Spectrum Sensing for Cognitive Radios with Transmission Statistics: Considering Linear Frequency Sweeping

The spectrum sensing performance of Cognitive Radios (CRs) considering noisy signal measurements and the time domain transmission statistics of the Primary User (PU) is considered in this paper. When the spectrum is linearly swept in the frequency domain continuously to detect the presence of the PU the time-domain statistics of the PU plays an important role in the detection performance. This is true especially when the PU's bandwidth is much smaller than the CR's scanning frequency range. We model the transmission statistics that is the temporal characteristics of the PU as a Poisson arrival process with a random occupancy time. The spectrum sensing performance at the CR node is then theoretically analyzed based on noisy envelope detection together with the time domain spectral occupancy statistics. The miss detection and false alarm probabilities are derived from the considered spectral occupancy model and the noise model, and we present simulation results to verify our theoretical analysis. We also study the minimum required sensing time for the wideband CR to reliably detect the narrowband PU with a given confidence level considering its temporal characteristics.