【作者】 王彤；

【导师】 何继爱；

【作者基本信息】 兰州理工大学 ， 通信与信息系统， 2022， 硕士

【摘要】 随着下一代通信网络、移动互联网等新型信息技术的蓬勃发展,用户对无线通信传输速率的需求日益提高,无线频谱资源十分紧缺成为了限制无线通信可持续发展的主要瓶颈。然而传统的静态频谱分配方式已经无法为新型业务提供大量频谱接入机会,亟需一种动态分配机制有效改善频谱利用率低下的问题。认知无线电(Cognitive Radio,CR)能够利用周边无线通信环境的交互信息,实时检测频谱空洞并充分利用。CR的提出显著提高了频谱利用率,保证了主用户的通信质量。频谱感知作为CR的关键环节,可以实时检测到主用户是否活跃。将其检测频段扩展至宽带范围内,采用宽带频谱感知技术可以检测到更宽频谱空间的信道状态,从而提高下一代通信系统的数据速率,是近年来受到研究学者广泛关注的技术之一。但是仍存在很多问题未得到有效解决,如在低信噪比通信场景中噪声方差不确定性高、重构算法计算复杂度高、宽频段范围内高速采样导致的硬件负担大、能耗高等。基于上述问题,本文针对宽带频谱感知中的信号压缩采样、量化信号重构展开研究,主要贡献如下:(1)为了解决现存的理论性随机矩阵占用存储空间多、硬件实现复杂的问题,本文提出了一种基于伪随机序列的测量矩阵构造方法。该方法将随机高斯矩阵与伪随机序列、哈达玛矩阵相结合,通过改变随机高斯矩阵的阶数来调整测量矩阵的大小,使得所构造的矩阵既保留了高斯随机矩阵所需测量数少和伪随机序列相关性高的优点,又具有良好的重构性能。同时,提出了相关的定理并验证了其合理性。最后,在MATLAB仿真平台对一维随机稀疏信号和二维图像进行了对比实验验证。实验结果表明,相比于随机高斯矩阵、伯努利矩阵、托普利兹矩阵等理论随机矩阵及已有的基于伪随机序列观测矩阵,该测量矩阵具有更优的重构性能以及良好的应用价值。(2)针对现存量化信号频谱感知方案采样压力大、功耗计算资源有限及信道衰落严重等问题,提出了一种基于1-bit压缩感知的局部重构方案。首先,每个次级用户进行1-bit符号位感知,完成信号恢复后得到支撑集估计信息。利用信号的联合稀疏性,采用了基于单跳通信、分布式计算的平均一致性技术从而得到支撑集估计信息融合,并将其作为先验信息以完成下一次局部信号重构。在该方案中,信号重构和支撑集估计信息分布式融合这两个过程交替迭代进行,直到实现可靠的频谱检测。最后,通过在MATLAB平台仿真,与其他算法进行对比,证明了本文提出的重构算法在分布式协作频谱感知中的有效性。更多还原

【Abstract】 With the rapid development of new information technologies such as next-generation communication networks and mobile Internet,users’ demand for wireless communication transmission rates is increasing,and the shortage of wireless spectrum resources has become the main bottleneck restricting the sustainable development of wireless communications.However,the traditional static spectrum allocation method has been unable to provide a large number of spectrum access opportunities for new services,and it is urgent to need a dynamic allocation mechanism to effectively improve the problem of low spectrum utilization.Cognitive Radio(CR)can use the interactive information of the surrounding wireless communication environment to detect spectral holes in real time and make full use of them,so as to improve the utilization of spectrum.At the same time,if the frequency band is occupied by the main user,the cognitive user will quickly and actively avoid it,avoid conflicts,and affect the communication quality of the main user.Spectrum awareness,as a key link in cognitive radio,can detect in real time whether the main user is active or not.Wideband spectrum sensing,which can detect channel states in a wider spectrum space,thereby increasing the data rate of next-generation communication systems,is one of the technologies that has received widespread attention from researchers in recent years.However,there are still many problems that have not been effectively solved,such as high noise variance uncertainty,high computational complexity of reconstruction algorithms,large hardware burden and high energy consumption caused by high-speed sampling in a wide frequency band.Based on the above problems,this paper conducts research on signal compression sampling and quantitative signal reconstruction in broadband spectrum perception,and the main contributions are as follows:(1)Aiming at the problem that the hardware implementation of the random measurement matrix is difficult and the storage cost is large,this paper proposes an improved measurement matrix construction method based on pseudo-random sequence.This method combines the stochastic Gaussian matrix with the pseudo-random sequence and the Hadama matrix,and adjusts the size of the measurement matrix by changing the order of the stochastic Gaussian matrix,so that the constructed matrix not only retains the advantages of the small number of measurements required for the Gaussian stochastic matrix and the high correlation of the pseudo-random sequence,but also has good reconstruction performance.At the same time,the relevant theorems are proposed and verified to be reasonable.Finally,the one-dimensional random signal and the two-dimensional image signal are experimentally verified on the MATLAB simulation platform.Experimental results show that the relative theoretical stochastic Gaussian matrix,Toplitz matrix and the existing observation matrix based on pseudo-random sequence have excellent reconstruction performance and good application value.(2)Aiming at the problems of large sampling pressure,limited power consumption and computing resources and serious channel decline of existing spectrum perception schemes,a distributed collaborative spectrum perception scheme based on 1-bit is proposed.First,each secondary user(SU)performs 1-bit symbolic bit perception and obtains support set estimation information from the signal reconstruction process.Using the joint sparsity of the signal,the average consistency technology based on distributed computing and single-hop communication is fused,and the fusion result of the support set estimation information is used as a priori information to complete the next local signal reconstruction.In this algorithm,signal reconstruction and distributed fusion of support set estimation information alternate until reliable spectral detection is achieved.Finally,simulation proves the effectiveness of our proposed scheme in distributed networks.更多还原