Experimental SDR implementation of cooperative spectrum sensing in cognitive radio networks

Abstract

In this paper, a Software Defined Radio (SDR) implementation of Cooperative Spectrum Sensing (CSS) is investigated. More precisely, various local spectrum sensing algorithms have been implemented and validated on USRP B210 using the Qt framework. The realized test bed is based on a client-server architecture. In the client interface, different spectrum sensing techniques have been implemented and analyzed. The fusion center interface implements all hard decision combination schemes, namely OR, AND, and the MAJORITY rule. Finally, several realistic scenarios have been considered to validate our SDR implementation. Moreover, these implementations allowed us to assess the efficiency of different hard-decision combination schemes for cooperative detection in the context of cognitive radio. Besides, the performance improvement is analytically evaluated and corroborated by the experimental results. These results regarding real scenarios confirm that Goodness of Fit-based (GoF) sensing methods outperform conventional methods. Furthermore, theoretical and simulation results from CSS have been validated experimentally that OR-rule gives clearly better results than other hard decisions combinations schemes.