Wait for Fresh Data? Digital Twin Empowered IoT Services in Edge Computing

The Mobile Edge Computing (MEC) paradigm gives impetus to the vigorous advancement of Internet of Things (IoT), through provisioning low-latency computing services at network edges. The emerging digital twin technique has grown in the community of IoT, and bridges the gap between physical objects and their digital representations in MEC, thereby enabling real-time data analysis, simulating the dynamics of systems, and optimizing network resource allocation. In this paper, we consider query services for various IoT applications in an MEC network, built upon digital twin data in the network, with the aim to optimize the freshness of query results, measured by the Age of Information (AoI) and query service delays simultaneously. We first formulate a novel minimization problem that explores a nontrivial trade-off between these two critical yet conflicted optimization objectives, and show the NP-hardness of the problem. We then propose an approximation algorithm for the problem with a provable approximation ratio, at the expense of a moderate computing resource violation. We finally evaluate the performance of the proposed algorithm via simulations. Simulation results demonstrate that the proposed algorithm is promising, and outperforms the benchmarks, improving by no less than 18.9% of the performance in comparison with that of the baseline algorithms.

I. Introduction

The last decade has witnessed the unprecedented explosion of Internet of Things (IoT) applications by culminating the proliferation of IoT devices around us connected with the Internet, thereby permeating the modern-day world and flourishing the potential for brilliant living [17]. However, most IoT devices have limited resources and energy for running IoT applications on themselves, instead, they usually offload their computing tasks to remote clouds for processing which causes high service costs and service delays [6]. Moreover, in traditional IoT architectures, IoT devices store data in their backlogs for future diagnosis and improvements, which however may lead to stale feedback, unverified updates, and severe malfunctions [22].