【作者】 郭瑜；

【导师】 冯文江；

【作者基本信息】 重庆大学 ， 电路与系统， 2008， 硕士

【摘要】 随着无线通信业务的快速增长,日益增长的频谱需求与有限的频谱资源之间的矛盾已成为制约无线通信发展的主要瓶颈之一。认知无线电作为一种智能频谱共享技术,依靠人工智能的支持,感知无线通信环境,根据一定的学习和决策算法,实时自适应的改变系统工作参数,动态的检测和有效的利用空闲频谱,理论上允许在时域、频域和空域上进行多维的频谱复用和共享。频谱感知是构建认知无线电系统的前提之一,通过频谱感知搜集无线环境的统计信息,既能为高层的频谱管理提供辅助,也能为系统的功率控制提供必要的参数支持,论文重点研究认知无线电中的频谱感知技术,主要内容有:(1)在全面总结已有认知无线电频谱感知技术的原理、特点和性能的基础上,针对未知信号类型,分析了能量检测法在理想和衰落信道环境中对授权用户信号的感知性能和存在的不足,将多个认知用户协作感知和分集接收技术融入能量检测法,提出一种改进算法。改进的频谱感知算法既保留了能量检测法易于实现和无需先验信息的特点,又能有效提高检测的可靠性。分别在加性白高斯噪声信道、典型的多径衰落信道环境下进行了性能仿真和分析。(2)针对确定的线性调频信号和相位调制信号模型,分析了采用匹配滤波器感知授权用户信号的性能。利用高斯噪声的高阶累积量恒为零的特性,提出一种将高阶累积量和匹配滤波器相结合来消除高斯噪声,提高感知能力的方法,对提出的频谱感知算法进行了性能仿真和分析。(3)通过对基于认知无线电的无线区域网接口标准--IEEE802.22的认识和理解,针对该标准适用的TV频段,研究了通过匹配滤波器检测接收机本振耗散来感知频谱的方法。(4)认知无线电中传输场景分析包括频谱感知、频谱分析和频谱判决,为了实施频谱分析和估计,针对Welch法和多窗谱估计法进行了性能仿真,分析了两种方法中重要参数对估计能力的影响。更多还原

【Abstract】 As the rapid growth of the wireless communication, the conflict between the increasing demands and the limited spectrum resources has become one of the primary bottlenecks that constrains the development of wireless communication. As a smart spectrum sharing technology, cognitive radio can sense the wireless communication environment with the support of artificial intelligence; changing the system parameters automatically according to certain learning and decision making algorithms; testing idle spectrums dynamically and making best use of them. Theoretically speaking frequency division multiplexing and sharing are allowed in time, space and frequency domain. Spectrum sensing is one of the basis of constructing a cognitive radio system. The statistical information about wireless environment can be collected by spectrum sensing. It can provide both assistance for higher level spectrum management and necessary data to the power control of the system. Research about cognitive radio spectrum sensing technology is carried out in this dissertation with the main contents:(1) Principles, characteristics and performance in cognitive radio spectrum sensing technology are summarized fully in this dissertation. And the shortcomings of energy detecting method in sensing authorized users in both ideal and fading conditions are analyzed according to the type of unknown signal. Then an improved algorithm combined multiple users collaborative sensing and diversity reception technology into energy detection is proposed. The advantages of easy to implement and needless of prior information are retained while the reliability of detecting is improved effectively. In this dissertation performance simulation and analysis are done in both additive white Gaussian noise channel and typical multi-path fading channel.(2) With respect to the linear frequency modulated signal and phase modulated signal model, the performance of using matched filter to sense the signal of authorized users is analyzed. As the Gauss noise’s higher order cumulant permanent is zero, a new method combined higher order cumulant and matched filter to eliminate Gauss noise and improve sensing ability is proposed. Performance simulation and analysis are done according to the proposed algorithm.(3) After recognizing and understanding cognitive radio-based wireless area network interface standard --IEEE802.22, which is applicable for the TV band, a method using the matched filter to sense spectrum by detecting local oscillator leakage of receivers is described.(4) The transmission scene analysis in cognitive radio is constituted by spectrum sensing, spectrum analysis and spectrum decision. In order to analyze and estimate the spectrum, performance simulations aiming at the Welch’s method and Multitaper spectral estimation method are carried out. Finally the impact of important parameters in the two methods on estimation ability was analyzed.更多还原