Research On Wireless Multipath Channel Estimation Methord Based On Compressive Sensing

With the growing demand for wireless communications, providing high-quality, high-speed data and multimedia services becomes the primary task of the mobile communications. It is a great revolution in mobile communication technology for the change from analog communication system to the digital communication system. The revolutionary improvement includes spectrum efficiency, the system capacity, confidentiality and the call quality. With the development of wireless communications, the proposal of orthogonal frequency division multiplexing (OFDM) has improved the spectrum utilization rate effectively, and reduced the channel fading and channel delay spread effect effectively. The multiple input multiple output (MIMO) technology has improved the system capacity effectively by increase the wireless system coverage using spatial diversity. At present, the research in relay technology could improve the communication quality of the community edge without increase the number of base station.Though the OFDM, MIMO and relay technologies have many advantages, they still face a lot of problems in some specific applications. The OFDM system is sensitive to synchronization error. In the process of signal transmission, the signal gets througth time delay and power attenuation due to the constructions and obstacles nearby. The time synchronization error can cause inter-symbol interference (ISI), while the frequency synchronization error can cause inter-carrier interference (ICI). The interference and distortion could be overcome effectively only the channel information is well understood. When the MIMO or relay is jointed, the ISI and ICI get lager, which influence the system performance seriously. Therefore, the receiver needs to get accurate channel state information (CSI), which is also very important for coherent detection and equalization. Concerned about the above issue, this paper researched the channel estimation of the wireless multipath channel. Specifically, emphatically analyzes the feasibility of the application of compressed sensing theory on wireless multipath channel estimation, mainly studies the channel estimation algorithm of single antenna system, multiple antenna system, two-way relay system and underwater acoustic system based on compressed sensing theory. Corresponding scientific research achievements have been obtained. This thesis mainly completed the following innovative research.First, the classic single antenna system has been studied. The research includes analyzing the channel characteristics, creating the sparse channel model, and so on. The research shows that compressed sensing theory can be used for sparse channel estimation. This method can transform the traditional linear algorithm into nonlinear algorithm, and recover the channel state information in time domain instead of frequency domain. In other word, this method makes full use of the wireless multipath channel inherent sparse characteristic, recovers the channel impulse response by the l1norm minimization, and obtains higher channel estimation accuracy with fewer training signals.Second, compare with the single antenna system, the multiple antenna system has the characteristics of more antenna number, higher data transmission rate, and more complicated channel state. Therefore, it is difficult to obtain accurate channel estimation in low SNR and fast fading conditions. Channel estimation aims for fast and accurate recovery of channel state information. For the multiple antenna system, in the fast fading conditions, the traditional linear channel estimation algorithm, i.e., least squares (LS), has lower computational complexity; however, the channel estimation accuracy is far from satisfied. The compressive sensing-based convex optimizing estimation algorithm, could significantly improve the performance of channel estimation, but implement complex. Based on the advantages and disadvantages of the method mentioned above, this thesis proposed a highly efficient channel estimation algorithm based on the compressive sampling matching pursuit (CoSaMP). For the high-speed multiple antenna systems, this algorithm can recover the channel state information with very high probability even in fast fading conditions. In the meanwhile, both the real-time performance and estimation accuracy have been greatly improved. By using the discrete Fourier transform (DFT), we could extend the application in the channel estimation of MIMO-OFDM systems.Third, the conventional point to point systems have not satisfied people’s demand, therefore, the relay networks have been intensively researched. For the two-way relay system, the two nodes interact and transfer information by the half duplex relay, and there is no direct path for the two nodes. The channel state information changed from the traditional single unknown channel variable into two unknown channel variables. Therefore, the channel estimation process is much more complicated than the point to point systems. Based on the thorough research of the channel characteristics of the two-way relay system, our thesis proposed an adaptive joint channel estimation algorithm based on the compressive sensing theory. Thus, transform the2-dimensional variables into a1-dimensional variable ingeniously, and improve the channel estimation performance of the two-way relay system significantly.Moreover, this thesis is focused on the application of the advanced theory in the practical scene-the research and application of compressed sensing-based underwater acoustic channel estimation. At present, sound wave is the only one carrier for remote underwater information transmission. The underwater acoustic channel is influenced not only by the frequency, but also by the sea movement, height and density and so on. Therefore, underwater acoustic channel is much more complicated than radio channels. This thesis presents a thorough study on the complexity of the underwater acoustic channel, constructs compressed sensing algorithms according to the sparse underwater acoustic channel model and non-relevant observation matrix, and selects the suitable reconstruction algorithm for the underwater acoustic channel estimation. This thesis provides a feasible solution to the channel estimation of underwater acoustic communication system and a new research direction for this field.