Real-time Cooperative Spectrum Sensing: Reliable Localization of Unknown Interferers in Shared Spectrum

Abstract:
Locating active users/transmitters and sources of spectral interference in real-time is a key challenge in next generation radio networks to enable robust wireless communication and highly dynamic radio resource allocation. We present and investigate a spectrum sensing network with multiple collaborating spectrum sensing units (SUs) which jointly find precise locations of active wireless transmitters in an observed area. The estimation of locations is based on measurements of received signal strength at SUs, taking into account that the measurements are perturbed by spatially correlated shadowing. We propose four localization methods for the cases of single and multiple active transmitters, namely, maximum likelihood search, generalized least squares, radio environment map (REM) based and compressed sensing based methods. The methods vary in their estimation accuracy and complexity. Simulation results present effects of various parameters on the localization performance and also show that there is no significant performance loss due to discretization.