【作者】 王春生；

【导师】 刘林；

【作者基本信息】 西南交通大学 ， 通信与信息系统， 2011， 硕士

【摘要】 频谱资源是无线通信系统最为宝贵及有限的资源。然而,现阶段频谱资源的匮乏与新业务大增对频带的需求之间的矛盾越来越突出。为解决该矛盾,认知无线电应运而生。频谱检测是认知无线电实现的基础和前提,也是最具有挑战性的难题之一。针对该问题,本文对本地频谱检测算法和合作频谱检测算法进行了较为深入的研究,并进行了大量仿真和系统性能分析。首先,本文详细介绍了认知无线电技术的研究背景、国内外发展现状及研究意义,接着阐述了各种本地频谱检测算法,并详细推导了非合作式能量检测算法的检测概率和虚警概率,对其性能进行了仿真分析。由于本地频谱检测易受阴影、多径衰落的影响,其检测性能大大降低。为解决该问题,引入了合作频谱检测技术。为了分析合作频谱检测的性能,本文对几种常见的软合并算法(如最大比合并、等增益合并、最优合并)进行了分析,并在此基础上引出了2bits量化方案,然后对其性能进行了仿真对比,结果表明2bits算法在复杂度和性能方面达到很好的折中。其次,本文在能量检测基础上对几种较为常见的硬融合算法(如“与”、“或”、“k秩”合并准则)进行了分析,并在其基础上提出了一种自适应的单门限合作频谱检测算法,结果表明该算法具有较好的鲁棒性。论文最后对频谱检测的几个关键参数(如检测概率、虚警概率、吞吐量及检测时间)进行了优化,并在其基础上提出了一种改进的吞吐量优化方案,结果表明该算法具有较好的性能。更多还原

【Abstract】 Spectrum is the most importance resource for wireless communication systems. The inconsistency between the lack of spectrum resources and new service demands which need lots of bands becomes more and more obviously. In order to solve the problem, cognitive radio is presented. Spectrum sensing is the basis and prerequisite of cognitive radio. It is also the most challenging issues in cognitive radio. Aiming at the problem, the local spectrum sensing algorithms and cooperative spectrum sensing algorithms are discussed in detail with lots of simulation analysis in this paper.First of all, the backgrounds, applications and developments and research significance are introduced, then all kinds of local spectrum sensing algorithms are describes, and the formula derivation of non-cooperative energy detection are deduced in detail with simulations and performance analysis results. However, the performances of local spectrum sensing algorithms degrade fast due to Multipath Rayleigh Fading. In order to eliminate the multipath influence, cooperative spectrum sensing is represented. The soft combination algorithms (for example maximum ratio combination, equal gain combination and optimal combination etc) and an innovation 2bits quantization scheme are introduced. The simulations and performance analysis have been done, and the simulation results indicate that the 2bits scheme is very well compromised between performance and complexity. Secondly, the conventional hard combination algorithms (for instance "and", "or", "k out of n") based on energy detection in cooperative spectrum sensing are represented. Based on the above, a self-adaptation single threshold spectrum sensing algorithm for cooperative users is proposed. The simulation results indicate that the proposed algorithm can boost the system performance. Eventually, the key parameters (such as detection probability, false alarm probability, throughput and sensing time) in cognitive radio are discussed and optimized in this paper. At the same time, an innovation optimal strategy for throughputs is introduced followed with the simulation results which indicate that the strategy is very well.更多还原