Optimal Resource Allocation in Ultra-low Power Fog-computing SWIPT-based Networks

TL;DR

This paper optimizes resource allocation in a fog computing system with an ultra-low power device using SWIPT, balancing energy harvesting, local computing, and offloading to minimize energy costs.

Contribution

It introduces an optimization framework for resource allocation in SWIPT-enabled fog computing with ultra-low power devices, considering energy efficiency and offloading strategies.

Findings

Optimized time slots improve energy efficiency.

Offloading strategies depend on energy harvesting efficiency.

Numerical results demonstrate reduced energy costs.

Abstract

In this paper, we consider a fog computing system consisting of a multi-antenna access point (AP), an ultra-low power (ULP) single antenna device and a fog server. The ULP device is assumed to be capable of both energy harvesting (EH) and information decoding (ID) using a time-switching simultaneous wireless information and power transfer (SWIPT) scheme. The ULP device deploys the harvested energy for ID and either local computing or offloading the computations to the fog server depending on which strategy is most energy efficient. In this scenario, we optimize the time slots devoted to EH, ID and local computation as well as the time slot and power required for the offloading to minimize the energy cost of the ULP device. Numerical results are provided to study the effectiveness of the optimized fog computing system and the relevant challenges.