Secure Prioritized Inter Cluster Channel Selection for Tracking User Attack in Cognitive Radio Networks

Abstract

Dynamic spectrum allocation evolved in order to allow the spectrum holes to be occupied by secondary users (SUs), but they have to handoff when primary users (PUs) come back, which brings new security problems like Tracking User Attack (TUA) into consideration. TUA might block the control information shared by a Cognitive Radio Network (CRN), if the attacker predicts the control channel frequency based on traffic analysis. This is a typical Denial of Service (DOS) attack preventing users from handoff to available channels. In cluster based channel selection, cluster heads will receive the channel information, about two users at a time and transmit the common preferred channel that was allocated to the base station. The base station selects a channel and returns it to the cluster head which returns them to the requesting CR users. For secure channel selection, the information could be encrypted by the base station and decrypted at the cluster heads to prevent DOS attacks. In this work, we propose selection of available channels based on the priority of the cluster which we term Prioritized Inter Cluster Channel Selection (PICCS). The priority is essentially the history based reputation of the cluster calculated from the attackers in the cluster. Thus, instead of a random choice of a cluster for the available channel, during inter-cluster channel selection, we select a channel from the best available cluster. We observe the effect of such selection on both the delay and the handoffs in the presence of attackers and compare it with the channel selection scheme based on the nearest available cluster.