Transmission Adaptation for Battery-Free Relaying

Abstract

Energy harvesting (EH)–enabled relaying has attracted considerable attention as an effective way to prolong the operation time of energy-constrained networks and extend coverage beside desired survivability and rate of transmission. In related literature, the Harvest-Store-Use (HSU) model is usually utilized to describe the energy flow behavior of the EH system. However, the half-duplex (HD) constraint of HSU that harvested energy can only be used after being temporally stored in energy buffer may reduce effective transmission time. Thus, we first construct the full-duplex (FD) energy flow behavior model of the EH system where the harvested energy can be tuned to power load and being stored simultaneously. The FD model is then proved to be equivalent with the HSU model when time interval is small enough. Considering some key physical variabilities, for example, the wireless channel and the amount of harvested energy, the transmission adaptation problem for multiple relays embedded with FD EH systems is formulated with the objective to improve the utilization of the harvested energy. We tackle the problem by using a centralized optimization algorithm by jointly tuning the factors, including power control for source and relay nodes, relay selection and dynamic switching among four relay transmission mode, namely HD amplify-and-forward (AF), HD decode-and-forward (DF), FD AF, and FD DF. The centralized optimization algorithm is proposed on the basis of dual decomposition and serves as a benchmark. To enable relays to individually make their own decisions, a distributed algorithm with relatively higher complexity is given by using consensus optimization in conjunction with the alternating direction method of multipliers, and a sub-optimal algorithm with low complexity is provided. The proposed algorithms are shown to have good performance via simulations for a range of different EH rates and prediction errors.