A Study On Secondary User Monitoring Strategies In Cognitive Radio Networks

Cognitive radio networking (CRN) is a promising technology to improve the spectrumutilization by allowing secondary users (unlicensed users) to opportunistically access whitespace (spectrum holes) in licensed bands of primary user. Monitoring the detailed charac-teristics of an operational cognitive radio network is critical to many system administrativetasks as well as security issues. However, very limited work has been done in this area.Firstly，this thesis studies the passive user-centric secondary user monitoring problemin a slotted cognitive radio network, where the users’ traffic statistics and behavior character-istics are all unknown in priori. The problem is formulated as a multi-armed bandit (MAB)problem with weighted virtual rewards. The thesis proposes several dynamic sniffer-channelassignment policies to enable the monitoring system to capture as much interested secondaryuser data as possible with limited monitoring resource. Afterwards, extensive simulationsare conducted to analyse and compare the proposed policies in different scenarios towardsthe change. The results verify the feasibility of the proposed strategies and determine whichpolicy should be adopted in the unslotted system.Then the same problem is investigated in an unslotted system. The UCB1policyis adopted to solve the sniffer-channel assignment problem in an unslotted system withunknown users’ traffic statistics and behavior characteristics. In order to use limitedmonitoring resource effectively, primary-user-avoid method and multiple-sniffer methodare proposed and applied to solve the problem. Simulation results show that our proposedpolicies integrated the proposed methods can achieve logarithmic regrets with relativescalability and converge in finite time in an unslotted systems.