Received signal interpolation for context discovery in Cognitive Radio

This paper addresses the context acquisition to characterize different features of a primary network based on the power measurements obtained by a sensor network. In particular, it presents a comparative study of the impact of natural neighbor, linear, and nearest neighbor interpolation techniques carried out over the measurements at different geographical positions. Extracted features include transmitter position, antenna pattern or propagation model. An evaluation is carried out in scenarios including the effect of both correlated and non-correlated shadowing.