Game Theory Based Network Selection And Dynamic Resource Allocation Strategy

The development trend of wireless network is that many kinds of coexisting radio access technologies service for users. These radio access technologies have their own advantages in coverage, system capacity, mobility, and so on, which are used complementarily and fuse into heterogeneous network so as to make communications with quality wherever and whenever. As one of the key technologies of heterogeneous network convergence, the network selection is to be an important research hotspot, namely, how the user terminal selects the optimal radio access technology according to the certain selection strategy among many coexisting access technologies. Meanwhile, with the underutilization of spectrum resource and increasing broadband communication demands, the spectrum becomes more and more shortage. The cognitive radio (CR) technology, which can allow the primary users (PUs) and secondary users (SUs) to share the same spectrum, and improve the utilization of the limited spectral resources, according to detection technologies, is proposed. In cognitive radio network, especially with the introduction of cooperative relay, the problem of efficient resource allocation becomes a key problem need to be solved,The key work in this thesis is included as following.First, a novel multiplicative multi-attribute auction based optimal network selection algorithm is proposed to maximize the user’s performance-cost-ration and achieve networks load balancing. The main process of algorithm is divided into three phases; firstly the set of candidate networks is determined under the consideration of the minimum bandwidth requirement. Secondary, the network’s utility can be gained by entering the observed network parameters into the defined utility function, similarly, the network’s cost also can be gained by entering the network load intensity and cost factor parameter into the defined utility function. Finally, the best suitable network with the maximum user’s performance-cost-ration is selected. Simulation results confirm that proposed scheme outperforms the existing scheme in terms of network selection’s fairness, user’s performance-cost-ration, load balancing and the number of accommodated users.Second, coalition graph theory for the joint relay selection and resource allocation in cooperative cognitive radio network is proposed. The objective of this optimization problem is to maximize both system throughput and user’s fairness under the quality of services (QoS) requirement of the primary system. The main process is divided into three phases:potential coalition heads selection which is determined by maximum weight bipartite matching algorithm; adaptive coalition graph formation which enables the nodes with more surplus resource to help the nodes needing more help; and power is allocated using iterative water-filling. Simulation results show that, the proposed algorithm outperforms other algorithms in terms of both throughput and fairness while adapting the topology to environment changes due to heterogeneous traffic and mobility.