Cooperative spectrum sensing by algorithmic techniques using spatial information

Spectrum sensing is the basic step in cognitive radio networks. Its accuracy directly affects spectrum utilization. In this paper we propose a new method for blind cooperative spectrum sensing using spatial information. We model our sensing matrix as a graph, taking advantage of signal and spatial correlation of antennas. We prove that the general settings of the problem are hard to approximate. Taking spatial information into account, we propose a new scheme to find white space subregions. Cognitive radio users residing in the region may be able to transmit, but must keep their transmission power to the extent that no interference is caused to primary receivers out of the region.