This paper presents an analysis of the performance of a multi-hop multi-relay network where all nodes are battery-assisted (BA) energy harvesting (EH) devices, harvesting energy from a dedicated power beacon. In each hop, the relay with the largest energy is selected for relaying from the nodes that have successfully decoded the symbols in the previous hop. This selected node then augments the harvested energy with some battery energy to relay information. Note that such BA EH nodes can be readily implemented to extend battery lifetimes. In any hop, signalling stops if the direct link to the destination is successful. Using a practical nonlinear EH model, analytical expressions are presented for outage probability, throughput, average battery energy consumption, and battery energy efficiency (ηBEE). The results demonstrate that, despite the nonlinearity in EH, substantial enhancement in the system throughput and ηBEEis possible due to the presence of the direct link. Furthermore, an optimal combination of EH duration, hop count, and target information rate is demonstrated to maximize the overall ηBEEfor a desired throughput. Monte-Carlo simulations validate the accuracy of the derived expressions.

This work investigates a multihop multirelay network in which all nodes are of the battery-assisted (BA) energy harvesting (EH) type, and harvest energy from a power beacon. In each hop, the node with the most harvested energy is selected for relaying and augments the harvested energy with some battery energy. Note that such BA EH nodes are immediately realizable to prolong battery lifetimes. Considering nonlinear EH, analytical expressions are derived for outage probability, throughput, average battery energy consumption, and battery energy efficiency (BEE). It is first shown how the parameters can be chosen to maximize the network throughput. Since unsuccessful relaying that results in outage causes battery energy to be consumed, optimizing the network to minimize the average battery energy consumption or to maximize the node-level BEE is well motivated. We show that the judicious selection of the EH duration can minimize average battery energy consumption while achieving the target throughput requirement. Moreover, a joint optimal choice of EH duration, hop count, and target information rate is shown to maximize overall BEE. Monte Carlo simulations validate the accuracy of the derived expressions.

Due to explosion in the number of Internet of Things (IoTs), and the need to ensure that they are of small cost and form factor, it is expected that most of these will be self-sustaining in the near future, deriving energy by energy harvesting. In this paper we analyze the performance of a multihop network consisting solely of energy harvesting IoTs. Clusters of IoTs in the vicinity assist in communication to the destination. In any hop, if the direct link to the destination is successful, the signalling is halted. The relay with the largest energy is picked in each hop from those that successfully decode the symbols in the previous hop. Using a practical nonlinear EH model, expressions are derived for the throughput of the scheme. It is shown that despite the nonlinearity in EH, the presence of the direct link significantly improves the throughput. Monte-Carlo simulations validate the accuracy of the derived expressions.

This paper investigates the delay performance of a battery-assisted (BA) multi-hop cluster-based network assuming that each hop consists of multiple relays. At each hop, all the nodes harvest energy from the power beacon utilizing the time-switching (TS) energy harvesting (EH) protocol, and augment it with limited energy from a battery in a battery-assisted EH framework. Note that such BA EH nodes are immediately realizable and can prolong battery lifetimes. For efficient relay's battery utilization, the relay with successful decoding status and the largest amount of the harvested energy is selected for forwarding the information in each hop. The symbols are retransmitted after energy harvesting if an acknowledgment is not received from at least one node in the next hop. Considering saturation-based nonlinear EH, an expression is derived for the average delay performance. We then demonstrate that the average delay can be minimized by optimal selection of the TS parameter. Our results also demonstrate that augmenting the harvested energy with a small amount of battery energy dramatically improves the average delay performance. In view of the explosion in number of machine-type devices (MTDs) in recent times, such analysis is of immense practical importance. Monte Carlo simulations validate accuracy of the derived expressions.