The Design And Implementation Of Adaptive Signal Reconfiguration System Base On Compressive Sensing Theory

With the growing demand for access to information, the signal of bearing information is also becoming more complex, its bandwidth is also becoming more and more big, the system sampling frequency is also more and more high; if still using basic Shannon sampling theorem, the design of data acquisition system to get information, is bound to face many difficult to overcome the problem. In order to solve this series of problems, the scientific research workers proposed the time interleaved sampling structure and periodic non uniform sampling structure, and the MWC sampling system and AIC sampling system which based on the theory of compressed sensing. However, the hardware design of sampling structure is too complicated, the accuracy requirement is high and the cost of production is relatively high, which brings some obstacles to the research work of the research workers. Therefore, the main works done in this paper are as follows:1. Based on the compressive sensing theory and the spatial sampling structure model, constructing the adaptive reconfiguration model for this paper, on the basis of this, designing an adaptive signal reconstruction system, with the low sampling rate, collecting the high frequency signal, and giving the overall scheme design of the whole system, which mainly includes the following sampling acquisition channels, signal adaptive tracking, signal reconstruction.2. The acquisition channel design and error calibration, combining with the system performance requirements, selecting the right ADC chip, building the acquisition channel; in order to reduce the complexity of hardware circuit design, the channel error calibration is performed by using the digital background calibration technique; at the same time, combining with the corresponding time difference estimation algorithm, estimating the time difference between the two signals, and then continuously adjusting the parameters which setting the sampling clock in clock circuit, to achieve accurate relative delay of two sampling clocks, for the construction of the whole system is precondition for sampling matrix.3. The adaptive tracking signal, according to the spatial of the sampling theory, dividing the signal space into multiple sub space, and representing the signal subspace corresponding to the signal subspace by the corresponding function; in the sample space, combining with a set of sampling function, deducing the sampling matrix of the system and the reconstruction function of subspace signals corresponding to the system; using compressed sensing in orthogonal matching pursuit algorithm, and combining with the derived sampling matrix and the acquisition signal, tracking the corresponding frequency band of the acquisition signal subspace.4. The signal reconstruction, on the basic of the basic interpolation reconstruction model, using the subspace signal corresponding to the reconstruction function and the multiphase interpolation filter structure, achieving a complete reconstruction of the signal subspace in PC terminal software, and the signal reconstruction after the sampling rate is equivalent to 10 GSPS.Based on the ARM and FPGA hardware platform and the PC software platform, realizing the whole signal adaptive reconfigurable system, and access the true signal source, testing and verifying respectively the acquisition channels, the signal tracking module and the signal reconstruction module, meeting the design requirements.