Delay-Aware Digital Twin Synchronization in Mobile Edge Networks With Semantic Communications

Bin Li, Haichen Cai, Lei Liu, Zesong Fei · 2025

Read Delay-Aware Digital Twin Synchronization in Mobile Edge Networks With Semantic Communications by Bin Li, Haichen Cai, Lei Liu, Zesong Fei (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2503-04387.txt.

Introduction

Bin Li and co-authors tackle sync delays in 2025 for digital twins running on mobile edge networks. A digital twin mirrors a real device, and keeping the mirror fresh is the core of the system. Wireless channels are narrow, though, so the team sends meaning instead of raw data, a method called semantic communication, to save bandwidth.

Their goal is to cut the average sync delay across user devices by jointly choosing sync plans, sending power, and computing shares at devices and the base station. Device movement makes the choices uncertain over time. A learning algorithm based on soft actor-critic handles the shifting problem. Simulations cut sync delay by up to about thirteen percent.

Catalog record

CREATORBin Li, Haichen Cai, Lei Liu, Zesong Fei
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE42 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2503-04387

Browse the full library · About the Tower of Babel

Full text

The opening of the work, from the library's hosted text:

Delay-Aware Digital Twin Synchronization in Mobile Edge Networks With Semantic Communications Bin Li, Haichen Cai, Lei Liu, Zesong Fei Abstract The synchronization of digital twins (DT) serves as the cornerstone for effective operation of the DT framework. However, the limitations of channel capacity can greatly affect the data transmission efficiency of wireless communication. Unlike traditional communication methods, semantic communication transmits the intended meanings of physical objects instead of raw data, effectively saving bandwidth resource and reducing DT synchronization latency. Hence, we are committed to integrating semantic communication into the DT synchronization framework within the mobile edge computing system, aiming to enhance the DT synchronization efficiency of user devices (UDs). Our goal is to minimize the average DT synchronization latency of all UDs by jointly optimizing the synchronization strategy, transmission power of UDs, and computational resource allocation for both UDs and base station. The formulated problem involves sequential decision-making across multiple coherent time slots. Furthermore, the mobility of UDs introduces uncertainties into the decision-making process. To solve this challenging optimization problem efficiently, we propose a soft actor-critic-based deep reinforcement learning algorithm to optimize synchronization strategy and resource allocation. Numerical results demonstrate that our proposed algorithm can reduce synchronization latency by up to 13.2% and improve synchronization efficiency compared to other benchmark schemes. I Introduction As an emerging technology in the era of 6G, digital twin (DT) is capable of real-time mapping of real physical entities and environments to virtual spaces, achieving communication, collaboration, and information sharing between the physical and virtual worlds [ 1 ] , making it a promising technology for network resource management [ 2 , 3 ] . Furthermore, DT can support powerful artificial intelligence technologies, applying intelligent learning algorithms in virtual spaces along with real-time data provided by DT can assist physical entities in making more accurate and timely decisions, reducing the resource consumption of user decisions and improving the efficiency of resource utilization [ 4 ] . The DT framework is primarily composed of three components: the physical space, the digital space, and the communication space [ 5 ] . DT synchronization is the cornerstone of this framework, conducting simulation and control operations through the digital replica of the physical world. To guarantee real-time monitoring of physical entities and processes, DT requires the processing and modeling of real-time data from physical entities [ 6 ] , ensuring real-time update and precise synchronization of information. However, the construction of DT models requires not only real-time data but also substantial computational capabilities. User devices (UDs) can sense real-time data, but their computational resources are limited, making the implementation of DT challenging [ 2 ] . Mobile edge computing (MEC), as a promising paradigm, can deploy computational resource near users, ensuring low-latency task execution and DT maintenance, effectively addressing the resource constraints on UDs [ 7 , 8 ] . Consequently, UDs can transmit sensed real-time data to MEC servers for modeling and updating the DT of physical entity, ensuring the communication efficiency and energy-saving of DT services [ 6 ] . However, the bandwidth resource and energy on UDs are limited, as well as transmitting large amounts of data from UDs to the base station (BS) also results in latency and energy consumption. Therefore, how to process and transmit the real-time data sensed by UDs to achieve efficient DT synchronization is a critical issue that urgently needs to be addressed. Recently, semantic communication has been recognized as a promising intelligent communication method for 6G. In contrast to traditional communication methods, semantic communication is task-oriented, extracting and transmitting only information related to tasks [ 9 ] . This emerging communication paradigm extracts semantic features from the source information, obtains semantic information, and transmits the required meaning of physical objects rather than raw data, providing reliable and robust wireless communication services [ 10 , 11 , 12 ] . The key goal of semantic communication has shifted from ensuring the accuracy of transmitted data and signal waveforms to ensuring the accurate understanding and correct use of transmitted semantic information, while prioritizing the conveyance of intended meaning with minimal data [ 13 , 14 ] . This mode of communication not only minimizes data redundancy through semantic extraction, improving data transmission efficiency, but also shows strong robustness under conditions of limited bandwidth and low signal-to-noise ratio, effectively alleviating the bottleneck of channel capacity limitations [ 15 ] . To reduce the latency of UDs transmitting synchronization data and enhance DT synchronization efficiency, we develop a semantic communication-based DT synchronization framework. This framework aims to transmit the semantic information of physical objects rather than raw sensed data, conveying the intended meaning with the minimal amount of data possible, thereby reducing communication overhead and effectively lowering synchronization latency. In contrast to traditional synchronization methods, our proposed framework requires semantic extraction of the raw sensed data collected by UDs before transmission, followed by the recovery of semantic information on the edge server. While transmitting more raw sensed data can improve synchronization accuracy, it also increases the time required for synchronization. Therefore, an efficient synchronization strategy is necessary to determine the optimal semantic extraction approach and find the best resource allocation for semantic extraction and recovery, adhering to constraints on energy consumption, synchronization latency, and computational resource. In conclusion, the contributions of this work are as follows: • We propose a semantic communication-based DT synchronization framework in MEC that considers UDs’ mobility and data sensing process. To balance synchronization latency and accuracy, a semantic extraction factor is introduced. The DT synchronization problem is formulated as a latency minimization problem, subject to constraints on energy consumption, synchronization latency, and computational resource. • The formulated problem is a time-varying optimization problem, where the mobility of UDs leads to dynamic changes in the wireless environment, making it challenging for traditional convex-based methods to solve. To tackle this challenge, an efficient deep reinforcement learning (DRL) algorithm based on soft actor-critic (SAC) is proposed by jointly optimizing the semantic extraction factor, transmission power of UDs, and computational resource allocation for both UDs and the edge server. • We analyze the complexity of the proposed SAC-based algorithm and demonstrate its convergence. By analyzing the numerical results, our proposed SAC-based algorithm demonstrates superior performance in reducing the synchronization latency of UDs, compared to other benchmark schemes. The remainder of this paper is structured as follows. Section II surveys related work. Section III describes the system model. Section IV formulates the optimization problem. Section V presents the solution to the problem under consideration. Numerical results and conclusions are detailed in Sections VI and VII, respectively. II Related Work In this section, we introduce the related work on DT and semantic communication. The DT technology can effectively assist MEC systems in building network topologies and optimizing resource allocation [ 2 ] . For example, Guo et al. [ 16 ] proposed a resource allocation method based on federated learning (FL) for device-to-device aided DT edge networks, aiming to enhance network performance and address privacy issues. Liu et al. [ 17 ] presented a DT based unmanned aerial vehicle (UAV) assisted MEC network and utilized multi-agent collaboration to minimize the weighted energy consumption of users and UAVs. Hazarika et al. [ 18 ] introduced a

Read the complete text