Flexible resource allocation in device-to-device communications using stackelberg game theory

In underlaid device-to-device (D2D) cellular networks, severe radio interference can be typical for cellular and D2D users, which causes each D2D user's quality-of-experience (QoE) to degrade significantly. Thus, in this paper, a dynamic Stackelberg game is formulated with a single-leader (base station) and multiple-followers (D2D pairs). The leader reduces interference within the network by charging a price to followers, whereas followers react to this price and compete to find optimal transmit power and resource block allocation. To enhance D2D user QoE, D2D users are categorized into one of three application classes with each class mapped to a different utility function. To address this, we propose a crucial innovation where the several different utility functions, available to followers, are solved using a non-scalarized approach, as a scalarized approach to this multi-criteria optimization problem is generally infeasible in real-time D2D cellular communications. To solve the game, a distributed algorithm is proposed, and it is shown to converge to a sub-game perfect Stackelberg equilibrium across all users. Simulation results show that the proposed approach reduces transmit power effectively and increases throughput, giving a Pareto-efficient outcome, while guaranteeing best social welfare and satisfaction across all D2D users.