Communication reconnaissance complex signal separation based on independent component analysis
Xudong Zhang · Journal of Tsinghua University(Science and Technology) · 2010
Traditional communication reconnaissance has limited separation capacity due to complex multiple signals.A blind reconnaissance technology is presented,which uses independent component analysis(ICA) combined with preprocessing and post-processing.The system first de-noises the noisy mixtures with discrete wavelet transforms.The signals are then separated using the ICA.Finally,the constant modulus algorithm(CMA) is used for post-processing to refine the output signal.Simulations show that excellent separation performance is achieved without any prior information.