Recovery algorithms of compressively sampled data include solving a sparse approximation problem that requires iterative search or optimization techniques. Software implementations of these algorithms are not fast enough for real-time applications. In this research, two different hardware approaches are used to sample and then recover of sparse signals. For high sparse signals, orthogonal matching pursuit (OMP) recovery algorithm is implemented in the hardware using a deterministic measurement matrix. The construction of the matrix is based on the parity check matrix of LDPC codes. Cyclic and binary structure of this matrix leads to the lower computational complexity and hardware cost. For low sparse signals, a hardware architecture of iterative method with adaptive thresholding (IMAT) is presented to recover a sparse signal from its random samples. To demonstrate the effectiveness of IMAT, a comparison is performed between the IMAT algorithm and OMP algorithm, in terms of complexity and reconstruction quality. Since IMAT employs discrete transform in each iteration, two multiplication-free transforms, Walsh–Hadamard transform (WHT) and approximate DCT, are used to reduce computational complexity of its implementation. Both OMP and IMAT are implemented on Virtex6 FPGA and the results are reported in terms of hardware resource utilization, power consumption, and recovery time. Key words Compressive sensing, deterministic, measurement matrix, OMP, random sampling, IMAT.
