Narrowband Spectrum Sensing of a 60Hz Bandwidth Signal in AWGN 

Context in source publication

Context 1

... of samples and T the sampling time. The performance of each ED was investigated for both narrow and wideband sensing as follows: For sparse wideband sensing : Two PU signals with equal amplitudes A and A were varied to achieve low and high SNR conditions with respect to AWGN; each had frequencies f = 500 Hz ; f = 2000 Hz respectively in a sensing span of 4KHz as shown in Figure 2 using the SP technique. It is noted that this dataset was generated for each ED technique examined here. We note that this sparsely dense spectrum (Figure 2) was simulated to represent obtainable real-life sensing scenarios with wideband frequency occupancy of about 10% at threshold of - 15dBm . Also, though low frequencies were used here, such low frequencies could easily fit for typical low baseband frequencies or down sampled versions of the high frequency signal; therefore, the concept is adaptable for higher frequency bands. For narrowband sensing : A typical PU signal of bandwidth W  60 Hz was simulated for a 100Hz narrowband sweep as shown in Figure 3. This sample result is shown for the SP technique; however, same process was applied to other techniques examined here. For analysis, the hypothesis testing for the case of signal plus noise generally expressed as H : y ( n ) = s ( n ) + w ( n ) ...