Theories And Methods Of Device-to-Device Communications In Cellular Networks For Services

To improve the spectrum efficiency, device-to-device (D2D) communications in cel-lular network enable users to directly communicate with each other over licensed or other spectrum with the assistance of the base station. However, most studies on D2D communi-cations did not consider scenarios and service types for D2D communications. To solve the above problem, this thesis thoroughly studies D2D communications based network frame-works and optimal transmission scheduling schemes for the data transmissions for different scenarios and services. The contents of this thesis are listed as follows.A D2D-assisted cooperative communication problem for non real-time data transmis-sions is studied. Specifically, a D2D-assisted cooperation framework is proposed for the downlink non real-time data transmissions in a cell, which can obtain white spaces through the cognitive radio and form full-duplex cooperations via out-of-band D2D communica-tions. Therefore, the proposed framework can break through the performance bottleneck of cooperative communications. Further, an optimal transmission scheduling scheme is given for the proposed framework to maximize the network utility, which contains transmission mode selection, relay selection, sub-channel allocation and power control. We obtain an op-timal scheduling plan via the dual decomposition method, and the network utility, network throughput and fairness can be improved.A D2D-assisted cache sharing problem for real-time video transmissions is studied. Specifically, a D2D-assisted caching framework is proposed for the mobile video multicast in a cell, which can stimulate cache sharing among users through D2D communications, by leveraging two types of social ties, i.e., social trust and social reciprocity. The proposed framework can restore as many as possible incomplete video frames with the consideration of the unique video encoding structure. Further, an optimal transmission scheduling scheme is given for the proposed framework to improve the video quality and user experience, which contains user group formulation and spectrum allocation. We obtain a group formation plan based on the coalitional game in a distributed manner, and a spectrum allocation plan is also given, which can improve the video quality and user experience.A D2D-assisted relay communication problem for smart grid data transmissions is studied. Specifically, a D2D-assisted relaying framework is proposed for the delay sensitive data transmissions in a cell under the smart grid environment, which can utilize heteroge-neous data rate requirements and enhance the weaker access link in relay communications via inband underlay D2D communications. Therefore, the proposed framework can break through the performance bottleneck of relay communications. Further, an optimal trans-mission scheduling scheme is given for the proposed framework to minimize the overall information loss ratio, which contains transmission mode selection, D2D/cellular link pairing, relay selection and power control. The optimal transmission scheduling problem is formulated as a two-stage stochastic programming problem to take into account the stochas-tic delays in wireless local area networks. An optimal D2D/cellular link pairing plan can be obtained in a distributed manner, based on which a distributed real-time scheduling al-gorithm is given to reduce the overall information loss ratio and the peak bandwidth demand.