基于压缩感知的无线通信信号处理方法研究
发布时间：2019-03-08 20:23
【摘要】：传统的信号处理方法是在Nyquist采样定理的基础上建立起来的,这就意味着为了不失真的恢复原始信号,发送信号的采样速率至少为信号带宽的两倍。然而,随着无线通信技术的飞速发展,信息需求量与日俱增,信号带宽也随之不断增加,这使得传统信号处理方法中信号的采样、传输、处理和存储过程都面临严峻的挑战。如何在获取足够多信息量的同时有效降低信号采样速率已经成为当前无线通信领域的研究热点。压缩感知(Compressed Sensing,CS)理论能够通过低维观测信号恢复出高维原始稀疏信号,将压缩感知理论应用于无线通信信号处理过程中,不仅能够显著降低信号的采样速率,而且可以减少信号处理过程中的信息量,提高通信系统性能。为此,本文对CS在无线通信领域中的应用进行了相关研究,主要内容包括基于CS的信号重构方法,基于CS的信道缩短方法以及基于CS的宽带频谱感知方法,主要创新点如下:1、在基于CS的信号重构方法的研究中,为了提高稀疏信号重构效率,提出了一种一次投影子空间追踪算法。首先将传统的子空间追踪算法的迭代过程分解为一次相关最大化过程和两次投影过程;然后通过减少迭代中观测向量在支撑集上的投影过程,降低了算法的复杂度,提高了稀疏信号的重构效率。同时,分析了现有的衡量算法重构性能指标的不足,并提出一种更可靠的参考指标。2、在基于CS的信道缩短方法的研究中,针对使用信道缩短均衡器的通信系统的复杂度会随着均衡器中非零抽头的增加而快速增大的问题,提出了一种半融合贪婪追踪算法,并用该算法实现了稀疏信道缩短均衡器。首先在最小均方误差准则下,将信道缩短问题转化为均衡器中非零抽头数目最小化问题;然后通过稀疏度预估计过程确定了均衡器稀疏度下界;最后通过回溯重构阶段和支撑集扩展阶段确定了均衡器的非零抽头,实现了稀疏信道缩短均衡器。3、在基于CS的宽带频谱感知方法的研究中,为了提高宽带频谱感知效率,设计了一种基于压缩感知的协作频谱感知方法。首先通过带通滤波器组和压缩采样方法的组合使用,进一步降低了宽带信号的采样速率;然后通过认知用户与融合中心的信息交互,确定了每个认知用户所需重构的子频段,该过程使得每个认知用户只需要重构部分频带,大幅度降低了信号重构阶段的复杂度;最后通过多用户的协作感知提高了感知方法的可靠性。
[Abstract]:The traditional signal processing method is based on the Nyquist sampling theorem, which means that in order to restore the original signal without distortion, the sampling rate of the transmitted signal is at least twice the bandwidth of the signal. However, with the rapid development of wireless communication technology, the demand for information is increasing day by day, and the signal bandwidth is also increasing, which makes the signal sampling, transmission, processing and storage procedures face severe challenges in the traditional signal processing methods. How to obtain enough information while effectively reducing the signal sampling rate has become a hot topic in the field of wireless communication. The compressed sensing (Compressed Sensing,CS) theory can recover the high-dimensional original sparse signal from the low-dimensional observation signal, and apply the compressed sensing theory to the wireless communication signal processing process, which not only can significantly reduce the sampling rate of the signal. Moreover, it can reduce the amount of information in the signal processing process and improve the performance of the communication system. In this paper, the application of CS in wireless communication is studied, including CS-based signal reconstruction method, CS-based channel shortening method and CS-based broadband spectrum sensing method. The main innovations are as follows: 1. In order to improve the efficiency of sparse signal reconstruction, a one-time shadow space tracking algorithm is proposed in the research of signal reconstruction method based on CS. Firstly, the iterative process of the traditional subspace tracking algorithm is decomposed into one-time correlation maximization process and two-time projection process. Then by reducing the projection process of the observation vector on the support set, the complexity of the algorithm is reduced and the reconstruction efficiency of sparse signal is improved. At the same time, the shortcomings of the performance index of the existing measurement algorithm reconstruction are analyzed, and a more reliable reference index is proposed. 2. In the research of channel shortening method based on CS, In order to solve the problem that the complexity of communication system using channel shortening equalizer will increase rapidly with the increase of non-zero tap in equalizer, a semi-fusion greedy tracking algorithm is proposed, and the sparse channel shortening equalizer is implemented with this algorithm. Under the minimum mean square error criterion, the channel shortening problem is transformed into the problem of minimizing the number of non-zero taps in the equalizer, and then the lower bound of the sparsity of the equalizer is determined by the sparsity pre-estimation process. Finally, the non-zero tap of the equalizer is determined by the backtracking reconstruction phase and the extension stage of the support set, and the sparse channel shortening equalizer is realized. 3. In order to improve the efficiency of broadband spectrum sensing in the research of broadband spectrum sensing method based on CS, A cooperative spectrum sensing method based on compressed sensing is designed. Firstly, a combination of band-pass filter bank and compression sampling method is used to further reduce the sampling rate of wideband signals. Then through the information interaction between the cognitive user and the fusion center, the sub-band that each cognitive user needs to be reconstructed is determined. In this process, each cognitive user only needs to reconstruct part of the frequency band, which greatly reduces the complexity of the signal reconstruction stage. Finally, the reliability of the perceptual method is improved by multi-user cooperative perception.
