Robust cooperative sensing via state estimation in cognitive radio networks

Abstract

Cooperative sensing, a key enabling technology for dynamic spectrum access, is vulnerable to various sensing-targeted attacks, such as the primary user emulation or spectrum sensing data falsification. These attacks can easily disrupt the primary signal detection process, thus crippling the operation of dynamic spectrum access. While such sensing-targeted attacks can be easily launched by an attacker, it is very challenging to design a robust cooperative spectrum sensing scheme due mainly to the practical constraints inherent in spectrum sensing, particularly the shared/open nature of the wireless medium and the unpredictability of signal propagation. In this paper, we develop an efficient, yet simple attack detection framework, called IRIS (robust cooperatIve sensing via iteRatIve State estimation), that safeguards the incumbent detection process by checking the consistency among sensing reports via the estimation of system states, namely, the primary user's transmit-power and path-loss exponent. The key insight behind the design of IRIS is that the sensing results are governed by the network topology and the law of signal propagation, which cannot be easily compromised by an attacker. Consequently, the sensing reports must demonstrate consistency among themselves in estimating system states. Our analytical and simulation results show that, by performing consistency-checks, IRIS provides high attack-detection capability, and preserves satisfactory performance in estimating the system states even under very challenging attack scenarios. Based on these observations, we propose a new incumbent detection rule that can further improve the spectrum efficiency. IRIS can be readily deployed in infrastructure-based cognitive radio networks, such as IEEE 802.22 WRANs, with manageable processing and communication overheads.