Reputation-Based Spectrum Data Fusion against Falsification Attacks in Cognitive Networks

Abstract

The cognitive radio network paradigm increases spectrum usage efficiency by allowing secondary users to perform shared access to licensed spectrum. This systematic improvement may be obtained in a practical way by implementing a distributed cooperative spectrum sensing mechanism. Although such decentralized sensing offers many advantages, it also opens the door to new security threats such as spectrum sensing data falsification attacks. In this work, we design a new mechanism that exploits sensing correlation through the concept of reputation to enhance resilience against this type of threat. By both theoretical analysis and simulations, we show that our proposal provides incentives for cooperation among honest devices and reduces the spectrum occupancy assessment error rate in the presence of malicious users.