Spectrum sensing for cognitive radios using Kriged Kalman filtering

A cooperative spectrum sensing algorithm for cognitive radios (CRs) is developed using the novel notion of channel gain maps. These maps capture the spatio-temporal variation of the RF propagation in the geographical area where the CR network is operated. They are tracked via Kriged Kalman filtering (KKF), a tool with well-appreciated merits in geo-statistics. This in turn enables the activity of an unknown number of primary users to be tracked using a sparse regression technique based on a weighted least-squares criterion regularized by the ℓ₁ norm of the regression coefficient vector. Simulations demonstrate considerable performance advantage of the proposed scheme over a crude path loss-based sensing algorithm.