Intelligent spectrum management techniques for wireless cognitive radio networks

Abstract

This thesis addresses many of the unique spectrum management chal- lenges in CR networks for the rst time. These challenges have a vital e ect on the network performance and are particularly di cult to solve due to the unique characteristics of CR networks. Speci cally, this thesis proposes and investigates three intelligent spectrum management tech- niques for CR networks. The issues investigated in this thesis have a fundamental impact on the establishment, functionality and security of CR networks. First, an intelligent primary receiver-aware message exchange protocol for CR ad hoc networks is proposed. It considers the problem of alleviat- ing the interference collision risk to primary user communication, explic- itly to protect primary receivers that are not detected during spectrum sensing. The proposed protocol achieves a higher measure of safeguard- ing. A practical scenario is considered where no global network topology is known and no common control channel is assumed to exist. Second, a novel CR broadcast protocol (CRBP) to reliably disseminate the broadcast messages to all or most of the possible CR nodes in the network is proposed. The CRBP formulates the broadcast problem as a bipartite-graph problem. Thus, CRBP achieves a signi cant successful delivery ratio by connecting di erent local topologies, which is a unique feature in CR ad hoc networks. Finally, a new defence strategy to defend against spectrum sensing data falsi cation attacks in CR networks is proposed. In order to identify malicious users, the proposed scheme performs multiple veri cations of sensory data with the assistance of trusted nodes.