Abstract
Compressed sensing CS has been an effective research area which it plays an efficient role in many applications such as cognitive radio, imaging, radar and many other applications. The main part of CS system is to recover an input signal by using minimum number of samples than used by conventional (Nyquist) sampling. In this paper, the minimum number of measurements and the number of sparsity level used to reconstruct the signal with minimum error, computational complexity and time will be optimized. Moreover, different recovery algorithms such as convex optimization, greedy algorithms, iterative hard thresholding and hard thresholding pursuit algorithms are used for optimal recovery of sparse signal from a small number of linear measurements. Furthermore, the effect of using CS as denosing will be investigated. Then comparisons study between CS as denoising and conventional denoising process will be made to reduce the effect of noise on the signal.