【作者】 程波；

【导师】 杨宗凯；

【作者基本信息】 华中科技大学 ， 信息与通信工程， 2007， 博士

【摘要】 近年来,无线传感器网络(Wireless Sensor Networks, WSN)以其巨大的应用潜力吸引了众多国内外研究者的关注。然而,设计一个能够达到实际应用需求的无线传感器网络却面临许多挑战。无线传感器网络一般都是由一些能量受限的微小节点组成,因此其中一个关键的问题就是怎样设计一种高效节能的传输方式。协作式传输技术的出现为上述问题提供了一种新的解决方案,其中尤以虚拟MIMO(Virtual Multiple Input Multiple Out)引人注目。该技术允许多个单天线节点进行协作式信息处理,构成一个虚拟的天线阵列,以MIMO的方式完成信息的传输,可以极大地提高传输的能耗效率。作为一个完整的传输方案不可缺少的组成部分,接收机的性能对整个传输的性能具有重要影响。本文在前人研究的基础上,重点研究了无线传感器网络虚拟MIMO传输中,汇聚节点的接收机设计。当前的研究中,对于接收机的设计,存在如下问题:1.研究者一般假定无线传感器网络协作式MIMO传输的信道条件为准静态,并且只是应用最简单的接收机来考察传输方式的性能,没有完全体现出虚拟MIMO传输的优势;2.无线传感器网络信道条件不是准静态的情况在实际应用中很多,如利用高速飞行的飞机收集一片传感器监测区域数据的应用等,信道为快衰落。然而,当前还缺少对上述信道条件下的接收机设计。针对上述问题,本文设计了一种WSN中基于虚拟MIMO的统一传输结构,基于此结构,分别就准静态和快衰落信道条件下的接收机进行了设计与改进,取得的成果如下:1. WSN中基于多簇协作的虚拟MIMO统一传输结构提出了一种新的基于多簇的协作传输结构,同时具有STBC和V-BLAST协作方式的优点,可根据应用的不同需求特征,改变两种协作方式的偏重。针对这种传输结构,提出了一种较为通用的基于Gram-Schmidt正交化块QR分解的接收解码算法。该算法利用了等效的传输信道矩阵所具有的部分正交结构,降低了等效信道矩阵QR分解的复杂度。2.准静态信道条件下WSN虚拟MIMO的接收机设计提出了一种基于分块的球形解码的接收算法以及球形预处理方法,它把要求解的未知信号矢量分为多个块,然后对每一块利用球形解码进行求解,使准静态信道条件下的接收机在具有高BER性能的同时具有较低的复杂度,提高了汇聚节点解码的速度,降低了传输所需的能量,使参与协作发送的传感器节点以较少的能量即可完成任务。3.快衰落信道条件下WSN虚拟MIMO的接收机设计提出了一种基于盲接收设计的接收机,使基于虚拟MIMO的传输在快衰落信道条件下能够实现。该接收机利用了序列蒙特卡罗方法,能够根据接收的到信号全盲地进行解码,并且能够自适应地跟踪信道状况的变化。以上研究工作得到了如下项目资助:1.国家自然科学基金项目,”可重构MIMO的无线传感器网络信息处理与传输”,项目编号No. 605720492.国家自然科学基金项目,”自组织认知无线电网络关键技术研究”,项目编号No. 606020293.湖北省智能互联网技术重点实验室开放基金项目,”自组织认知无线电网络关键技术研究”,项目编号No.HSIT200605更多还原

【Abstract】 Recently, the Wireless Sensor Networks(WSN) attract many researchers’great inter-ests for its tremendous application potential. However, there exists many challenges todesign a WSN that can satisfy the real requirements of applications. Usually, the WSN ismade up from energy constrained tiny nodes. So, an one of key problems is how to design ahigh energy efficiency transmission strategy.Cooperative transmission technology gives a new direction to resolve the above issue,among which the virtual MIMO is the most attractive. It lets single antenna node cooperatewith each other, constituting a virtual antenna array, and transmit the data as the traditionalMIMO does. This kind of scheme greatly saves the energy of transmission.However, as a part of a complete transmission solution, the performance of the receiverhas significant effects on the whole performance of a transmission. Basing on the previouswork, this thesis lays focus on the receiver design at the data gathering node for virtualMIMO transmission in wireless sensor network.There are following issues in the current research on the receiver:1. The researchers usually assumes that the channel condition, over which the coop-erative MIMO transmission is carried out, is quasi-static, and only utilizes the mostsimple receiver to check the performance of a transmission strategy. The advan-tages of virtual MIMO transmission are not completely utilized.2. The applications, for which the channel are not quasi-static, is numerous, e.g. a?ying plane collects data from a wide monitor area, etc. In these applications,the channel are fast fading. However, currently, there are no receivers have beendesigned for this condition.After conceiving these issues, this thesis designed a uniform transmission form basingon the virtual MIMO, and for this form, the receiver design and optimization are discussedover the quasi-static and fast fading channel conditions, respectively. The contributionsinclude three aspect:1. Proposed a uniform multiple clusters based cooperation form. It combines the ad-vantages of the STBC and the V-BLAST cooperation schemes, and can be config-ured to lay particular stress on either of the schemes. To check the performanceof the new transmission scheme, an algorithm called Gram-Schmidt orthogonaliza-tion based block QR decomposition decoding is proposed. It utilized the orthog- onal structure of the channel matrix, and the complexity of QR decomposition isreduced.2. Proposed a blocking based sphere decoding algorithm and an optimized preprocess-ing. In the new decoding algorithm, the unknown signal vector is firstly split intomultiple blocks, and then, the sphere decoding is applied to each block. It makesthe receiver for quasi-static channel not only has high BER performance, but alsohas low complexity at the meantime. The complexity of the data gathering node isreduced, and the data collecting nodes can complete their tasks with less energy.3. Proposed a sequential Monte Carlo based Bayesian adaptive blind receiver. It canblindly work only with the received signals, and can adaptively track the channel.It makes the virtual MIMO transmission over fast fading channel become possible.The work is supported by:1. the National Natural Science Foundation of China,“Information Processing andTransmission in Re-configurable MIMO based wireless Sensor Networks”, No.605720492. the National Natural Science Foundation of China.“Key Technology of CognitiveRadio Network”,No. 606020293. the Foundation of Hubei Provincial Key Laboratory of Smart Internet Technology,“Key Technology of Cognitive Radio Network”, No.HSIT200605更多还原