Principal-Minimum Eigenvalue Algorithm for Signal Sensing

This paper proposes a signal sensing method based on principal components to minimum eigenvalue of the received signal. The signal model and theoretical performance analysis of the principal-minimum eigenvalue (PME) algorithm are provided in detail. A major advantage of the PME algorithm is that it does not need any prior information of noise or signal, making it highly promising in scenarios with uncertain noise power. To evaluate the performance of PME algorithm, simulations are made and compared with state-of-the-art algorithms, including energy detection (ED) algorithm, energy to minimum eigenvalue (EME) algorithm, maximum-minimum eigenvalue (MME) algorithm and arithmetic to geometric mean (AGM) algorithm. Simulation results show that the PME algorithm has superior performance under very low signal-noise-ratio (SNR) signals, the performance of PME algorithm is close to the MME algorithm, and has advantages over other methods.