Spatial-Temporal Cooperative Spectrum Sensing in Flat Fading Channels for Cognitive Radio Using Extend Kalman Filter

This paper investigates a new channel gain map tracking by Space-Time Extended Kalman Filtering (STEKF) for a flat channel, and a novel spectrum sensing via Time Spatial Weighted Non-negative Lasso (TSWNL) algorithm. STEKF enables CRs to estimate and interpolate channel gain map for the entire geographical area of interest with a limited number of CRs measurements. In order to sense primary users (PU) activities, include the transmission power by each PU, location and number of active PUs, TSWNL algorithm is proposed. Numerical results illustrate that the proposed STEKF channel estimation and TSWNL sensing algorithms outperforms linear methods.