【作者】 卢为党；

【导师】 张乃通；

【作者基本信息】 哈尔滨工业大学 ， 信息与通信工程， 2012， 博士

【摘要】 随着无线通信技术的快速发展，新的无线应用和服务（如实时的视频会议，在线游戏等）呈爆发式的增长，这些新的应用和服务通常需要较宽的带宽和较高的速率。然而有限的资源和现有这种固定式的分配机制，使得这些新的应用和服务没有资源可用，制约着无线通信技术的发展。作为有效缓解这对矛盾的方法，资源优化设计一直以来是科学家们的研究热点。优化的资源不仅可以有效的提高资源的利用率，同时满足人们对高质量服务和应用的需求。协作分集利用了无线电波的广播特性，使得中继用户可以接收到源用户发送的信号，中继用户经过一定的处理后把源用户的信号转发到目的终端，这样不同的用户就可以共享彼此的天线形成虚拟的多天线阵列，获得MIMO(Multiple Input and Multiple Output)的效果，从而有效地利用频谱资源，改善用户的通信质量。认知无线电通过自适应地检测周围无线环境，实时地调整系统参数（如调制技术、工作频率和传输功率等），在保证授权用户的正常通信不受影响的前提下，伺机接入授权频谱，从而充分利用有限的频谱资源，有效地提升频谱资源的利用率，提高通信系统的容量。OFDM(Orthogonal Frequency Division Multiple)是下一代无线通信网络的一种关键技术，不仅可以有效提高频谱效率，而且还能有效对抗宽带通信中频率选择性衰落带来的影响。本文以下一代无线通信网络为背景，在分析国内外文献的基础上，主要利用OFDM技术对协作和认知无线电系统中的资源优化和频谱接入方法进行了深入研究。此外，我们还利用博弈论的方法对认知无线电系统中的频谱博弈问题进行了分析，提出了一种基于纳什博弈的协作频谱接入方法，并对这个频谱接入方法中的资源优化问题进行了分析和研究。论文的主要研究内容如下：第一，介绍了协作分集和认知无线电的系统模型及关键技术。首先，我们介绍了协作分集的系统模型，并详细描述了几种经典的中继协作协议，分析了它们的性能和特点。然后我们介绍了认知无线电中的几个关键技术，包括频谱检测、频谱分享和频谱管理等技术。第二，针对协作系统中资源优化算法资源利用率不高的问题，提出了一种OFDM固定中继系统中单中继多源用户资源优化算法，在该算法中，一个固定的中继节点利用OFDM符号可以同时帮助多个源用户转发信息，这样可以有效地利用频谱资源，进一步提高系统容量，降低中断概率。我们以信道容量最大化为原则，提出了源用户和中继节点间子载波和功率优化算法。第三，在现有固定频谱分配机制中，由于频谱资源受到分配机制的限制，利用率得不到根本上提高。我们把协作分集和认知无线电技术相结合，分别提出了基于放大转发（Amplify-and-Forward，AF）和解码转发（Decode-and-Forward，DF）协作的OFDM频谱接入方法，在我们所提的频谱接入方法中，如果授权用户在直传时候达不到自己的目标速率，认知用户作为授权用户的中继，利用一部分子载波来转发授权用户的信息帮助其达到目标速率，然后利用剩余的正交子载波来发送自己的信息，以此来获得频谱接入。这样他们之间就不会互相产生干扰，从而有效地利用了有限的频谱资源，进一步提高了系统性能。首先，我们对基于AF协作频谱接入方法做了一下理论分析，提出了一种集中式的AF协作频谱接入方法，该方法能够获得性能理论上界。但是该方法要交互很多信息，不能在分布式系统中实现，所以我们在此基础上又提出了一种分布式的AF协作频谱接入方法。如果认知用户能够解码授权用户的信息，认知用户就可以通过DF协作来帮助转发授权用户的信息，所以我们又提出了一种分布式的DF协作频谱接入方法，在该方法中认知用户解码授权用户的信息后，在协作的第二个时隙，就不用和授权用户的子载波进行配对。第四，现有的大部分频谱接入方法都是在不影响授权用户正常通信的前提下，认知用户才能接入授权用户的频谱，并且这些方法的优化目标都是最大化认知用户的性能。针对这个限制，我们提出了一种基于纳什博弈的协作频谱接入方法。在我们所提的方法中，如果认知用户能够帮助改善授权用户的性能，则授权用户就会主动分配一部分自己的发送时间给认识用户来改善自己的性能。这样不仅授权用户的性能能够得到改善，认知用户也能够获得频谱接入的机会来发送自己的信息，从而双方获得互赢的效果。更多还原

