LRT Detectors for Spectrum Sensing of Weak OFDM Signals With Periodic Pilots

Orthogonal frequency division multiplexing (OFDM) signals are widely employed in most wireless comm- unication systems, and the problem of spectrum sensing for OFDM signals has been extensively investigated. However, most of the available research reports on spectrum sensing for OFDM signals ignore the existence of pilots, which are usually inserted in the OFDM signals for synchronization, channel estimation and control information delivery. To sense weak OFDM signals with periodic pilots reliably in cognitive radio networks, we propose two likelihood ratio test (LRT) detectors for the scenarios with and without time synchronization information. In the ideal scenario that an unlicensed secondary user is perfectly synchronized with a licensed primary user, we propose a near-optimal LRT (NOLRT) detector, which constructs the test statistic by correlating the locally generated pilots with the received OFDM signal samples. Since the NOLRT detector relies on perfect time synchronization that is difficult to realize in practice, we further propose a more practical buffer-aided LRT (BALRT) detector that does not require time synchronization. To formulate the test statistic, the BALRT detector stores the averages of historical OFDM signal samples in a buffer and then correlates them with the currently received OFDM signal samples. Theoretical closed-form expressions for the probability of detection and probability of false alarm are derived for the proposed detectors. To alleviate the negative effects of noise uncertainty in low SNR scenarios, we also propose an effective method based on noise power estimation. Monte Carlo simulation results match with theoretical results very well and show that the NOLRT detector outperforms the counterparts significantly in additive white Gaussian noise (AWGN) channel with perfect time synchronization, while the BALRT detector outperforms the counterparts noticeably in multipath propagation environments without time synchronization.