Coordinated Ambient/Dedicated Radio Frequency Protocol Design for Wireless Powered Communication Networks

Abstract

Harvesting energy from radio frequency (RF) sources to wirelessly power electronic devices have been of tremendous interest in both academia and industry in recent years. This thesis tackles the problem of maximizing the sum-throughput while ensuring fairness in a multiple source, multiple sensor environment; where the sensors reliably harvest energy from ambient sources when dedicated sources are unavailable. This thesis proposes a new protocol called the coordinated ambient/dedicated (CA/D) protocol, which has backscatter-enabled combination sensors optimized to harvest energy from intended RF energy sources when available (referred to as dedicated mode) and fall back to harvesting energy from unintended sources when only unintended sources are available (called ambient mode). It is shown that the CA/D protocol, in dedicated mode, delivers up to approximately 296% increase in sum-throughput compared to the reference state-of-the-art time-switching (TS) RF energy harvesting protocol in a dynamic environment. Further, the proposed CA/D protocol increases the Jain’s fairness index from J = 0.76 under the TS protocol to up to 0.90 and 0.80 with and without sensors operating in backscatter mode, respectively. When operating in ambient mode, the proposed CA/D protocol is integrated with two new machine learning techniques, the linear forecaster with near-time linear regression-based enhancer (LFNTLRE) algorithm and an artificial neural network with environment detection (ANN-ED), to determine the optimum EH schedule. These machine learning algorithms can reliably operate in environments where there is unpredictable availability of unintended sources and ongoing changes in channel conditions between the sensors and the unintended sources. Numerical results show that sensors using the ANN-ED and LFNTLRE algorithms can successfully sense up to 99.5% and 99.6% of the data, respectively; very close to the 100% sensing achieved by an ideal sensor. Sensors using the ANN-ED and LFNTLRE algorithms achieve an accuracy rate of up to 99% and 100%, respectively, as well. These findings on the proposed CA/D protocol are significant, necessitating the protocol’s adoption in sensors deployed in the future wireless networks.