Research On Data Compression And Reconstruction Algorithm Based On Compressed Sensing And FPGA Implementation

Due to the limitations of the transmission interface efficiency,the processing speed of the chip and the memory capacity,the ultra-high speed data acquisition system is faced with the problems of mass capacity data transmission and real-time storage in the process of implementation.In view of the above problems,this paper proposed a data compression method and FPGA hardware implementation based on Compressed Sensing(CS)theory,and conducted in-depth research on the orthogonal matching pursuit(OMP)type algorithms in Compressed Sensing reconstruction algorithm.The main contents of this thesis are as follows:First of all,based on Compressed Sensing theory,this paper compresses the sampled data in the data acquisition system according to the idea that CS obtains measured values through non-adaptive linear projection operations,and uses Verilog HDL language and VIVADO software to complete data compression method of FPGA hardware circuit design,the simulation results verify the correctness and effectiveness of the hardware design.Secondly,based on the theory and simulation analysis of OMP type algorithms,this paper proposed an improved algorithm,called preselected stage-wise orthogonal matching pursuit(PSt OMP)algorithm,which requires the secondary selection of the number of fixed steps for the pre-selected atoms that meet the threshold requirements.Compared with other traditional matching pursuit algorithms,the improved algorithm can recover the original signal under the condition that the signal sparsity is not known before.Compared with other traditional matching pursuit algorithms,the improved algorithm can restore the original signal without knowing the signal sparsity.The simulation results show that the PSt OMP algorithm can obtain a higher reconstruction accuracy than the stage-wise orthogonal matching algorithm,and the reconstruction time of the algorithm is much lower than that of the sparsity adaptive algorithm.Finally,in view of the problem that the reconstruction effect of the PSt OMP algorithm is easily affected by the parameters,using particle swarm optimization algorithm for solving the optimal parameters,refactoring makes PSt OMP algorithm to improve reconstruction performance.Through a lot of simulation experiments,the optimal values of the parameters and the recommended selection range of the PSt OMP algorithm under different conditions are given,and the feasibility of parameter optimization is verified.