Nicole Sawyer

A Nash stable cross-layer coalition formation game for device-to-device communications

Device-to-Device (D2D) communications is a key aspect of future 5G cellular networks, as it can help improve spectral efficiency, system capacity, resource utilization, and quality-of-service (QoS), for the overall network. In this paper, a dynamic cross-layer leave-and-join based coalition formation game with non-transferable utility (NTU) is proposed. Our proposed game jointly optimizes mode selection (at the network layer), and resource allocation from the cell and cellular user/s, including power control (at the physical layer), in a distributed wireless network with D2D communications. Moreover, we aim to maximize channel rate and minimize transmit power for both cellular users and D2D pairs, while selecting optimal transmission mode for all D2D pairs. The coalition partition of the proposed game converges to a Nash stable outcome, which is socially efficient. The simulation results illustrate that once a Nash stable coalition partition is achieved, then optimal transmit power and channel rate are achieved for all D2D pairs and cellular users.

Pareto-efficient cross-layer repeated game for Device-to-Device (D2D) communications

Device-to-Device (D2D) communications underlaying cellular communications can increase resource utilization, system capacity, and interference, while increasing quality of service (QoS) for cellular and D2D users. In this paper, we design a non-cooperative cross-layer repeated game, which jointly optimizes mode selection (i.e. either operating in cellular mode or reuse mode) at the network layer and power control at the physical (PHY) layer for a distributed wireless network. The proposed game aims to minimize transmit power, and select optimal transmission mode, while achieving acceptable packet delivery ratio (PDR). The effects on PDR with respect to communication reliability and optimum mode switching point over time are determined. The game is constituted of a non-cooperative power control game and a two-armed bandit game, and by combining these two games we obtain rapid convergence to a Pareto-optimal outcome for both mode selection and transmit power.