Vehicle as a Computational Resource: Optimizing Quality of Experience for connected vehicles in a smart city

Today, communication technologies enable smart cities to become more effective in terms of satisfactory provision of services. Connected and Autonomous Vehicles (CAVs) can be considered key elements for today's Intelligent Transportation Systems. A small number of research investigations is found to aim at exploiting the resources of CAVs for services provisioning within a smart city. Moreover, and to the best of our knowledge, no research investigations have yet investigated exploiting the computational capabilities of CAVs to become an added value to the computational power of smart cities. In this paper, we introduce the concept of Vehicle as a Computational Resource (VaCR) and develop a system that enables vehicles to share their computational resources within smart cities. Users rely on Service Provider (SP) to acquire the required services that meet their Quality of Experience (QoE) preferences. The device capabilities of VaCR are one of the main indicators in QoE preferences. To that end, evaluation models are developed based on aggregate statistics and Machine Learning (ML) techniques for the discovery and selection of the appropriate VaCRs to participate in the provisioning of services. The deployment is modeled using a multiagent system. Then, game theory is used to model the recruitment that aims at optimizing QoE for Connected Vehicles (CVs). Thorough simulations, analyses, and evaluations of the proposed VaCR system are carried out. The performed validations confirm the effectiveness of the proposed system in optimizing QoE and the successful VaCR system operation, while attaining appealing performance characteristics. (C) 2021 Elsevier Inc. All rights reserved.