Game-Based Resource Management Research For Heterogeneous Cellular Networks

With the rapid development of the wireless Internet and popularity of new services and mobile applications, mobile users put forward a higher requirement for data rates. Wireless network must constantly improve their communication ability. An effective method is introduced in femtocell base stations(FBS) based on cellular communication system in the range of macrocell base station(MBS) to improve the network coverage of a particular region and form a two-tier heterogeneous femtocell networks.The femtocell base stations and macrocell base station both operate over the same frequency band. Since the frequency spectrum resource of MBS is limited, to maximize the use of limited spectrum resources, a game-based resource allocation strategy is adopted in this paper. First of all, MBS protects itself by pricing the interference from femtocell user equipments(FUEs), and FUEs need to pay for their interference. The MBS can adjust interference prices to maximize its revenue within its tolerable aggregate interference threshold Q, while the FUEs optimize their profits by trading off between transmission rate and payment of buying interference quota. To model this problem, a stackelberg game is formulated between MBS and FUEs. At the same time, considering the conflict of interest among FUEs since they compete for limited frequency spectrum resources, we establish a non-cooperative subgame to solve the problem of resources allocation in the same layer.Two practical femtocell network models are investigated: densely deployed scenario and sparsely deployed scenario. As for pricing schemes, we consider uniform pricing and non-uniform pricing. Uniform pricing scheme means MBS sets a uniform interference price for all the femtocell users, while for the non-uniform pricing scheme, MBS sets different interference prices for different femtocell users. In the sparsely deployed scenario, we can find fixed form of optimum resources allocation. In the densely deployed scenario, FUEs power allocation changes dynamically due to mutual influence between femtocells, and accordingly this paper puts forward an adaptive expectation parallel iterative power update algorithm. And for the interference price selection problem, we present interior penalty function method to acquire the approximate optimal solution of resource allocation by iterating algorithm. Finally we construct a specific network environment, and it is verified by MATLAB simulation that these algorithms are superior. Simulation results show that the algorithms are mature and can greatly improve system performance.