硕士学位论文认知无线电系统功率控制和频谱分配技术研究与闭环功率控制实现申请人：冯冀学科专业：信息与通信工程指导教师：罗新民 副教授2008年05月Study of Power Control and Spectrum Allocation and Implementation of Close Loop Power Control in Cognitive Radio SystemA thesis submitted toXian Jiaotong Universityin partial fulfillment of the requirementsfor the degree ofMaster of Engineering ScienceByJi Feng(Information and Communication Engineering)Supervisor: Associate Prof. Xinmin LuoMay 2008摘 要论文题目：认知无线电系统功率控制和频谱分配技术研究与闭环功率控制实现 本研究得到国家高技术研究发展计划（“863”计划）（编号：2005AA123910），陕西省自然科学基金项目（编号：2006F41），陕西省科技攻关计划项目（编号：2005K04-G11）资助。学科专业：信息与通信工程申请人：指导教师： 副教授摘 要认知无线电技术能够主动检测频谱使用情况，自适应的改变自身通信参数。在不影响授权用户的前提下，择机选择授权用户的空闲频段进行通信，具有灵活、频谱利用率高等优点，因而引起了国内外研究者的广泛关注。由于认知无线电用户与授权用户共享频段，且授权用户具有最高优先级，使认知无线电相关技术具有新的特点。论文对认知无线电系统的功率控制和频谱分配技术展开研究，并设计实现了认知无线电实验系统上的闭环功率控制模块。首先，论文在分析认知无线电自适应、交互式决策过程的基础上，建立了功率控制的博弈论模型。针对采用非共享机制的认知无线电系统，提出了基于函数的功率控制博弈算法。仿真结果表明：与已有算法相比，新算法在保证认知无线电用户信干噪比的前提下，提高了系统吞吐量。针对采用共享机制的认知无线电系统，改进了效用函数，提出了基于指数惩罚系数的功率控制博弈算法。仿真结果表明：新算法保障了授权用户足够低的掉线概率和认知无线电用户的信干噪比，有效控制了发射功率。其次，针对采用OFDM技术的认知无线电系统中占用信道数和时间开销的矛盾，论文提出实际吞吐量的概念及基于此概念的一种最优频谱分配策略，并设计了基于遗传算法的最优策略求解方法。仿真结果表明：当分配给认知无线电用户的频谱资源较少时，实际吞吐量随着分配的频谱资源的增多而增大。随着占用的频谱资源进一步增多，与授权用户发生冲突的概率增大，实际吞吐量反而呈单调下降趋势。论文提出的策略对认知无线电用户占用信道数进行了优化，提高了系统性能。最后，论文在调试认知无线电实验系统射频模块的基础上，设计并实现了闭环功率控制模块。调试结果表明：在射频模块工作正常的基础上，本模块能够对发射功率进行准确、实时的控制；在移动情况下，能够对信道增益的变化做出及时补偿，保障了接收端足够高的信噪比和足够低的误码率。关 键 词：认知无线电；博弈论；功率控制；频谱分配；闭环论文类型：应用研究IABSTRACTTitle:Study of Power Control and Spectrum Allocation and Implementation of Close Loop Power Control in Cognitive Radio SystemThis research was funded by the National High-tech Research and Development Plan（”863” Plan）（No.2005AA123910）,Shaanxi Province Natural Science Fund（No.2006F41）and Shaanxi Province Scientific and Technological Plan（No.2005K04-G11）.Speciality:Information and Communication EngineeringApplicant:Ji FengSupervisor:Associate Prof. Xinmin LuoABSTRACTCognitive Radio technology can detect the utilization of the spectrum initiatively, and change its communication parameters adaptively. On the premise of no prejudice to licensed users, cognitive radio can choose the idle spectrum of licensed users to communicate in proper opportunities. For its advantages of agility, and high spectrum utilization, cognitive radio has been concerned widely by the scholars at home and abroad. Because licensed users share spetrum resource with cognitive users and have the highest priority, the technologies in cognitive radio have new characteristics. The technologies of power control and spectrum allocation in cognitive radio systems is researched, and close loop power control module in cognitive radio experimental system is designed and implemented in this thesis.Firstly, based on the analysis of adaptive and interactive decision process in cognitive radio, a game theory model of power control is established. Aimed at cognitive radio system based on exclusive model, a power control game algorithm based on arctan function is proposed. The simulation results show that new algorithm not only guarantees the cognitive users SINR,but also imporves the system throughput compared with other algorithms. Aimed at cognitive radio system based on common model, the utility function is improved,and a power control game algorithm based on exponent chastisement factor is proposed. The simulation results show that new algorithm guarantees the SINR of cognitive users and low enough outage probability of licensed users, and the transmitting power is controlled effectively.Secondly, aimed at the contradiction between number of occupied channels and time cost in OFDM cognitive radio system. a goodput concept and an optimal spectrum allocation strategy based on this concept are proposed. And a method based on genetic algorithm for the solution of the optimal strategy is designed. The simulation results show that when the spectrum resource allocated to the cognitive users is small, the goodput increases with the increase of allocated spectrum resource; when the occupied spectrum resource by cognitive users increases further, the conflict probability to licensed users increases, and the goodput decreases monotonically with the increase of allocated spectrum resource. This strategy proposed in the thesis optimizes the number of occupied channels by cognitive users, and improves the OFDM cognitive radio system performance.Finally, close loop power control module in cognitive radio experimental system is designed and implemented based on debugging of RF module in experimental system. The debugging results show that transmitting power can be controlled accurately and in real-time by this module based on RF modules normal operation; when the experimental terminal is moving, the change of channel gain can be compensated in time, and SNR and BER in receiver can