Signal Assemblage Based Wideband Spectrum Sensing for OFDM Cognitive Radios

This paper investigates power-law detector (PLD) based wideband spectrum sensing schemes for OFDM cognitive radios. By exploiting signal assemblage (SA) feature of the primary user (PU), we first propose a heuristic iterative threshold test (HITT) to roughly detect the SA set boundary of the noise and PU signal; Subsequently, a signal assemblage cardinality measuring (SACM) method is developed to refine the HITT decisions. Operating in SA-based wideband sensing manner, the PLD greatly overcomes the ambiguity in differentiating weak PU subcarriers from noise and significantly meliorates the sensing performance, compared with the non-SA methods.