Digital Twin (DT) is an emerging technology surrounded by many promises, and potentials to reshape the future of industries and society overall. A DT is a system-of-systems which goes far beyond the traditional computer-based simulations and analysis. It is a replication of all the elements, processes, dynamics, and firmware of a physical system into a digital counterpart. The two systems (physical and digital) exist side by side, sharing all the inputs and operations using real-time data communications and information transfer. With the incorporation of Internet of Things (IoT), Artificial Intelligence (AI), 3D models, next generation mobile communications (5G/6G), Augmented Reality (AR), Virtual Reality (VR), distributed computing, Transfer Learning (TL), and electronic sensors, the digital/virtual counterpart of the real-world system is able to provide seamless monitoring, analysis, evaluation and predictions. The DT offers a platform for the testing and analysing of complex systems, which would be impossible in traditional simulations and modular evaluations. However, the development of this technology faces many challenges including the complexities in effective communication and data accumulation, data unavailability to train Machine Learning (ML) models, lack of processing power to support high fidelity twins, the high need for interdisciplinary collaboration, and the absence of standardized development methodologies and validation measures. Being in the early stages of development, DTs lack sufficient documentation. In this context, this survey paper aims to cover the important aspects in realization of the technology. The key enabling technologies, challenges and prospects of DTs are highlighted. The paper provides a deep insight into the technology, lists design goals and objectives, highlights design challenges and limitations across industries, discusses research and commercial developments, provides its applications and use cases, offers case studies in industry, infrastructure and healthcare, lists main service providers and stakeholders, and covers developments to date, as well as viable research dimensions for future developments in DTs.

Digital twin (DT) technology has recently gathered pace in the engineering communities as it allows for the convergence of the real structure and its digital counterpart throughout their entire life-cycle. With the rapid development of supporting technologies, including machine learning (ML), 5G/6G, cloud computing, and Internet of Things, DT has been moving progressively from concept to practice. In this article, a DT framework based on cloud computing and deep learning (DL) for structural health monitoring is proposed to efficiently perform real-time monitoring and proactive maintenance. The framework consists of structural components, device measurements, and digital models formed by combining different submodels, including mathematical, finite element, and ML ones. The data interaction among physical structure, digital model, and human interventions are enhanced by using cloud computing infrastructure and a user-friendly web application. The feasibility of the proposed framework is demonstrated via case studies of damage detection of model bridge and real bridge structures using DL algorithms, with high accuracy of 92%.

Smart structural health monitoring (SHM) for large-scale infrastructure is an intriguing subject for engineering communities thanks to its significant advantages such as timely damage detection, optimal maintenance strategy, and reduced resource requirement. Yet, it is a challenging topic as it requires handling a large amount of collected sensors data continuously, which is inevitably contaminated by random noises. Therefore, this study developed a practical end-to-end framework that makes use of physical features embedded in raw data and an elaborated hybrid deep learning model, namely 1-DCNN-LSTM, featuring two algorithms--convolutional neural network (CNN) and long-short term memory (LSTM). In order to extract relevant features from sensory data, the method combines various signal processing techniques such as the autoregressive model, discrete wavelet transform, and empirical mode decomposition. The hybrid deep learning 1-DCNN-LSTM is designed based on the CNN's capacity of capturing local information and the LSTM network's prominent ability to learn long-term dependencies. Through three case studies involving both experimental and synthetic data sets, it is demonstrated that the proposed approach achieves highly accurate damage detection, as accurate as the powerful 2-D CNN, but with a lower time and memory complexity, making it suitable for real-time SHM.

This article studies secure communication in an Internet of Things (IoT) network, where the confidential signal is sent by an active refracting reconfigurable intelligent surface (RIS)-based transmitter, and a passive reflective RIS is utilized to improve the secrecy performance of users in the presence of multiple eavesdroppers. Specifically, we aim to maximize the weighted sum secrecy rate by jointly designing the power allocation, transmit beamforming (BF) of the refracting RIS, and the phase shifts of the reflective RIS. To solve the nonconvex optimization problem, we propose a linearization method to approximate the objective function into a linear form. Then, an alternating optimization (AO) scheme is proposed to jointly optimize the power allocation factors, BF vector, and phase shifts, where the first one is found using the Lagrange dual method, while the latter two are obtained by utilizing the penalty dual decomposition method. Moreover, considering the demands of green and secure communications, by applying Dinkelbach’s method, we extend our proposed scheme to solving a secrecy energy maximization problem. Finally, simulation results demonstrate the effectiveness of the proposed design.

Device-to-device (D2D) communications as an underlay to cellular networks can not only increase the system capacity and energy efficiency, but also enable national security and public safety services. A key requirement for these services is to provide alternative access to cellular networks when they are partially or fully damaged due to a natural disaster event. In this paper, we employ energy harvesting (EH) at the relay (R) with simultaneous wireless information and power transfer (SWIPT) to prolong the lifetime of energy constrained network. In particular, we consider a user equipment relay (UER) that harvests energy from radio frequency signal via base station (BS) and use harvested energy for D2D communications. We integrate clustering technique with D2D communications into cellular networks such that communication services can be maintained when the cellular infrastructure becomes partially dysfunctional. Simulation results show that our proposed EH based D2D clustering model performs efficiently in terms of coverage, energy efficiency and cluster formation to extend the communication area. Moreover, a novel concept of power transfer in D2D clustering with user equipment relay and cluster head (UERCH) is proposed to provide a new framework to handle critical and emergency situations. The proposed approach is shown to provide significant energy saving for both mobile users and clustering heads to survive in emergency and disaster situations.