器网络组网协议，具体包括频谱感知协议，认知MAC协议，认知时-空接入协议，认知分簇协议设计，认知路由协议。. (2) 认知组网优化：考虑资源严格受限认知无线传感器网络难以保持数据的安全、可靠传输的问题，提出一系列组网优化方法，具体包括实时传输调度算法，资源优化分配方法，中继部署策略，恶意设备检测方法。. (3) 认知无线网络共存：针对2.4 GHz ISM频段的多异构无线网络共存问题，提出了基于生物控制论的认知网络共存方法；并进一步针对一般性TDMA多网共存问题，提出了基于共存容限分析的资源分配方法。. 综上，项目所取得的研究成果覆盖了项目任务书的全部内容，在国内外一流学术期刊和主要国际会议发表论文40篇，授权和申请专利10项。