Distributed and centralized test approaches to reinforce distributed spectrum sensing in cognitive radio networks

To reliably detect primary users (PUs) even on the existence of compromised nodes generating forged sensing reports, secure distributed spectrum sensing (DSS) schemes in cognitive ratio networks (CRNs) have been proposed. However, they have the limitation of sensing accuracy for the existence of PU signals due to the absence of exact signal patterns of PUs. It is caused by FCC restriction (no modification of) on PUs, and thus the CRNs cannot communicate with PUs in order to obtain such patterns. To address this challenge, we propose a test framework utilizing primary user emulation signals that can be applied to existing DSS schemes for reinforcing the robustness against forged sensing values. The proposed framework can be applied using two approaches: distributed and centralized. We evaluate these approaches through simulation in comparison with the existing scheme. The results show that our approaches improve the sensing accuracy and fusion speed in the case of attack, and that the centralized approach, with an extra system, outperforms the distributed approach.