Compressed sensing based jammer detection algorithm for wide-band cognitive radio networks

This paper proposes a new algorithm for jammer detection in wide-band (WB) cognitive radio networks. We consider a WB which comprises of multiple fixed length narrow-band sub-bands (SB). These SBs are occupied by narrow-band signals which can be legitimate users or a jammer. To reduce the overhead of the analog-to-digital conversion (ADC), compressed sensing (CS) is performed first. CS allows us to estimate a WB spectrum with sub-Nyquist rate sampling. After that, energy detection is applied to identify the occupied sub-bands (SB). Then, for each occupied SB, some waveform parameters such as signal bandwidth and power spectral density (PSD) levels are compared with licit user database to classify the observed signal as a licit user or a jammer. In the end, performance of the proposed algorithm is shown with the help of monte carlo simulations under different empirical setups.