Energy Efficient Software Matching in Vehicular Fog

TL;DR

This paper proposes a vehicular fog computing architecture that optimizes software matching to reduce power consumption, demonstrating up to 27% savings by diversifying software packages across vehicles.

Contribution

It introduces a MILP-based model for software assignment in vehicular fogs, analyzing how software variety impacts power efficiency.

Findings

Power saving of up to 27% with four or more software packages.

Software diversity significantly reduces overall power consumption.

The model effectively guides software distribution for energy efficiency.

Abstract

Along with the development of Internet of Things (IoT) and the rise of fog computing, more new joint technologies have been proposed. Vehicular Ad-hoc Networks (VANET) are one of the emergent technologies that come with a very promising role, where the spare processing capabilities of vehicles can be exploited. In this paper, we propose a fog architecture to provide services for end users based on a cluster of static vehicles in a parking lot referred to as a vehicular fog. The proposed vehicular fog architecture is connected to the central data center through an optical infrastructure. As the processing requests from users require specific software packages that may not be available in all vehicles, we study the software matching problem of task assignments in vehicular fog. The goal of this paper is to examine the effect of software packages variety in vehicles on the assignment decision and the overall power consumption. A mixed integer linear programming (MILP) model was utilized to optimize the power consumption of the overall architecture, considering different numbers of software packages in the vehicular fog. The results reveal a power saving of up to 27% when vehicles are uploaded with four or more different software packages out of a library of ten software packages in this example.