An MDP Approach for Defending Against Fraud Attack in Cognitive Radio Networks

作者：Shahhoseini Hadi Shahriar^*; Jafari Amir Hosein; Afhamisisi Khadijeh

来源：IETE Journal of Research, 2015, 61(5): 492-499.

DOI：10.1080/03772063.2015.1023749

• Cognitive radio
• Secondary users
• MDP
• Fraud attack

Cognitive radio networks (CRNs) have been introduced in recent years to solve frequency leakage. In these networks, several challenges such as Lion and jamming attacks, giving false sensing information, low accuracy in detection of primary users, and sensing error can lead to decrease in CRN's performance. Selecting appropriate free bands by secondary users (SUs) can prevent interference and increase frequency use. In this paper, we present a decision making method based on reinforcement learning to find the optimal strategy for SU in best channel selection. In our model, effect of collision fraud and sensing error is considered simultaneously. In the proposed model, to overcome falsification of sensing reports, the abstract of spectrum sensing data falsification attack is extracted and modelled by using model-based reinforcement learning method. We formalize different states of system, actions of SUs, and probability transitions between states in the framework of Markov decision process (MDP). The proposed model is evaluated by numerical results and optimal strategies according to different effective parameters in network performance are studied.

• 出版日期2015-9-3