Discrete practical RF energy harvesting models for wireless communication networks

Abstract

Optimal resource management in wireless communication networks is usually modeled in discrete state space, for example, using Markov decision process. This requires the energy harvesting (EH) model to be discrete. Since the existing discrete EH models in the literature use simple discrete distributions, for example, Bernoulli and Poisson, with no regard to the properties of the EH circuits, the proposed work handles this dilemma by deriving discrete EH models based on the practical continuous radio frequency to direct current (RF‐to‐DC) power conversion models. General closed‐form expressions are derived for the probability density functions of the harvested energy in the case of linear, piecewise linear, and non‐linear EH models. These general closed‐form expressions are applied to some special fading models, for example, the Nakagami‐m$$ m $$ fading model. Simulation results verify the derived models. Moreover, statistical analysis is performed to compare the derived models to other discrete models in the literature. Finally, the derived models are applied to a case study where the effect of the number of discrete EH levels and the importance of the derived models compared to other models in the literature are presented. These comparisons show the deviation in the performance of the discrete models commonly used in the literature from the actual performance when the proposed piecewise linear EH model is used. Based on these comparisons, the proposed models, especially the piecewise linear EH model with an adequate number of segments, are recommended to be used for accurate discrete modeling of the EH process.