【学位授予单位】：宁波大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN911.7
本文编号：2437178
[Abstract]:The traditional signal processing method is based on the Nyquist sampling theorem, which means that in order to restore the original signal without distortion, the sampling rate of the transmitted signal is at least twice the bandwidth of the signal. However, with the rapid development of wireless communication technology, the demand for information is increasing day by day, and the signal bandwidth is also increasing, which makes the signal sampling, transmission, processing and storage procedures face severe challenges in the traditional signal processing methods. How to obtain enough information while effectively reducing the signal sampling rate has become a hot topic in the field of wireless communication. The compressed sensing (Compressed Sensing,CS) theory can recover the high-dimensional original sparse signal from the low-dimensional observation signal, and apply the compressed sensing theory to the wireless communication signal processing process, which not only can significantly reduce the sampling rate of the signal. Moreover, it can reduce the amount of information in the signal processing process and improve the performance of the communication system. In this paper, the application of CS in wireless communication is studied, including CS-based signal reconstruction method, CS-based channel shortening method and CS-based broadband spectrum sensing method. The main innovations are as follows: 1. In order to improve the efficiency of sparse signal reconstruction, a one-time shadow space tracking algorithm is proposed in the research of signal reconstruction method based on CS. Firstly, the iterative process of the traditional subspace tracking algorithm is decomposed into one-time correlation maximization process and two-time projection process. Then by reducing the projection process of the observation vector on the support set, the complexity of the algorithm is reduced and the reconstruction efficiency of sparse signal is improved. At the same time, the shortcomings of the performance index of the existing measurement algorithm reconstruction are analyzed, and a more reliable reference index is proposed. 2. In the research of channel shortening method based on CS, In order to solve the problem that the complexity of communication system using channel shortening equalizer will increase rapidly with the increase of non-zero tap in equalizer, a semi-fusion greedy tracking algorithm is proposed, and the sparse channel shortening equalizer is implemented with this algorithm. Under the minimum mean square error criterion, the channel shortening problem is transformed into the problem of minimizing the number of non-zero taps in the equalizer, and then the lower bound of the sparsity of the equalizer is determined by the sparsity pre-estimation process. Finally, the non-zero tap of the equalizer is determined by the backtracking reconstruction phase and the extension stage of the support set, and the sparse channel shortening equalizer is realized. 3. In order to improve the efficiency of broadband spectrum sensing in the research of broadband spectrum sensing method based on CS, A cooperative spectrum sensing method based on compressed sensing is designed. Firstly, a combination of band-pass filter bank and compression sampling method is used to further reduce the sampling rate of wideband signals. Then through the information interaction between the cognitive user and the fusion center, the sub-band that each cognitive user needs to be reconstructed is determined. In this process, each cognitive user only needs to reconstruct part of the frequency band, which greatly reduces the complexity of the signal reconstruction stage. Finally, the reliability of the perceptual method is improved by multi-user cooperative perception.
【学位授予单位】：宁波大学
【学位级别】：硕士
【学位授予年份】：2015
【分类号】：TN911.7
【参考文献】
相关期刊论文 前1条
1 张成;杨海蓉;韦穗;;基于随机间距稀疏Toeplitz测量矩阵的压缩传感[J];自动化学报;2012年08期
，本文编号：2437178
本文链接：https://www.wllwen.com/kejilunwen/wltx/2437178.html
