认知无线电中基于隐马尔可夫模型的频谱感知和预测研究

发布时间：2018-06-25 08:01

本文选题：认知无线电技术 + 频谱感知　；参考：《安徽理工大学》2017年硕士论文

【摘要】：近年来,随着无线电技术的快速发展,移动数据业务的爆炸性增长,固定式的频谱分配方案已不能满足当今世界的无线电通信的需求。然而研究表明,当前全球范围内被授权的频段利用率是非常低的,而认知无线电能够自适应地与电磁环境交互信息,探测感知到频谱空洞,实现动态频点转换和传输参数的调整,其这一特性在提高频谱利用率方面有着广泛的研究。频谱感知是认知无线电系统的首要任务,快速而准确的频谱感知微观上能够减少主、次用户之间的碰撞机率,提高通信质量,提升系统的吞吐量。宏观上则可以有效提高空闲频谱的利用率。本文就认知用户如何快速准确地捕捉到频谱空洞展开讨论。为了实现快速准确的频谱感知以及弥补本地频谱感知信任度不足的问题,本文首先提出一种基于离散隐马尔可夫模型的自适应协作频谱预测方案,根据本地频谱感知的信任度赋予相应的权重,通过数据融合的方式对下一时刻的信道状态作出预测。仿真结果表明,本方法在频谱预测方面有较高的准确性。由于离散隐马尔可夫模型只能处理数值离散的观测序列,在处理数值连续的观测序列时会产生信息失真。另外,为了充分利用频谱感知的历史信息,提高频谱感知的准确性和时效性,本文将连续隐马尔可夫模型引入协作频谱感知。研究证明,在不同的信道状态下授权用户的能量值满足不同的高斯分布,所以采用模式识别的方式判定信道两种不同的状态。通过仿真验证,本文提出的基于连续隐马尔可夫模型通过模式识别的频谱感知方案在一定程度上能够克服离散隐马尔可夫的信息失真问题,具有较高的频谱感知性能。
[Abstract]:In recent years, with the rapid development of radio technology and the explosive growth of mobile data services, the fixed spectrum allocation scheme can no longer meet the needs of the world's radio communications. However, research shows that the current authorized frequency band utilization rate is very low worldwide, and cognitive radio can adaptively interact with the electromagnetic environment to detect spectrum holes. The dynamic frequency point conversion and the adjustment of transmission parameters are widely studied in improving the spectral efficiency. Spectrum sensing is the primary task of cognitive radio system. Fast and accurate spectrum sensing can reduce collision probability between primary and secondary users improve communication quality and improve system throughput. Macroscopically, the efficiency of idle spectrum can be improved effectively. This paper discusses how cognitive users capture spectrum holes quickly and accurately. In order to realize fast and accurate spectrum sensing and make up for the lack of trust in local spectrum sensing, this paper first proposes an adaptive cooperative spectrum prediction scheme based on discrete hidden Markov model. According to the trust degree of local spectrum perception, the channel state at the next moment is predicted by data fusion. The simulation results show that the proposed method is accurate in spectrum prediction. Because discrete hidden Markov models can only deal with discrete observation sequences, information distortion will occur when processing numerical continuous observation sequences. In addition, in order to make full use of the historical information of spectrum sensing and improve the accuracy and timeliness of spectrum sensing, this paper introduces the continuous hidden Markov model into cooperative spectrum sensing. It is proved that the energy values of authorized users satisfy different Gao Si distributions in different channel states, so the mode recognition is used to determine two different states of the channel. The simulation results show that the proposed spectrum sensing scheme based on continuous hidden Markov model through pattern recognition can overcome the information distortion problem of discrete hidden Markov model to a certain extent and has high spectrum sensing performance.
【学位授予单位】：安徽理工大学
【学位级别】：硕士
【学位授予年份】：2017
【分类号】：TN925

【相似文献】