Budget-Constrained Digital Twin Synchronization and Its Application on Fidelity-Aware Queries in Edge Computing

With the advance of mobile edge computing (MEC) and the Internet of Things (IoT), digital twin (DT) has become an emerging technology for provisioning IoT services between the real world and the cyber world. In this paper, we consider the state updating of DTs in an MEC network through synchronizing DTs with their physical objects. We make use of an energy-constrained UAV for data collection in a sensor network, as an illustrative example for the DT state updating of each object (sensor), and then use the DT data of objects (sensors) later for fidelity-aware query services. To this end, we first formulate a novel DT state staleness minimization, under a given update budget per update round. We then propose an optimal algorithm for a special case of the problem where the budget per update round is exactly $K$ objects synchronizing with their DTs. We then devise an algorithm for the DT state staleness minimization problem by reducing to the award collection maximization problem, assuming that the volume of the update data generated by each object per update round is given. Otherwise, we adopt a deep learning method to predict the volume of the update data. To demonstrate the importance of the DT state staleness in practical applications, we consider fidelity-aware query services in the MEC network, and we develop a cost-effective evaluation plan for each query. We finally evaluate the performance of the proposed algorithms through simulations. Simulation results demonstrate that the proposed algorithms are promising.

I. Introduction

In the past decade, digital twin (DT) has gained significant impetus as a breakthrough technological development that has the potential to transform the landscape of manufacturing today and tomorrow [5]. The DT of a physical object refers to a virtual representation of the object in a virtual environment, allowing for real-time monitoring and analysis of its behavior by synchronizing the DT with the object. The DT market is forecast to reach $15.66 billion by 2023 at an annual growth rate of 37.87% according to a market research in 2017 [4], and more organizations and institutions have gained a lot of interest in DTs. The wide adoption of DTs will further be fuelled up by strong service demands from users in the edge of networks, such as Internet of Things (IoT), autonomous driving, healthcare, smart city, AR/VR, and so on. However, the DT study and its applications is still in the early stage, accelerating the usage of DT technologies will bring unprecedented benefits in service provisioning, intelligent resource allocation for 6G wireless networks through real-time digital representation of physical objects [6], [9], [28], [32], [36].