In order to solve the modulation recognition problem of single-mixed signal under multiple-input multiple-output orthogonal frequency division multiplexing(MIMO-OFDM)system in non-cooperative communication,this paper proposes a modulation and recognition method based on multi-feature fusion and decision fusion.Firstly,the joint approximate diagonalization of eigenvalue matrix(JADE)algorithm is used to recovery the transmitted signal from the receiving end.The quartic spectrum,cyclic spectrum and higher-order cumulants of the recovery signal are then extracted as input samples.Secondly,the multi-feature fusion network is used to train the input samples.Finally,the decision fusion strategy is used to obtain the final judgment result according to the output of each branch.Simulation result shows that the proposed method can effectively identify seven signals in MIMO-OFDM system,and the recognition accuracy can reach 99.4％at a signal-to-noise ratio of 10 dB.
维普
万方数据