Edge Computing Services for Smart Cities: A Review and Case Study

Abstract

Translate Article

Edge computing emerged as a promising paradigm that provides cloud-like services in the vicinity of mobile users. Benefitting from the geographical and social similarities among users, edge computing can intelligently provide on-demand services with low latency. However, the unprecedented amount of tasks along with the heterogeneous nature of applications in future smart cities such as healthcare, energy management, transportation, cybersecurity, and so on, have created a bottleneck on the performance of edge computing. Therefore, edge devices need to be equipped with adaptive machine learning techniques to cope with the continuously changing computing environment and to provide mobile users with reliable smart services. Furthermore, establishing a cloud-to-edge and edge-to-edge cooperative learning can help not only to relieve the burdens on edge servers but also to facilitate the learning process in which the knowledge earned by one computing node can be shared by other nodes in the network. In this paper, edge computing services are presented in light of state-of-the-art machine learning and computing technologies. In addition, a case study on edge-assisted multi-hop D2D communications for IoT devices is proposed aiming to provide seamless connectivity and reduce delay in dense IoT environments.