Machine Learning Meets Point Process: Spatial Spectrum Sensing in User-Centric Networks

This letter introduces a novel machine learning (ML)-based approach to approximate the distributions of the aggregated interference power in wireless networks. We focus on the application of spatial spectrum sensing (SSS) in user-centric networks where Poisson cluster process (PCP) is used to model the primary users. A nonlinear regression method, i.e., kernel regression, is introduced to learn the distributions of the aggregated interference power of the PCP modeled primary user network. Simulation results demonstrate the accuracy of our approach.