Experimental Study of Spectrum Estimation and Reconstruction based on Compressive Sampling for Cognitive Radios

This paper addresses the experimental study of the wide band signal estimation and reconstruction using the established compressive sampling (CS) methods. For this purpose, a hardware test bed was setup inter-connecting a wide band SDR based hand held military radio (SWAVE HH or HH), vector signal generator, bi-directional coupler, attenuators, PC and other auxiliaries. Real-world communication signals were created by the signal generator and SWAVE HH was used to scan these signals. The discrete samples from the HH were collected on PC for reconstruction and application of CS. It was shown that good reconstruction of the acquired wide band signal is possible with sub-Nyquist rate sampling by means of signal reconstruction under CS framework. In the end, mean squared error (MSE) performance is shown to indicate better estimation and reconstruction of the signal with higher compression rate and higher sparsity.