Research On Array Signal Processing Under Channel Inconsistency And Mutual Coupling

Smart antenna system plays an important role in wireless communication.It adaptively changes the pattern by combining array antenna and array signal processing technology to ensure normal communication between users and base stations.In the research of array signal processing,classical DOA estimation and adaptive beam-forming(ABF)algorithms are all discussed in ideal cases.Taking into consideration of channel inconsistency and mutual coupling of array elements,thesis introduces gain and phase calibration,array decoupling,DOA estimation and adaptive beam-forming algorithms,and implements these algorithms in engineering setting.Our contribution is as following:Firstly,in terms of gain and phase calibration algorithm,we proposed a single frequency point version based on FFT for narrow-band signals.Combining with the gain and phase calibration module of digital multi-beam forming,we also introduce the implementation of calibration algorithm in Engineering setting.The module calculates the calibration coefficient mainly by DSP and completes the gain and phase calibration by FPGA.The results of hardware operation prove the feasibility of the algorithm.Then,in the research of array mutual coupling matrix and DOA,a cascade estimation method is proposed.Based on the property of mutual coupling matrix,the position matrix and mutual coupling coefficient are separated using transformation matrix,which achieves the purpose of "decoupling".The algorithm has low complexity and good direction finding and decoupling performance.For broadband signals,combining with the 5G Massive MIMO base station decoupling,we design a decoupling method based on cell pattern reconstruction in the digital domain,where the decoupled array is estimated by ISM algorithm.Computer simulation results illustrate the efficacy of the decoupling algorithm.Finally,we introduce the classical LMS algorithm in non-blind adaptive beam-forming.For broadband signals,ISM and CSM adaptive beam-forming techniques are proposed by using DOA estimation for broadband sources.Combining with the beam-forming module of the digital multi-beam forming project and considering the specific application scenarios,we design a "blind adaptive" beam-forming technology: The system updates the weights and adjusts the working state in real time according to the feedback information of the receiving demander.It forms a beam-forming through the cooperation of DSP and FPGA to find the satellite in real time,and tracks the satellite by conical scanning once it is found.The whole process is adaptive.This technology has the advantages of less computation and real-time implementation.It can change the way of calculation of weights to suppress interference even in the presence of strong interference signal in space.The results of computer simulation and hardware operation show the efficiency of the algorithm.