《基于认知用户协作的频谱检测方法研究》由会员分享，可在线阅读，更多相关《基于认知用户协作的频谱检测方法研究（123页珍藏版）》请在金锄头文库上搜索。
1、中国科学技术大学 博士学位论文 基于认知用户协作的频谱检测方法研究 姓名：麦磊鑫 申请学位级别：博士 专业：通信与信息系统 指导教师：戴旭初 2011-04-15 摘 要 I 摘摘 要要 随着无线业务快速发展，对无线频谱资源的需求日益增长，无线频谱资源 匮乏问题也越发严峻。在现有的固定频谱分配政策下，只允许授权用户（主用 户）使用授权频段，致使频谱使用效率低下，大量授权频谱资源被闲置，而日 益增长的频谱资源需求却无法得到满足。 认知无线电是提高频谱资源利用率、缓解频谱资源相对匮乏的有效途径之 一，其中频谱检测是实现认知无线电技术的关键基础，其目标是快速可靠地获 取空间域和时间域上的频谱空穴信息
2、，为后续的频谱管理、频谱共享和频谱移 动操作提供信息依据。目前国内外科研工作者针对频谱检测已开展了大量的研 究，取得许多进展，但仍然存在着一些具有挑战性的问题。 本文针对认知用户（次用户）协作的频谱检测中的任务分配、检测信息融 合，以及空时联合检测等问题进行探讨和研究，给出了相应的理论模型和算法， 并通过理论分析和实验仿真验证了算法的有效性。 本文的主要研究内容和贡献有： 1）考虑次用户硬件能力受限的约束条件，研究了多信道频谱的快速检测问 题。从信息论的角度分析了系统检测信息的准确性与系统熵的关系，提出了基 于熵最小化准则的协作频谱检测策略。该策略核心思想是根据次用户反馈的检 测结果，按照系统
3、熵最小化的准则分配次用户检测不同的信道，自适应地调整 每个信道的检测时间间隔，减少检测系统对整体信道状态估计的不确定性，从 而提高信道状态估计的准确性。 2）考虑阴影衰落以及隐藏终端问题，研究了单信道频谱的可靠检测问题。 将带有错误检测信息的协作频谱检测问题建模为隐马尔可夫模型，并根据隐马 尔可夫模型的理论，提出了相应的信息融合算法。算法的主要思想是通过观测 序列学习模型参数，然后收集次用户的检测信息，在线估计信道状态，提高了 单信道频谱的检测可靠性。另外，还讨论了次用户选择问题，给出了选择次用 户的方法。理论分析和实验仿真验证了所提出算法的有效性，检测性能优于已 有的协作检测算法。 3）考虑
4、频谱空穴的时空分布特点，研究了空时联合的频谱检测问题。将空 时联合的频谱检测建模为系统参数和状态的联合估计问题，提出了基于粒子滤 波的协作频谱检测方法。该方法能够对频谱进行空时联合检测，即对主用户空 间上的位置，以及时间上的信道状态同时进行在线估计。此外，还给出了次用 户选择方法，结合所估计的空间和时间信息，根据与主用户距离、空间分布对 摘 要 II 称性等标准选择合适的次用户参与协作，有效提高了时间域和空间域上的频谱 检测性能。 关键词：关键词： 认知无线电，伺机频谱接入，协作频谱检测，硬件能力约束，阴影衰 落，信息熵，隐马尔可夫模型，粒子滤波，空时联合频谱检测 Abstract III A
5、BSTRACT Presently, the demand for wireless spectrum resource is growing with the rapidly developing wireless service, and the increasing demand makes wireless resources more and more scarce. However, only authorized users (primary users) are allowed to use the licensed band with the current fixed sp
6、ectrum allocation policy, which makes a large number of licensed spectrum resources sometimes are idle. That is to say, the fixed spectrum allocation policy results in inefficient usage of spectrum and can not well satisfy the growing demand for spectrum resources. Cognitive radio is one of effectiv
7、e ways to improve the efficiency of spectrum usage, while spectrum sensing is the key basis for realizing cognitive radio. The objective of spectrum sensing is to acquire the information of spectrum hole in spatial domain and temporal domain quickly and reliably, so as to support the follow-up funct
8、ions of cognitive radios, such as spectrum management, spectrum sharing and spectrum mobility. Up to now, although some researchers have carried out a lot of research and made a mighty advance in spectrum sensing, there still exist some challenging issues. In this dissertation, secondary user cooper
9、ation based spectrum sensing is investigated, mainly including sensing task allocation, sensing information fusion, and jointly spatial-temporal cooperative spectrum sensing. The some theoretical models and relative algorithms are proposed, and extensive simulations are also presented to demonstrate
10、 the validity of the proposed methods. The main contents and contributions of this dissertation are listed as follow. 1) Under the constraints of limited hardware capabilities of secondary users, how to rapidly sense multi-channel spectrum is studied. Based on information theory, the relationship be
11、tween sensing accuracy and the system entropy is analyzed. Then a least entropy based strategy for cooperative spectrum sensing is proposed. The main idea behind the proposed strategy is based on the least entropy criteria to allocate secondary users to sense different channel according to the feedb
12、ack of sensing results, which can adaptively adjust sensing interval of each channel to maximumly reduce the uncertainty of the estimation of channel state, and improve the accuracy of the channel estimation. 2) Taking the impact of shadow fading and hidden node problem into Abstract IV consideratio
13、n, how to increase the reliability of spectrum sensing for single-channel is studied. In the dissertation, the cooperative sensing with error information is modeled as hidden Markov model. Based on hidden Markov model theory, an online algorithm for sensing information fusion is proposed. The basic
14、idea of the proposed algorithm is to learn the model parameters through the observation sequence, then collect sensing information of secondary users to estimate the channel state, which can improve the reliability of single-channel spectrum sensing. Meanwhile, the selection strategy of secondary us
15、er and a selection method are also given. Theoretical analysis and simulations demonstrate the efficiency of proposed algorithm. It is shown that the proposed approach outperforms the existing methods for cooperative spectrum sensing. 3) Considering the character of spectrum hole with spatial-tempor
16、al distribution, how to jointly sense the spatial-temporal spectrum is studied. The joint sensing problem is modeled as parameter-state combined estimation, and a particle filtering based solution strategy is proposed. The proposed method can sense the spectrum in spatial domain and temporal domain simultaneously, which means to online estimate the location of primary user in spatial domain and the channel state in tempo
