Research On Resource Optimization Allocation In Cognitive Radio Networks

Cognitive radio (CR), which can greatly improve spectrum utilization, is adynamic spectrum allocation technology, and it is expected to alleviate increasinglyserious spectrum shortage problem that the development of telecommunication industryis now facing. An important goal of CR development is to improve spectrum utilization.Therefore, for a communication network which is equipped with CR technology, theproblem that how to make use of spectrum holes most efficiently is consistent with thegoal of the research of CR technology, and it has the same great significance as CR.There are two primary resources in CR network (CRN): spectrum bandwith andtransmit power. In order to efficiently make use of spectrum holes, and to improvespectrum utilization as much as possible, the two resources must be optimally allocated.In a CR scenario where there is only one CR user, the CR user can occupy all spectrumresource, thus resource optimization problem can be reduced to power optimizationallocation problem. Because spectrum sensing has a huge impact on the accuracy ofspectrum sensing, the sensing time can be regarded as another important resource inCRN, and thus it must also be optimally allocated.In this dissertation, under the guidane of information theory, network theory andconvex theory, resource optimization allocation problem in CRN which is CR'sdifferent application scenario, is explored and investigated.The main results are as follows:1. Resource optimization allocation problem in spectrm sharing based CRN isstudied. Aimed at the problem that existing most resource optimization allocationalgorithm often lead to the interruption of the primary transmission, considering the factthat exploiting primary users' channel state information (CSI) can increase thesecondary thoughput, based on the average and valley signal to noise plus interferenceratio (SINR) constraints, a power optimization allocation algorithm is proposed, whichcan not only ensure the communication quality of the primary transmission, but alsoobtain more secondary throughput; aimed at the problem that theory research forspectum sharing can not be used in practice due to the selfish of primary users, based on seller's market trading mechanism, an actual implementation menthod of spectrumsharing system is proposed.2. Resource optimization allocation problem in orthogonal frequency divisionmultiplexing (OFDM) based CRN is studied. Aimed at the unfair problem that theremay exist in resource allocation in a multiuser CRN, considering the interference causedby OFDM signal energy leakage, a resource allocation algorithm is proposed, which cannot only control the interference to primary users, but also balance the fairness andefficience in resource allocation; aimed at the problem that resource allocation may leadto unbalanced interference to multiple receiving terminal of primary users (or multipleprimary users), a resource optimization allocation algorithm is proposed, which canprotect primary users fairly; aimed at the problem that error detections may lead theinterference to primary users and the waste in resource usage, an energy efficient powerallocation algorithm is proposed.3. Dynamic primary users' traffic model based sensing-based throughput problemis studied. Aimed at the deficiency of traditional, stationary primary users' traffic model,considering a dynamic traffic model closer to the reality, and further considering theimpact of the new model on the performance of spectrum sensing and the interference toprimary users, an optimization allocation problem of resource including spectrum hole,transmit power and sensing time is studied, and an optimal resource allocationalgorithm is also proposed.This dissertation deep studies the resource optimizaiton allocation problem incognitive radio networks, and obtains a number of meaningful research results, whichhave some theoretical value and practical significance for the design andimplementation of future cognitive radio networks.