Simulation And Improvement Of Rendezvous Algorithms In Cognitive Radio Networks

Recently, cognitive radio network(CRN) has emerged as a new and promising networking paradigm to alleviate the shortage and improve the utilization of the scarce spectrum resource. Conceptually, a CRN is a group of cognitive users equipped with cognitive radios. Via the cognitive radios, the cognitive users can identify and then access the idle spectrum bands(channels) which are temporarily not occupied by any licensed user nearby. However, once the licensed users reclaim the spectrum for their data communications, the cognitive users must immediately release the operated channels if their continued using of the channels will cause any harmful interference to the licensed users’ communications. In this sense, channel availability of the cognitive users in a CRN is usually time-varying and location-varying.In CRNs, rendezvous refers to such a process that two cognitive users meet on a channel commonly available to them at the same time and thereby establish a communication link between them. Rendezvous is a fundamental and crucial operation in CRNs, since many network tasks including message exchange and data communication cannot be efficiently carried out without rendezvous. However, achieving rendezvous is usually very challenging for the cognitive users. The main reasons for this challenge are as follows: 1) Two cognitive users are usually unaware of the existence of each other before rendezvousing; 2) They may have non-uniform available channels which would also dynamically change by time; 3) They may start their own rendezvous operations at different instances of time.The rendezvous problem has attracted lots of attentions in the community of CRN research, and a number of rendezvous algorithms have been proposed in the literature. We have also carried out the study on this problem and eventually formed this thesis. Specifically, this thesis contributes mainly through the following three aspects:1) Presenting a moderate but comprehensive survey of the representative and state-of-the-art approaches for the rendezvous problem. Specifically, we focus on channel-hopping algorithms for blind rendezvous and presents a comprehensive review of existing rendezvous algorithms. Also we summarize some reliable metrics to assess the performance of various rendezvous algorithms.2) Providing a thorough comparison study of the representative rendezvous algorithms based on extensive simulations. Based on Matlab software, we propose a simulation method that is applicable to all channel-hopping algorithms by a full account of various network environments and measurement metrics. Then, We conduct a comprehensive simulation to compare the performances of 6 representative rendezvous algorithms.3) Proposing a channel virtual expansion strategy to improve efficiency of existing rendezvous algorithms. This strategy allows for a sound improvement in efficiency of rendezvous algorithms by increasing the ratio of quality channels amid the process of user’s hopping sequence without interfering guarantee rendezvous. Furthermore, we strengthen the security of the algorithms by deploying a random replacement strategy. We prove the effectiveness of the proposed strategy through theoretical analysis and experiment result.