Measurement Matrix Optimization For CS-DCSK UWB Wireless Communication System Based On Chaotic Compressive Sensing

Ultra-wide band(UWB)transmission technology has been widely applied in wireless body area network,internet of things,and so on.As a kind of non-coherent UWB communication systems,compressive sensing based code shifted differential chaos shift keying(CS-DCSK)UWB communication system has been demonstrated to be feasible with a subNyquist sampling rate.In this dissertation,measurement matrix optimization for CS-DCSK UWB communication system based on chaotic compressive sensing is studied to improve bit error ratio(BER)performance.The main works are summarized as follows:(1)Taking the Walsh-code structure of the CS-DCSK UWB signal into account,this paper firstly proposes a novel sparse cyclic measurement matrix.Compared with the existing measurement matrix,the proposed measurement matrix has stronger noise removal ability and better column orthogonality.Moreover,the merit of the generalized rotation matrix is introduced to further reduce the correlation between the columns.Simulation results indicate that when the compression ratio is 0.5,the compressive sensing based CS-DCSK UWB system using the proposed measurement matrix has about 2.0dB BER performance improvement,which has almost the same performance as the conventional system sampled a full Nyquist rate.(2)The optimization of two important system parameters(i.e.guard interval and integration interval)is studied based on the theoretical formula of BER performance.The numerical results display that the optimal guard interval of compressive sensing based CSDCSK UWB system is linear with that of the conventional system sampled a full Nyquist rate and the slope depends on the compression ratio.An adaptive optimization algorithm for integration interval is proposed.Simulation results indicate that the proposed optimization algorithm has about 1.0 dB BER performance improvement for the conventional Nyquistsampled system.However,since each element of the low-dimensional signal contains information of the primitive signal after compressed by the measurement matrix,this optimization algorithm cannot be applied to the compressive sensing based system.(3)A sparse signal-matched measurement matrix is proposed to realize the optimization of the integration interval from the view of compressive sensing.The proposed sparse signalmatched measurement matrix has stronger noise removal ability and captures more useful signal energy,thereby possessing better BER performance.Simulation results show that the compressive sensing based CS-DCSK UWB system using the sparse signal-matched measurement matrix has almost the same performance as the Nyquist-sampled system using the optimal integration interval when the compression ratio is 0.5.