【Abstract】 With the development of wireless communication technology, the new wireless ap-plication and high quality serves (i.e. real-time video conference, online game) increaseexplosively. These wireless application and serves always need wide bandwidth and largerate. However, they do not have enough resource to use due to the limit resource andstatic resource allocation method. It is the main obstructions to develop wireless com-munication technology. As an efective way to relieve this conflict, resource optimizationalgorithms design is always the hot topic for scientists. Optimized resource can efective-ly improve the utilization of the resource. Meanwhile, it can satisfy the requirement ofthe high qualified service and application.Cooperative diversity makes the relay possible to receive the signal of the source byusing the broadcast nature of radio waves. After some proceeding, the relay help forwardthe signal of the source to the destination node. Thus, diferent users can share theirantennas and obtain MIMO (Multiple Input and Multiple Output) efects. It can efectivelyuse the spectrum and improve the quality of the communication. Cognitive radio canopportunistic access the primary spectrum by adaptively sensing the neighboring wirelessenvironment and timely adjusting the parameter of the system (i.e. modulation technique,working frequency, transmission power) while guaranteeing the primary transmission. Itmakes the best use of the limited spectrum, which can efectively improve the utilizationof spectrum. Thus it can help improve the performance of the wireless communicationsystems.OFDM(Orthogonal Frequency Division Multiple) is one of the key technology inthe next generation wireless communication network. It can efectively use the spectrum.Also, it can combat the influence of the frequency selective channel. This dissertationinvestigates the next generation wireless communication network. Based on the analysisof literature in China and abroad, the resource optimization in cooperative and cognitivesystem mainly by using OFDM technology is studied. Meanwhile, the spectrum bargain-ing problem in cognitive system by using bargaining method is analyzed. A cooperativespectrum access method based on Nash bargaining is proposed and the resource optimiza-tion in this method is investigated. The main work and innovative points are expressed asfollows. Firstly, the system model and key technology of cooperative diversity and cognitiveradio is introduced. The system model of cooperative diversity is introduced at first. Thenseveral classical relay protocols is describe, and the performance and characteristics ofthese protocols is analyzed. At last, the key technology of cognitive radio which includesspectrum detection, spectrum sharing and spectrum management is introduced.Secondly, a resource optimization in OFDM fixed relay system with one relay multi-source is proposed, in order to combat the low resource utilization of resource optimiza-tion algorithms in cooperative system. In our propose algorithms, one fixed relay is ableto help forward information of multiple sources in one OFDM symbol. It can efcient-ly utilize the available bandwidth, improve the system capacity and reduce the outageprobability. A joint sub-carrier matching and power allocation algorithm to maximize thechannel capacity is proposed.Thirdly, the spectrum utilization can not be fundamentally improved with the staticspectrum allocation. Cooperative OFDM spectrum access methods base on amplify-and-forward (AF) and decode-and-forward (DF) is proposed respectively, by combining thetechnology of cooperation and cognition. In our proposed methods, the cognitive user actsas a relay of the primary user, in which the cognitive user uses some of the subcarriersto help forward the primary signal, then uses the left orthogonal subcarriers to transmitits own signal to gain spectrum access. Thus, there will be no interference anymore. Itcan efciently use the limited spectrum to further improve the system performance. Acentralized AF spectrum access method by theoretically analyze is proposed at first. Thiscentralized method can achieve the upper bound performance, but it needs exchange toomuch information. It is hard to use in distributed system. Thus, a distributed method isproposed. If the cognitive user can decode the signal of primary user, the cognitive usercan use DF cooperation to forward the signal of primary user. A spectrum access methodbased on DF cooperation is proposed, in which the cognitive user do not need to pair thesubcarrier with primary user in the second slot.Fourthly, in most existing underlay spectrum access methods, the cognitive user canaccess the spectrum of primary user if the primary transmission is guaranteed. And thetarget of these methods is to maximize the rate of cognitive user. A cooperative spectrumaccess method based on Nash bargaining to extend these limit is proposed. In our pro-posed method, if the cognitive user can help improve the performance of the primary user,the primary user will lease a part of the transmission time to cognitive user to improve its own performance. In this spectrum access method, not only primary user can improveits performance, but the cognitive user can obtain spectrum access. Thus gain win-winefect.更多还原