Research On Cooperative Game Based Resource Allocation Algorithms In Energy Harvesting Heterogeneous Cellular Networks

With the rapid development of telecommunications constantly,mobile services and traffic are increased as well as service is diversified,which leads to the shortage of system capacity and high energy consumption in current wireless networks.In coming 5G,ultra dense network(UDN)of small cells is proposed to efficiently improve the capacity of wireless network by sharing spectrum.Energy harvesting is a technology of collecting energy of surrounding environment and turning it into electric energy,which is used for power of devices.For the concept of green communication being so popular lately,energy harvesting must be seriously considered.However,UDN of small cells may also introduce severe interference among small cells and macrocells,and the harvested energy may be limited.Therefore,it is important to reasonably allocate the wireless resource and power to mitigate the interference and increase the capacity.Hence,the study of resource allocation algorithms in energy harvesting UDN has been a heated point.Therefore,this thesis studies the cooperative game based resource allocation algorithms in the energy harvesting UDN.A coalitional game based algorithm is introduced in small cells for optimizing user access problem,in order to minimize the interference among small cells and allocate transmission time and power reasonably.Considering NOMA in small cells,this thesis proposes a matching game to solve the user access problem and derive the algorithm of many-to-one matching game with externality.In small cell networks with caching,a stackelberg game jointly optimizes user access problem and backhaul bandwidth allocation is studies.This thesis proposes a stackelberg game to tackle the user access problem and caching content update for caching considered UDN.The main contributions of this thesis are presented.(1)For downlink Energy-harvesting small cell networks,a coalitional game-based optimal resource allocation algorithm is proposed.First,the energy-harvesting small cell network model is established,in which small cells are connected by transmit lines and hold the ability to transfer energy.Then,a decentralized coalition formation algorithm is presented with spectrum efficiency as u.The strategies of sharing energy and time are also given.For ensuring the convergence of the algorithm,the small cell transition rules are put forward to control the coalition formation,which would also guarantee the optimization of the game.The simulation results show that the proposed algorithm could enhance the system frequency efficiency by sharing harvesting energy and thesatisfaction of service is also demonstrated to be improved.(2)This thesis proposes a rate-based two-sided many-to-one matching game algorithm for energy-harvesting small cells with NOMA.First,acting as players in this game,users and small cells establish their preference lists in terms of rate.Then,each user sends application to the first small cell in lists and small cells choose to reject or accept users according to their lists.Considering externality in matching game,we propose an algorithm which involves swap-matchings to find the optimal matching.After finite iterations,this game will converge to a stable matching result.Simulation results show that this algorithm outperforms one-to-one matching game in efficiency and rate.(3)A stackelberg game jointly optimizes user access and backhaul bandwidth allocation of the energy-harvesting small cell network in heterogeneous cellular networks is proposed.Considering this scenario of caching files in small cells,users accessing the small cells download files they need from small cells first,and then request the rest file from backhaul.The user accessing problem is established as a student-project allocation matching game with energy efficiency acting as the utility function.Then based on the user accessing and file request of energy-harvesting small cells,a backhual bandwidth auciton game is proposed to adjust the price of backhaul bandwidth as well as the backhaul bandwidth of each small cells.Simulation results proves that our algorithm outcomes than the other one which merely considers the user access matching game in energy efficiency and value of download files.(4)To improve the users' satisfaction of download files,a caching based resource allocation algorithm in energy-harvesting small cell networks is proposed,which is established as a stackelberg game that jointly optimizes user accessing and small cells' content updating.This thesis introduces a coalitonal game for users to transfer between coalitions with user satisfaction being utility function,and the transition is accepted once the utility of user after transferring satisfies the transition rules.The algorithm allows several users to be served by a same cell,and users accessing the same small cell will share the same channel using NOMA.Then,to maximize the benefit,a cache updating scheme is put forward to update the files cached by small cells.The equilibrium solution can be achieved after finite iterations of the two algorithms.The simulation results show that our novel algorithm performs better in user satisfaction and energy-saving.