Communication, sensing, computing and energy harvesting in smart cities

Abstract

A smart city provides diverse services based on real-time data obtained from different devices deployed in urban areas. These devices are largely battery-powered and widely placed. Therefore, providing continuous energy to these devices and ensuring their efficient sensing and communications are critical for the wide deployment of smart cities. To achieve frequent and effective data exchange, advanced enabling information and communication technology (ICT) infrastructure is in urgent demand. An ideal network in future smart cities should be capable of sensing the physical environment and intelligently mapping the digital world. Therefore, in this paper, we propose design guidelines on how to integrate communications with sensing, computing and/or energy harvesting in the context of smart cities, aiming to offer research insights on developing integrated communications, sensing, computing and energy harvesting (ICSCE) for promoting the development ICT infrastructure in smart cities. To put these four pillars of smart cities together and to take advantage of ever-increasing artificial intelligence (AI) technologies, the authors propose a promising AI-enabled ICSCE architecture by leveraging the digital twin network. The proposed architecture models the physical deep neural network-aided ICSCE system in a virtual space, where offline training is performed by using the collected real-time data from the environment and physical devices.