Optimal and Sub-Optimal Spectrum Sensing of OFDM Signals in Known and Unknown Noise Variance

摘要

We consider spectrum sensing of OFDM signals in an AWGN channel. For  the case of completely known noise and signal powers, we set up  a vector-matrix model for an OFDM signal with a cyclic prefix and  derive the optimal Neyman-Pearson detector from first  principles. The optimal detector exploits the inherent correlation  of the OFDM signal incurred by the repetition of data in the cyclic  prefix, using knowledge of the length of the cyclic prefix and the  length of the OFDM symbol. We compare the optimal detector to the energy  detector numerically. We show that the energy detector is  near-optimal (within 1 dB SNR) when the noise variance is  known. Thus, when the noise power is known, no substantial gain can  be achieved by using any other detector than the energy detector. For the case of completely unknown noise and signal powers, we  derive a generalized likelihood ratio test (GLRT) based onempirical second-order statistics of  the received data. The proposed GLRT detector exploits the  non-stationary correlation structure of the OFDM signal and does not  require any knowledge of the noise power or the signal power. The  GLRT detector is compared to state-of-the-art OFDM signal detectors,  and shown to improve the detection performance with 5 dB SNR in  relevant cases.