Upper bounds for Neyman-Pearson cooperative spectrum sensing

We consider a cooperative spectrum sensing scenario where the local sensors at the secondary users are viewed as one-level quantizers, and the quantized data are to be fused under Neyman-Pearson (N-P) criterion. We demonstrate how the N-P fusion results in a randomized test, which represents the total performance of our spectrum sensing scheme. We further introduce an upper performance bound for the overall primary user signal detection. An analytical procedure towards the upper bound and its relevant quantization setup at the local sensors are proposed and examined through simulations.