Mitigating SSDF attack using k-medoids clustering in Cognitive Radio Networks

Collaborative sensing is preferred to individual sensing in Cognitive Radio Network (CRN) since it helps in achieving a more accurate sensing decision. In infrastructure-based cognitive radio network, each node sends its local sensing report to the fusion center which uses a fusion rule to make the final decision. The decision of the Fusion Center plays a vital role. Attackers may try to manipulate the decision-making of the Fusion Center (FC) for selfish reasons or to interfere with the primary user transmission. In SSDF attack, malicious users try to manipulate the FC by sending false sensing report. In this paper we present a method for detection and isolation of such malicious users. Our method is based on the k-medoids clustering algorithm. The proposed approach does not require the use of any predefined threshold for detection. It mines the collection of sensing reports at the FC for determining the presence of attackers. Additionally, we also present how we can use the proposed approach on streaming data (sensing reports) and thereby detect and isolate attackers on the fly. Simulation results support the validity of the approach.