Research On Resource Allocation And Cooperative Mechanism In Wireless Networks Based On Game Theory

The design and deployment of a centralized-control based wireless network is a time-consuming and manpower-intensive series of tasks, so that the centralized-control based approach gives way a new approach to network configuration and management that places the decision-making burden on the individual terminals. In this scenario, control is distributed and local, and network scalability is enhanced. However, since the network consists of a community of local agents, design and operational decisions are made without explicit representations of the global environment or even of the other users. Game theory provides a wealth of tools that can be applied to the design and operation of this kind of communication systems. In this article, we use game theory to study and improve the global network configuration and performance when they are determined solely by the decisions of individual agents.In the physical layer, firstly, using cooperative game theory, we consider the problem of resource sharing between two nodes in cooperative relay networks. In the system, each node can act as a source as well as a potential relay, and is willing to achieve an optimal signal-to-noise ratio (SNR) increase by adjusting the power value it should contribute to cooperative relaying. We formulate this problem as a two-person bargaining game, and use the Nash bargaining solution (NBS) to achieve a win-win strategy for both nodes. Simulation results indicate that the NBS resource sharing is fair in that the degree of cooperation of a node only depends on how much contribution its partner can make to its SNR increase.Next, we continue carrying research on the problem of stimulating cooperation and resource allocation in cooperative relay networks. Differing from the former study, we formulate the resource allocation problem as a sellers' market competition where a relay is willing to share its resource among multiple user nodes. We use a Stackelberg game to jointly consider the benefits of the relay and the users. First, the relay determines the price of relaying according to the user demand. Secondly, the users purchase the optimal amount of resource to maximize their utilities. Although the Nash equilibrium (NE), i.e., the solution of the game can be obtained in a centralized manner, we develop a distributed algorithm to search the NE, which is more applicable to practical systems. Also, the convergence conditions of the algorithm are analyzed. Simulation results show the game could stimulate cooperative diversity between the selfish nodes effectively. And, by using the distributed algorithm, the relay and the users could determine what price should ask for and how much bandwidth should buy, respectively.In MAC (media access control) layer, firstly, we propose a game theoretic MAC scheme to improve the performance of IEEE 802.11 DCF in wireless LANs. The channel contention process between the nodes is formulated as a dynamic game with incomplete information. According to the Nash equilibrium of the game, a novel MAC scheme, called G-DCF (Game theoretic DCF), is proposed. Using the G-DCF, each node adjusts its local contention parameters for data transmission to the current game state (i.e. the number of competing nodes), and thereafter updates the game state through the transmission feedbacks. This process is finitely repeated to get the optimal performance. Additionally, to help the idle nodes estimate the game state accurately, a virtual frame scheduling mechanism, called VFS, is developed. The VFS is incorporated into the G-DCF, so an idle node can obtain his equilibrium strategy when he gets ready to transmit real frames. Simulation results show G-DCF can increase the system throughputs, decrease the delay-bound and frame-loss-rate.In the following, a novel MAC scheme referred to as P-EDCA is proposed to resolve the unfairness problem in IEEE 802.11e WLANs (i.e., the flows with lower priorities can't get any throughput of data transfer in cases of heavy network loads). P-EDCA provides weighted fairness for differentiated service between the flows, and supports the stations with the 802.11e based multi-queue structure. Use of the inner centralized queuing discipline, called the S_intra-node, P-EDCA guarantees fair transmission opportunity between the queues within a station. And by using an incomplete cooperative game theoretic access method, called the S_inter-node, fairness of channel accessing between the stations is reached. Simulation results show that the weighted-fair differentiation is accurately implemented by P-EDCA with no information shared between the users. And, without decreasing the performance of higher priority flows, P-EDCA outperforms the original EDCA in terms of its QoS assurance for lower priority flows in the network.As for using game theory in cross-layer design, we propose a generalized proportional fair (GPF) resource (i.e., sub-channel and power) allocation scheme for downlink OFDMA networks. A user's payoff is defined as a function of his data rate. And the resource allocation can be formulated as a cooperative game to maximize the sum of the users' payoffs. To obtain the Nash bargaining solution (NBS) of the game, a suboptimal subcarrier allocation is firstly performed by assuming an equal power allocation. Then an optimal power allocation algorithm is given to maximize the sum of the users' payoffs and meanwhile ensures QoS demand (the minimum data rates) of the users. Compared with the other two resource allocation algorithms, i.e., the maximizing system capacity and the max-min fairness algorithm, the cooperative game achieves a good tradeoff between the fairness and the overall system capacity. Thus the user's QoS demand is ensured, and the system resource is utilized efficiently.Finally, we propose a non-cooperative game to perform the sub-carrier assignment and power allocation for the multi-cell OFDMA systems. The objective is to find a balance between the QoS satisfaction and battery life for applications where energy efficiency is important. We define a game player as a cell formed by the unique base station and the served users. The utility function considered here measures the user's achieved utility per power. Each individual cell's goal is to maximize the total utility of its users. To search the Nash equilibrium (NE) of the game, an iterative and distributed algorithm is presented. Since the NE is inefficient, we introduce pricing of user's transmission power to improve the NE in the Pareto sense. Simulation results show the proposed game outperforms the water-filling algorithm in terms of fairness and energy efficiency. Moreover, through employing a liner pricing function, the energy efficiency could be further improved.