Integrated Sensing and Communication Driven Digital Twin for Intelligent Machine Network

TL;DR

The paper addresses the challenge of integrating ubiquitous sensing with real-time communication in intelligent machine (IM) networks by proposing an Integrated Sensing and Communication driven Digital Twin (ISAC-DT) architecture. It introduces a three-layer architecture (Physical Network and Environment Layer, ISAC Digital Twin Layer, Network Application Layer) and three DT-layer modules (data repository, service mapping, and twining model management) within a framework of two closed-loops to enable policy-driven optimization of sensing, communication, and networking. The ISAC-DT enables DT-based environment sensing, channel estimation, beam management, MAC, neighbor discovery, and resource/reliability management, all supported by DT-split edge-cloud construction and DT-driven data fusion and learning. Performance evaluations on cooperative localization, environment reconstruction, beam tracking, and neighbor discovery illustrate tangible improvements in sensing accuracy, spectral efficiency, and network reliability, highlighting the practical potential for flexible manufacturing and other IM network scenarios.

Abstract

Intelligent machines (IMs), including industrial machines, unmanned aerial vehicles (UAVs), and unmanned vehicles, etc., could perform effective cooperation in complex environment when they form IM network. The efficient environment sensing and communication are crucial for IM network, enabling the real-time and stable control of IMs. With the emergence of integrated sensing and communication (ISAC) technology, IM network is empowered with ubiquitous sensing capabilities, which is helpful in improving the efficiency of communication and sensing with the mutual benefit of them. However, the massive amount of sensing information brings challenges for the processing, storage and application of sensing information. In this article, ISAC driven digital twin (DT) is proposed for IM network, and the architecture and enabling technologies are revealed. ISAC driven DT structurally stores the sensing information, which is further applied to optimize communication, networking and control schemes of IMs, promoting the widespread applications of IMs.

Figures (5)

• Figure 1: Scenarios of IM network, including flexible manufacturing, IoV, and UAV network.
• Figure 2: Architecture of ISAC-DT enabled IM network.
• Figure 3: Use case of ISAC-DT enabled IM network.
• Figure 4: Cooperative target localization and environment reconstruction.
• Figure 5: ISAC-DT based beam management and neighbor discovery in IM network.