Research On Spectrum Allocation Technology In Cognitive Radio Network

With the rapid development of wireless communication and the continuous expansion ofwireless applications, it is not enough for the existing spectrum resources to fulfill the needsof different businesses. But current spectrum management mechanism is barely satisfying inincreasing the spectrum utilization. As the communication technology in next generation,Cognitive Radio (CR) makes the remission of spectrum resources scarcity and theimprovement in spectrum utilization possible. In cognitive radio system, cognitive users getaccess to the use of spectrum through spectrum sensing technology and they could get a“second opportunity” in the use of the spectrum resources with the help of spectrum sharingsystem. Spectrum allocation technology is one of the core technologies in CR. Effectivespectrum allocation means that spectrum resources can be used in a reasonable and balancedway without any disturbance in primary users so that the spectrum utilization rate can beincreased.This thesis conducts researches on spectrum allocation technology of cognitive radiosystem. It mainly illustrates current researches on allocation technology and the existingmodels of spectrum allocation and proposes corresponding spectrum allocation algorithm onthe basis of graph coloring model.Firstly, it studies different spectrum allocation modes and compares the advantages anddisadvantages of every mode. Then it illustrates the performance index of the algorithm fromthe perspective of fairness, system efficiency and time expenditure aiming at problems theremay occur in allocation. The study also summarizes and analyzes current allocation model.Secondly, the study builds up the model of the allocation algorithm based on the graphcoloring model via mathematical matrix and analyzes and summarizes the classical modelalgorithm. It also studies the fairness among users in the process of spectrum allocation on theabove basis. It introduces a fairness factor in the algorithm through the combination of nodepriority parameter defined in distributed greedy algorithm and max proportional-fair, andbalances the node priority in allocation. The simulations show that the algorithm proposed inthis study takes account of the system efficiency and users fairness.Finally, it builds up a new model on the basis of graph model in accordance to the actualcommunicate needs of cognitive users. It also introduces the evolution algorithm andconducts the analysis on the optimization of spectrum allocation on the basis of above model.This study combines the ergodicity in chaotic initialization, high efficiency in quantumcomputing and the convergence of the artificial bee colony algorithm, and proposes thespectrum allocation algorithm of cognitive radio spectrum on the basis of chaos quantum beecolony (CQBC). It also proves the probability of algorithm convergent is1. At last, it presentsthe algorithm feasibility with the result of spectrum allocation via effective encoding mode.The simulations reveal that the algorithm in this paper has better results in convergence andperformance.