Distributed Link Scheduling for Device-to-Device Communications via Power Control

In this work, we consider the problem of distributed link scheduling for D2D communications without a centralized controller. Our design goal is to maximize a chosen global objective function. For example, maximizing the long-term proportional fair throughput amongst distributed links can be such a goal. In principle, each link evaluates whether its rate increase by raising its transmit power is sufficient enough to compensate others' rate decrease due to increased interference power towards the objective function maximization. To this end, an iterative distributed power control algorithm, called PowerD2D, and supporting frame structure and signaling method are proposed. As a result of link scheduling via PowerD2D, some links are completely turned off by setting their transmit power to zero and some scheduled links are assigned non-negative transmit power, which are all set for optimality. In addition to its superior performance, PowerD2D algorithm is also flexible enough to enforce any level of fairness amongst links in a distributed manner.