Research On Spectrum Allocation Algorithms Based On Game Theory In Cognitive Radio Networks

With the rapid development of wireless communication technologies, and the increasing of wireless subscriber and new radio business, the radio frequency resources had tended saturated, which do not seem to satisfy the new demand of radio technologies. Actually, the long-time measurements show very little usage of the allocated frequency bands. Recent research has shown that the scarcity of radio resources is caused by unreasonable assignment. Cognitive radio technology can make the best use of radio spectrum through intelligent spectrum management. With the ability of sensing and reusing the idle spectrum resources, it can effectively preclude the possibility of conflicting. According to the graph and game theories, this dissertation mainly analyzes the spectrum sharing and allocation problems through optimal method and sub-gradient algorithm in cognitive radios.There are two approaches for the spectrum access, namely, spectrum orthogonal and spectrum sharing. In orthogonal access, each channel or subcarrier only permits one user to access at some time; but in sharing access, the users can simultaneously access to all usable channels. In cognitive radios, according to the usable spectrum from primary (i.e. licensed) users, there are another two spectrum access kinds, namely, spectrum overlay and spectrum underlay. The spectrum overlay increases the spectrum efficiency by granting cognitive (i.e. unlicensed) users to opportunistically exploit unused frequency bands of primary users. In contrast, the spectrum underlay permits simultaneous sharing of all the frequency bands, at this moment, the cognitive users is limited by the interference constraint, but the spectrum efficiency is much higher.This dissertation studies and analyzes the spectrum allocation problems under orthogonal access and sharing access for cognitive radios in detail. After a full account of the features about spectrum overlay and spectrum underlay, the dissertation mainly centered on the allocation algorithm designing, the effectiveness of the algorithms and the performance analysis. In addition, the article also deeply discussed the spectrum pricing game between primary users as well as primary and cognitive users, the pricing games under competitive equilibrium and Nash bargaining is the focus research target.Firstly, according to the characteristics of spectrum allocation in cognitive radio, the dissertation explains and analyzes the present spectrum allocation algorithms and mechanisms, including water-filling theorem, spectrum allocation mechanisms based on graph coloring and game theory. The graph based channel allocation models are introduced, such as the distributed list-coloring algorithm, color-sensitive graph coloring algorithm and parallel algorithm, the flows and characteristics of these algorithms are particularly analyzed and compared. Then according to the game theory, the channel allocation mechanism based on potential game between cognitive users, and the spectrum pricing mechanism based on repeated game and auction mechanism between the primary users are introduced.Secondly, combining the interference graph model with maximal independent set, a improved algorithm of the parallel algorithm and an algorithm based on maximum weighted independent set is proposed, while reducing time overhead efficiently, it improves the user fairness effectively. These algorithms are based on"0/1"model, which cannot effectively solve the channel allocation problems in spectrum underlay, so we consider introducing game theory. According to the characteristics about spectrum overlay and spectrum underlay, the corresponding water-filling algorithms are proposed to solve the spectrum allocation problems between multi-users. Theory and simulation analysis show that the spectrum allocation algorithm based on subcarrier partition can not only complete the subcarrier and power allocation for cognitive users, but also effectively protect the primary users.Moreover, game theory is used to analyze the spectrum allocation problem under sharing access in cognitive radios. For spectrum overlay, the power iterative algorithm is proposed. Simulation analyses verify the convergence and effectiveness of this algorithm. For the spectrum underlay, due to the interference constraint, the users'strategy spaces are coupled. By introducing duality principle and sub-gradient method, the coupled strategy game can be decomposed into a noncooperative game with no coupled constraints and an optimization problem. And then the spectrum sharing algorithm for spectrum underlay is proposed.Finally, the pricing games between primary users as well as primary and cognitive users are researched by equilibrium and bargaining theories. By using the utility function and Bertrand equilibrium, a dynamic Bertrand algorithm is proposed for channel pricing of primary users, theory and simulation analysis show that the stability of Nash equilibrium related to the speed of adjustment. Due to the equilibrium solution is not the optimal one, a distributed bargaining algorithm based on Nash product is proposed, through which we can not only obtain the near optimal utility, but also consider the fairness. Taking advantages of NBS into account, three asymmetry Nash bargaining models between the primary and cognitive users are proposed to analyze the spectrum price, spectrum demand and system utility finally.