认知无线电论文：认知无线电中基于簇二次加权合作频谱检测算法研究.docVIP

认知无线电论文：认知无线电中基于簇的二次加权合作频谱检测算法研究 【中文摘要】频谱检测作为认知无线电的核心技术之一,可以发现没有被主用户占用的空闲频谱,是认知无线电实现的前提和先决条件。目前,频谱检测已经有很多研究成果,主要集中在两个方面：单用户本地频谱检测与多用户合作频谱检测。单用户检测由于受到多径衰落和遮蔽的影响,会产生隐藏终端问题,使得整个认知系统的检测性能降低；而合作检测虽然可以有效解决隐藏终端问题,显著提高系统的检测性能,但因通信量增大而增加了系统开销。本文针对上述问题,提出一种基于簇的二次加权合作频谱检测算法。本文首先综述了课题的和国内外研究现状,然后介绍了各种单用户频谱检测算法并比较分析,着重研究了能量检测算法,并对其在高斯信道和瑞利衰落信道下的检测性能进行了仿真分析。针对单用户频谱检测的局限性,研究了多用户合作频谱检测算法。对合作检测中基于软、硬两类数据融合算法的几种常见合并准则进行了仿真分析,并对传统硬合并数据融合算法提出两种改进方案,以“或”准则为例仿真分析改进方案的有效性,仿真结果表明改进方案能获得更好的检测性能。本文在合作检测中引入分簇和加权的思想,提出了一种基于簇的二次加权合作检测算法。第一次加权是在簇内,采用等权重的软合并数据融合方法;第二次加权是在簇外,采用动态加权的硬合并数据融合方法。在瑞利衰落信道下,对该算法进行仿真验证。仿真结果表明,相比传统的合作检测算法和分簇算法,这种基于簇的二次加权算法检测性能得到了明显提高。而且对那些信噪比(SNR)比较低的次用户,可以通过降低它们的权重因子来减少其对最后判决结果的不利影响,从而节省了系统的开销。 【英文摘要】One of the key technologies for CR is spectrum sensing, which can find the vacant frequency band that is not currently occupied by the primary users (PU). At present, there is a lot of research about spectrum sensing technology, mainly in two aspects: the single-user local spectrum detection and multi-user cooperative spectrum detection. Due to multipath fading and shadowing,single-user detection has hidden terminal problem, which makes the performance of cognitive system degrade; cooperative detection can significantly improve the detection performance of cognitive system, which can overcome multipath fading and shadowing, but it increases the cost of cognitive system. In this thesis, to address the above problems, a twice-weighted cooperative spectrum sensing algorithm based on clusters is proposed.In this thesis, the research background and research development are reviewed first, then several typical detection methods of single user are described and compared with each other. The classical energy detection is researched in detail and is simulated under Additive White Gaussian Noise channel and Rayleigh channel. Single-user detection is compromised when it experiences multipath fading or shadowing. In order to deal with the limitation of single-user detection, cooperative