Spectrum Sensing in the Presence of RF Impairments in Cognitive Radio

The spectrum detection performance of a matched filter (MF) in the presence of carrier frequency offset (CFO), phase noise (PN) and time offset is analyzed. Detection performance is derived theoretically and confirmed through simulation in the presence of CFO. The results have shown that in the presence of CFO, the MF may be outperformed by an energy detector (ED) over a range of CFO values. The authors propose three spectrum sensing techniques that are robust to CFO. The first technique is called the block-coherent detector (BLCD) with a suboptimal number of blocks (N/2). The second technique is called second-order matched filter-I (SOMF-I), the detection performance of which has been studied both theoretically and confirmed through simulation. The last technique is named the second-order matched filter-II (SOMF-II) and is a modified version of (SOMF-I) but with a superior performance. The presence of PN and time offset and their effect on the detection performance is then examined via simulation. The results have shown that PN has not affected the detection performance of SOMF-I and has only a slight effect on the MF performance. But the time offset has seriously degrades the MF, ED and SOMF-I. Also, the results have shown that the performance of SOMF-I is much less affected by noise uncertainty compared to the ED. Finally, the throughput of SOMF-I has been studied in the presence of CFO and the results have shown that its performance is only marginally degraded compared to the MF.