Sampling Of Non-cooperative Ultra-wideband Signal Based On Compressed Sensing

As a new-type wireless communication technology, ultra wideband(UWB) has already been an extraordinary important content of research in the area of signal processing. With bandwidth of UWB signal going more and more widely, there is enormous challenge in capability of signal processing and design of hardware when the sampling of UWB signal is implemented based on the Nyquist theorem. In recent years, Compressed Sensing(CS) breaks the limit of traditional sampling theorem for sparse signal acquisition. But in non-cooperative environment, we are short of signal of a priori information, and then it appears troubles in application of compressed sensing theory. This paper studies the data acquisition of UWB signal in non- cooperative environment, it presents the hybrid sampling method based on signal projection(HSSP) based on compressed sensing. On the basis of the HSSP, and further simplify the structure of the sampling system, designed parallel sampling method based on orthogonal projection(PSOP).1. In view of the UWB signal compression sampling problems in non-cooperative environment, we present the HSSP, its concrete steps are as follows.(1) With analog information converter based on CS for the random measuring of UWB signal.(2) The orthogonal measuring of UWB signal in frequency domain, reconstruction signal roughly with making use of projection coefficient and linear combination of basis function.(3) Using principal component analysis(PCA) for K-L decomposition to achieve eigenvectors of autocorrelation matrix.(4) In eigenvector matrix as a signal sparse domain to reconstruction the signal again in compression perception theory framework. The structure can sample slowly and restrict precisely to UWB signal in non-cooperative environment.2. To make the structure of sampling system simplified, lower complexity of the system, we present the PSOP. The design idea of the structure is analyzing the parallel projection based on random sampling structure functionally consistent with the AIC, and Use the parallel projection based on random sampling structure and the frequency domain projection of the sample structure of the similarity in structure, optimize the combination of the two together, finally get the PSOP. There is no difference between this structure and HSSP, both of them can make sampling slowly and reconstruction precisely to UWB signal in non-cooperative environment. But this structure has higher compression ratio of signal samples.Through simulation tests, both of the HSSP and the PSOP can reconstruct signal precisely. It can be used in the low speed of UWB signal in non-cooperation environment. In order to further the performance analysis of two kinds of sampling structure, the precision and error for reconstruction actual evaluation research, put forward to further improve the sampling in order to improve the performance of the structure, found the deficiency on the algorithm, the improved measures are put forward.