Research On Sparse Representation Separation Method For Time-Frequency Overlapping Commun Ication Signals

With the rapid development of radio communication technology,electromagnetic environment becomes more and more complex,frequency spectrum resources become more and more limited,and time-frequency overlapping signals widely exist.In the received signal,the useful signal and the interference signal are mixed together in the time domain,frequency domain or time frequency domain in the form of partial or complete overlap.The traditional methods of signal interference suppression and separation are often ineffective.In this thesis,the interference suppression and separation of time-frequency overlapping communication signals based on signal sparse representation are studied.Based on the signal sparse representation theory,this thesis,makes an exploratory study on the over-complete dictionary construction and signal sparse decomposition of time-frequency overlapping communication signals in signal sparse representation.The main contents of this thesis are as follows:1.The related theories of signal sparse representation are described,i ncluding signal sparse representation model,dictionary construction of signal sparse representation and decomposition algorithm of signal sparse representation.Theoretical derivation and computer simulation of the classical K-SVD dictionary construction algorithm and OMP sparse decomposition algorithm.2.Aiming at the problem of interference suppression and separation of time-frequency overlapping communication signals,a method of interference suppression and separation of overlapping signals based on joint dicionary is proposed.In view of the di fficulty of sparse decomposition of overlapping signals in a single overcomplcte dictionary,the idea of dictionary union is appied to overcomplete dictio,nary construction,and the original K-SVD dictionary construction algorithm is improved.3.The interference suppression and signal reconstruction of single-carrier signals such as 2ASK and 2PSK and multi-carrier signals such as OFDM are performed.Based on common learning dictionary and joint learning dictionary algorithm,the time-frequency overlapping single carrier signal and multi-carrier signal under different signal-to-interference ratios are sparsely decomposed,and the waveforms,spectrograms,constellation diagrams and bit error rate of the signals before and after the sparse decomposition are analyzed by computer simulation.The results show that the algorithm based on joint learning dictionary can realize the interference suppression and separation of communication signals with time-frequency overlap.And the larger the signal-to-interference ratio of the overlapping signals,the better the separation effect.This thesis provides a theoretical basis for future research based on signal sparse representation separation.