A futuristic green service computing approach for smart city: A fog layered intelligent service management model for smart transport system

With the widespread adoption of technologies like the Internet of Things (IoT), context awareness, and decentralized computing at the edge of the network, service delivery in smart city parlance has been rapidly evolved. Management of Information and Communication Technologies (ICT) infrastructure for such dynamic environment at scale brings about new challenges. Existing fog layer resource management involves context-sharing and migration for real-time vertical and cross-vertical services. However, improper context migration might affect performance negatively. In this paper, the authors have envisioned improving the Quality of Service (QoS) of smart transportation while employing context-aware computing and Artificial Intelligence, which helps alleviate massive data transfers among Fog nodes and intelligent vehicles in real-time. The proposed Context-Aware Intelligent Transportation System (CAITS) manages the services of intelligent vehicles and manages the road traffic of traditional vehicles in an effective manner, with a three-layered learning model that accounts for onvehicle, on-co-vehicle, and on-fog-and-vehicle learning by means of a platoon control algorithm as well as federated learning at the Fog level. Simulations are carried out on CloudSim simulator under different scenarios and the results demonstrate that the proposed scheme improves prediction efficacy of contexts at the Fog layer by approximately 8%-24% than existing models which in turn reflects in reduced service time and energy consumption of EVs and reduces the CO2 of environment.

Automated summary

This paper focuses on improving the Quality of Service (QoS) of smart transportation by employing context-aware computing and Artificial Intelligence. The proposed Context-Aware Intelligent Transportation System (CAITS) effectively manages intelligent vehicles and traditional vehicle traffic, utilizing a three-layer learning model and platoon control algorithm. Simulations demonstrate that the proposed scheme improves context prediction efficacy by approximately 8%-24%, leading to reduced service time and energy consumption of EVs and lower CO2 emissions.