VISE: QoS-Aware Virtual Sensor Orchestration in Green Multi-Sensor-Cloud

Abstract

This work presents the problem of optimal resource orchestration for green Sensor-as-a-Service (Se-aaS) in the presence of multiple sensor-cloud service providers (SCSPs) and vehicle owners (VOs). It leads to an oligopolistic market, where the shareholders may adopt unfair means to maximize revenue. Therefore, we consider the trust factors among the SCSPs and the VOs and ensure high QoS to minimize service delay, maximize service availability, and maximize energy efficiency. We propose a virtual sensor orchestration scheme, VISE, to ensure high QoS in green multi-sensor-cloud systems. VISE is a multi-level expected utility game-theoretic scheme. In VISE, the requested SCSP initially aims to find an optimal set of vehicles to send the sensed data from its own sensor-equipped vehicles to the roadside units (RSUs). We consider that the intermediate nodes may belong to other SCSPs. Further, to address the inadequacy of resources, the requested SCSP in VISE relies on selecting an optimal SCSP from the other SCSPs to serve vehicular Se-aaS. We evaluate the performance of VISE both theoretically and experimentally and compare it with benchmark schemes. Simulation results depict 33.33 – 53.34% and 21.57 – 29.13% increase in end-user satisfaction and network lifetime, respectively, while ensuring a 37.25 – 42.85% increase in revenue sharing among the VOs.