Statistical parameters analysis of non-iterative algorithm for reconstruction of Compressive Sampled signals

摘要

Analysis of the parameters for Compressive Sampling reconstruction algorithm is presented in the paper. Compressive sampling allows reconstruction of the signal from a small number of randomly selected samples, it the signal satisfies sparsity property in certain transform domain. The analyzed algorithm is based on non - iterative procedure and allows successful reconstruction if the threshold is properly selected, regardless the number of signal components or the number of measurements taken during the signal sampling. As a measure of reconstruction quality the mean absolute error between original and reconstructed signal is used. The optimal choice of parameters which are used in the algorithm is shown graphically and demonstrated on examples.