【作者】 叶丹；

【导师】 汪一鸣；

【作者基本信息】 苏州大学 ， 通信与信息系统， 2010， 硕士

【摘要】 认知无线电CR(Cognitive Radio)作为一种革命性智能共享技术,可显著提高频谱利用率。认知无线电使用动态频率接入,应具备自动检测无线电频谱空洞功能。建立认知无线电网络CRN(Cognitive radio networks)首要解决的核心问题是如何采用频谱感知技术准确地识别频谱空穴,检测授权用户出现。在认知用户和授权用户工作在相同频段的情况下,认知用户如何采用功率控制技术调整传输功率在保证授权用户服务质量的同时实现动态频谱接入和频谱共享。本文主要研究认知无线电网络中的频谱感知技术和功率控制技术,创新点主要体现在认知用户智能感知与传输功率控制技术方面。文章首先分析了课题的研究背景与意义,国内外发展现状,简要介绍认知无线电主要关键技术,然后总结分析了现有频谱感知技术。重点讨论一种基于认知无线电技术的802.22无线区域网系统授权用户频带特征感知方案。研究中主要针对DVB-T、AM无线麦克风两类授权用户信号特点进行精细感知,并着重分析了循环周期长度和延时等参数对检测性能的影响。结果表明增加计算自相关函数时的循环周期长度或采样点数,降低延时,都有利于信号检测,从而验证了该算法的优良性。该算法可以改善系统频谱感知性能,更好地保护授权用户,充分挖掘频谱资源。其次,在分析认知用户与授权用户共存模型的基础上,提出了基于授权用户功率检测的自适应感知阈值的设定方法和接入控制方案,并针对IEEE802.22无线区域网中的授权用户,分别进行数学分析和实际模型仿真。通过自适应感知阈值调整控制认知用户发射功率,能保持恒定的漏检和虚警概率,获得最佳的频谱感知性能,使认知用户在不对授权用户产生干扰的同时与授权用户最佳共存。最后,基于非视距(NLOS)环境下的单次反射统计信道模型,由双基站测得多组经由多个散射体反射的移动台的电波到达角(AOA)和主基站提供的电波到达时间(TOA),提出了一种结合神经网络和加权质心算法来实现授权用户定位的新方法即基于移动终端周围散射体信息利用主次双基站来定位主用户。描述了定位模型和定位算法,结合神经网络对散射信息进行修正估计和定位计算,将授权用户定位结果代入自适应感知阈值算法。该算法能有效地抑制NLOS误差,进一步提高定位精度,便于认知用户更准确地自适应调整感知阈值,降低阈值误差。通过分析比较各种定位算法的定位误差对感知检测性能产生的影响,表明本文的定位算法可以改善检测性能,获得较准确的共存概率。更多还原

【Abstract】 As a revolutionary smart spectrum sharing technology, cognitive radio (CR) can significantly improve the spectrum utilization, which can be able to seek and use in a dynamic way the frequencies for network access by autonomous detection of vacant bands in radio spectrum. The key problem in cognitive radio network is how to use spectrum sensing technology to find the spectrum hole and identify the license users. And how to achieve the optimum coexistence between SU and PU at the same frequency using transmission powers control (TPC) technology to guarante the PU’s communication required by the quality-of-service (QoS).The creative work of this paper is mainly manifested in the improvement of intelligent sensing algorithm and power control method in cognitive radio network.In this paper, the subject background and significance of CR is presented. And the development of the status quo at home and aboard is described. The chief introduction on key technology in cognitive radio is made. The conclusions and analysis on current spectrum sensing technology are made. This paper focuses on methods for feature sensing of primary bands in IEEE 802.22 WRAN systems. This method facilitates the detection of signals by increasing the number of cyclic period or the duration of cyclic period used in the calculation of the correlation function as well as reducing the delay. Simulations in fine sensing by characteristic of primary user signal are presented. The algorithm proposed can promote the improvement of the performance of system spectrum sensing and protection of license user as well as make full use of spectrum resource.Secondly, a novel setting and access control method of adaptive sensing threshold based on power detection of the primary user (PU), on the basis of the analysis of the coexistence model between PU and SU, is proposed. The mathematical and practical simulations are made, and results analysis of the proposed mechanism in IEEE 802.22WRAN is carried out. The sensing threshold calculated by power detection of PU, not only maintains constant probabilities of missed detection and false alarm, but also obtains the optimum performance of spectrum sensing. The optimal SU’s transmission power, adaptively according to sensing threshold, enables both SU and PU to coexist in the same channel without interfering each other.Finally, the multi-group TOA (Time of Arrival) and AOA (Angle of Arrival) provided by two base stations are measured based on the single reflect statistics channel model under NLOS environment. The paper initials new sensing method using scatter information around mobile user to locate the primary user. A new algorithm combined neural networks with the weighted centroid algorithm to achieve primary user’s location is proposed. The system location model and the algorithmic principle of location are described. The location results of primary user are applied into adaptive sensing algorithm and analysis of the impact on detection performance towards location error in the proposed mechanism. The algorithm proposed can restrain NLOS error effectively and further improve localization accuracy. In order to adaptively adjust sensing threshold for cognitive radio and lower threshold error, the analysis and comparison of various kinds of localization algorithm error affect sensing detection performance. The result shows the current localization algorithm improves detection performance and obtains more exact probability of spectrum coexistence.更多还原