Energy Detection Under Interference Power Uncertainty

This letter concerns energy detection (ED) under interferences from surrounding users, where the challenge is that the distribution of the aggregate interference power has an unbounded support. With the assumption of log-normal large-scale fading of interferences and high interference-to-noise ratio, we derive an approximate but explicit expression for the decision threshold of ED that guarantees a given average false-alarm probability, and analyze its signal-to-interference ratio (SIR) wall. Simulation results verify the theoretical results and demonstrate that the proposed ED has a lower SIR wall and a better performance than the bounded worse behavior-based ED and the estimated background (interference and/or noise) power- based ED.