Research On Resource Allocation Techniques In Cognitive Radio Network

As a promising potential technology to improve spectrum utilization, the newly emergingcognitive radio（CR） has attracted increasing attention. Basically, there are two categories of CRoperations: opportunistic spectrum access (OSA) and spectrum sharing (SS). OSA is asensing-based technology, which allows a secondary user (SU) in the secondary communicationnetwork (SCN) to access the spectrum that is originally allocated to the primary users (PUs) whenthe spectrum is not used by any PU. In this way, the spectrum utilization efficiency can be greatlyimproved. However, to precisely detect a vacant spectrum is not an easy task. Alternatively, SSallows simultaneous transmission of PUs and SUs.In this dissertation, we begin by characterizing the key technologies of cognitive radionetworks, investigate the resource allocation and management problem in CR network. Morespecifically, on the premise of guaranteeing the QoS of PUs, we deals with the problem of optimallydeploying SUs’ communication resources such as link space-time scheduling, transmit power andbeamforming weights in order to most efficiently share the licensed spectrum, e.g., achievemaximum signal to noise ratio, information rate and throughput. The main contribution of thisthesis can be concluded as follows:(1) Downlink Space-time Scheduling Algorithm in Cognitive Radio Cellular Networks. Inthis thesis, we consider the downlink space-time scheduling algorithm problem of a typicalcognitive cellular networks. Through spatial multiplexing technology, the multi-antenna cognitivebase station transmit multiple data stream to different SUs which is equipped with a single antenna.A dual orthogonal spatial multiplexing transmission technique is proposed in this paper. Bydistributing the orthogonal space multiplexing vector to SUs, PUs interference generated by SUs isto be zero. And we further address a low complexity user scheduling algorithm based on ant colonyoptimization (ACO). The algorithm can get close to the performance of the most optimal methodonly by increasing a certain amount of calculation.(2) Source Transmit Power Control and Distributed beamforming in Amplify-and-Forward Two-Way Relaying Networks. we focus on an amplify and forward cognitive relaynetwork, in which cognitive relay nodes forward cognitive source node signal to cognitivedestination node through distributed beamforming. We investigate, jointly subject to constraints onindividual transmit power constraints, total transmit power and interference caused to primary users,how to optimize cognitive source tramit power and cognitive relay beamforming weights in order tomaximize the signal-to-noise ratio (SNR) at the cognitive destination. Firstly, we assume that thesource transmit power is given to solve the problem of latter. we decompose the problem into a series of quadratically constrained quadratic programming (QCQP) problems through Dinkelbachtype method. the QCQP problems further are resolve into convex semidefinite programming (SDP)problems which can be easily solved by interior point method. We prove that the above two types ofproblems are equivalent. Secondly,we design an evolutionary computation technique based onparticle swarm optimization to achieve the optimum source transmit power.(3) Distributed beamforming in SU two-way relay networks. We consider a SU two-wayrelay network consists of a pair of source nodes and a bunch of single-antenna SU relay nodes.Subject to PU interference temperature and single user transmit power of SU relay nodes, we studyhow to choose beamforming weights of SU relay nodes in order to maximize the minimum receivedSINR of sources, i.e., SINR balancing problem. By using semidefinite relaxing technique, theoriginal SINR balancing problem is converted into a convex feasibility prolem and we propose abisection search algorithm to solve it efficiently.