Energy-Efficient D2D Overlaying Communications with Spectrum-Power Trading.
Qingqing Wu,Geoffrey Ye Li,Wen Chen,Derrick Wing Kwan Ng
DOI: https://doi.org/10.1109/twc.2017.2698032
IF: 10.4
2017-01-01
IEEE Transactions on Wireless Communications
Abstract:In this paper, we investigate device-to-device (D2D) overlaying communications with spectrum-power trading, where D2D users (DUs) consume transmit power to relay cell-edge cellular users (CUs) for uplink transmission in exchange for bandwidth from CUs for D2D communications. The proposed spectrum-power trading aims at exploiting individual disparities from both the spectrum and the power perspectives. Recently, energy efficiency (EE) defined by the ratio of the date rate to the power consumption has become increasingly important for devices due to their limited capacity batteries. As such, our goal is to maximize the weighted sum EE (WSEE) of DUs via a joint D2D relay selection, bandwidth allocation, and power allocation while guaranteeing the quality of service of each CU. Specifically, we study WSEE maximization problems for two different cases, i.e., public-interest DUs and self-interest DUs, depending on whether the DUs are willing to share their obtained bandwidth with each other or not. For the case of public-interest DUs, we show that for a given D2D relay selection, the objective function of the WSEE maximization problem in a fractional form can be transformed into a subtractive form that is more tractable based on the fractional programming theory. To perform D2D relay selection, we first reveal a fundamental relationship between the WSEE and two other EE metrics, i.e., system-centric EE and fairness-centric EE, which, to the best of our knowledge, has never been found in the existing works. Based on this insight, the D2D relay selection problem can be cast as a minimum weighted bipartite matching problem. For the case of self-interest DUs, we show that the corresponding problem can also be solved with optimality by the algorithm proposed for the previous case. Simulation results demonstrate the effectiveness of the proposed algorithm.