Game-Theoretic Resource Allocation Methods For Multicell Device-to-Device Communication In Cellular Network

With the rapid development of wireless communication technology, mobile communication has been an indispensable part of people's daily life. It is predicted by ITU that, the mobile devices in 2020 will play a more important role in the future. Device-to-Device(D2D) communication, which is a key component of the future communication technology, has attracted a great interest. With D2 D communication, the system throughput and resource utilization can be improved dramaticlly in a cellular system. However the employment of D2 D technology brings new challenges to the system. If a D2 D pair reuses the current in-use spectrum with cellular users, it will generate strong interference to D2 D communication links and cellular communication links. If there are multiple D2 D pairs reuse the same spectrum, they will interfere each other. Resource allocation is the key step to address the above issues. Therefore, how to design a reasonable and effective resource allocation mechanism becomes the research focus of the communities.The main contributions in this thesis are summarized as follows:(1) This thesis presents three intercell D2 D communication scenarios with cellular infrastructures. For each dedicated scenario, the sum rates of base stations are defined and the resource allocation problem is formulated. To solve the problem develop a static game model that D2 D communication users and cellular users share the same licensed spectrum. With the game model, we analyze its Nash Equilibrium properties. Meanwhile we then discuss a new resource allocation method of Nash Equilibrium not exists and propose a repeat version on the basis. Otherwise we propose a resource allocation protocol according to the interactions among BSs. And we examine the sensitivity factors of the model and system performance through extensive numerical experiments. The results through extensive numerical experiment verify the resource allocation model developed in this thesis considerably improves the system performance in term of the sum rates and the sum rate gains, which further provides a systematic insight on resource allocation for intercell scenarios.(2) This thesis then consider the scenario that multiple D2 D pairs reuse the common uplink resource of two cells, in which there are several cellular communication links. In such a scenario, we define the sum rates the BSs and D2 D links, and formulate the resource allocation problems. And for the problems, we develop a static game model to address the problem that multiple D2 D communication links and cellular communication links share the common resource. Based on the static game model, we propose a repeated game model and then analysis the stability of the model. Meanwhile we examine the impacting factors of the model and system performance through extensive numerical experiments. Theoretical results show that the game model can reach stable by choosing reasonable adjustment parameter. Numerical experiments further verify the developed validness of model is reasonable.In conclusion, this thesis investigates the resource allocation problem under the condition of complete and incomplete assumption for intercell D2 D communications underlaying cellular networks. It establishes the corresponding game models and designs algorithms for them. The works presented in the thesis systematically advance the intercell D2 D communication research and set a representative example for developing game theory.