Research On SAMP For Nonlinear Distortion Estimation And Narrowband Interference Detection

MIMO-OFDM has become the core technology of the fourth generation communication system because of the ability of improving the utilization of spectrum resources and resisting the frequency selective fading channel.For a broadband system like MIMO-OFDM,signals in such systems often have high peak to average ratio,which will bring in nonlinear distortion when a power amplifier is operating on them,and also comes the narrowband interference when sharing the frequency band with other systems.With the number of the broadband signals to be processed in unit time badly increasing,the nonlinear distortion estimation and the narrowband interference detection in the receiver becomes more and more difficult.Fortunately,the compressed sensing theory,which can reconstruct the original sparse signals from a small amount of samplings containing the complete signal information using the specific reconstruction algorithms,provides an effective solution for solving these problems.Therefore,the nonlinear distortion estimation and the narrowband interference detection based on compressed sensing is becoming the research hotspot in wireless communication.But the existing reconstruction algorithms still cannot meet the real-time requirements in MIMO-OFDM due to the high complexity and the poor convergence.The aim of this dissertation is to improve the performance of the recover algorithms in solving nonlinear distortion and narrowband interference problems.After analyzing the existing reconstruction algorithms in CS,this article applies the SAMP algorithm to solve the above two problems for a better performance.The main work in this dissertation contains the following aspects:Firstly,on the basis of the sparsity of the nonlinear distortion in time domain,SAMP is introduced to estimate the nonlinear distortion in OFDM after analyzing the shortcomings when using the basis pursuit algorithm;Then SAMP is introduced to solve the nonlinear distortion estimation in MIMO-OFDM after the theoretical analysis and modeling in time domain of the nonlinear distortion.Secondly,the SAMP,which can converge to the true sparsity of signal without knowing it quickly,is very suitable to detect the narrowband interference in MIMO-OFDM to have a better performance than the existing reconstruction algorithms.