【作者】 孙萌；

【导师】 张钦宇；

【作者基本信息】 哈尔滨工业大学 ， 信息与通信工程， 2012， 硕士

【摘要】 随着无线通信事业的高速发展，固定的频谱分配方式导致频谱利用率低的问题日益显著，并逐渐成为无线通信发展的物理瓶颈。认知无线电技术的适时提出，有效的解决了频谱利用率低下的问题。频谱共享作为认知无线电的关键技术，能够帮助认知用户在降低对主用户干扰的前提下，更充分有效地利用频谱资源。本文针对认知无线电中的动态频谱共享技术进行了深入研究，主要研究内容如下：（1）首先针对固定式频谱分配方式导致的认知用户灵活性偏低、系统性能指标提高不明显的问题，提出了重建具有可变带宽机制的认知用户系统模型。模型中加入用户的平均服务等级标准来衡量用户的灵活性，认知用户的阻塞率和掉话率来衡量系统的频谱利用率。（2）接着提出了认知用户能够理想感知条件下，基于可变带宽机制的贪婪算法。算法中认知用户根据系统的子信道的开关情况进行不间断的服务等级调整，从而使认知用户得到更好的服务质量，达到有效利用频谱资源的目的，之后根据系统模型建立马尔可夫转移模型，求解系统的性能指标。最后通过与基于不完全可调带宽的贪婪算法进行仿真对比，证明了新算法在频谱利用率方面的有效性。（3）为进一步优化贪婪算法的性能，提出了基于马尔可夫决策过程的可变带宽动态频谱共享算法。算法中对系统中用户的接入与调整策略进行重新建模，利用系统的收益作为用户执行某一策略的约束条件，之后选用微粒群策略迭代法求解系统最大收益，进而得出最优动态频谱共享策略。最后通过与贪婪算法的仿真结果对比，结果表明马尔可夫决策过程算法的QoS性能以及系统灵活性都优于贪婪算法，从而验证了马尔可夫决策过程算法的有效性。更多还原

【Abstract】 With the rapid development of wireless communications, the fixed spectrumallocation method leads to an increasingly significant problem of low spectrumutilization, and it becomes the main factor that hinders the development of wirelesscommunication. Cognitive radio technology is a timely manner, it solves theproblem of low spectrum utilization. Spectrum sharing is the key technology ofcognitive radio, it helps cognitive users to make full use of spectrum resources,while reduce the interference on primary users. This paper makes a deep study ondynamic spectrum sharing of cognitive radio technology, the main contents are asfollows:Aiming at the problem of low cognitive user flexibility caused by fixedspectrum allocation, reconstructive cognitive user system model with variablebandwidth mechanism is proposed. In the proposed model, average user servicelevel is used to measure the flexibility of the user, while the blocking rate and calldrop rate of cognitive users is used to measure the spectrum efficiency of thesystem.The greedy algorithm based on variable bandwidth mechanism while cognitiveusers can percept ideally is proposed. In the algorithm, in order to get the bestquality of service and use the spectrum resources effectively, cognitive users adjustthe service level uninterruptedly according to the sub-channel switch of the system.After that, a Markov transfer model is created according to system model, and thenthe system performance is evaluated. Compared with the greedy algorithm whichbased on partly adjustable bandwidth, simulation results show that the newalgorithm can solve the problem of low spectrum utilization better.In order to further optimize the performance of greedy algorithm, the variablebandwidth and dynamic spectrum sharing algorithm based on Markov decisionprocess is proposed in this paper. In the algorithm, the user access and adjustmentstrategies are re-modeled, the system revenue is the constraint for user behavior,and the particle swarm iterative method is selected for solving the maximum benefitof the system, then the optimal dynamic spectrum sharing policy is obtained.Compared with the greedy algorithm, simulation results show that both the QoSstandards and system flexibility of MDP algorithm are better than greedy algorithm.更多还原