Networked Twins and Twins of Networks: an Overview on the Relationship Between Digital Twins and 6G

I. INTRODUCTION

• Researchers debate which proposed technologies will be mature enough for 6G, while the generation itself remains not fully defined.Proposals include higher frequency bands, OAM multiplexing, intelligent surfaces, AI, and terrestrial-satellite integration,,.Source: §I · p. 1
• Immersive and autonomous applications are expected to require 99.999% reliability and one millisecond latency, beyond what 5G can achieve at scale,.The paper also identifies 100 Gbs throughput and 99.999% reliability as requirements for realising DTs’ potential.Source: §I · p. 1
• DT popularity depends on application popularity and necessity; potential high-connectivity-demanding, rapidly emerging applications range from aerospace, with very high mobility, to Industry 4.0, with many devices in one location.Source: §I · p. 1

1) Pillars:

• A complete DT system comprises physical, digital/virtual, and connection pillars.The digital pillar maintains asset history and current information for data-supported decision-making, distinguishing DTs from augmented virtuality.Source: 1) Pillars · p. 2, §II · p. 2
• DT applications could drive 6G, while DT technology integrated with AI could facilitate 6G development.The paper outlines its discussion of DTs, future communication networks, adoption enabled by 6G, and research directions.Source: 1) Pillars · p. 2

A. Key Features of Digital Twin

• A DT system links a physical twin to a digital pillar that hosts models, historical data, decision support, AI, and visualisations.The connection pillar carries data and control commands, and its QoS requirements may differ by direction.Source: A. Key Features of Digital Twin · p. 2

2) Modularity:

• Modularity allows component DTs to be interconnected into a mega-DT and combined with simulations of subsystems without physical twins.Hybrid systems can support prototyping and performance testing within the boundaries of the data models representing the physical twins.Source: 2) Modularity · p. 2

3) Remote Intelligence:

• Remote intelligence lets resource-limited devices or older machinery use analysis or AI running on a DT at the edge or in the cloud.Source: 3) Remote Intelligence · p. 2

B. Standardisation

• DT standardisation efforts address data collection, storage, and exchange, while comprehensive standards remain lacking, especially for edge deployment.DTDL addresses interoperability for Azure-based DTs; FMI and existing standards such as OPC-UA are also discussed.Source: B. Standardisation · pp. 2–3

III. DIGITAL TWIN OF COMMUNICATION NETWORKS

• Telecommunications DTs may improve network design and maintenance across the network life cycle.Companies have begun developing DTs for 5G network components, activities described as paving the way for 6G adoption.Source: §III · p. 3

1) Design and deployment:

• DT-based design combines existing component twins with simulation, then incorporates each deployed subsystem’s twin as deployment proceeds.The resulting hybrid system becomes a complete DT connected to physical subsystems.Source: 1) Design and deployment · p. 3

2) Smart operation, maintenance, and resilience:

• AI-enabled DTs use real-time data and prior experience to optimize operational parameters and plan for potential disruptions beyond self-optimization and self-healing.DT modularity supports edge intelligence, federated learning, and transfer learning; prediction can reduce reliance on costly redundant copies.Source: 2) Smart operation, maintenance, and… · p. 3
• DTs can learn equipment lifetimes from manufacturer data, physical-twin data, and external factors, then adjust working conditions or schedule maintenance.In real scenarios, optimization should also consider service and other criteria.Source: 2) Smart operation, maintenance, and… · p. 3

3) Knowledge transfer, and robust expansion:

• Disconnected component or network twins can support new-network design and service testing, while operators may monetize experience by selling data and knowledge.The paper gives transfer of knowledge from a 5G network to 6G design as an example.Source: 3) Knowledge transfer, and robust… · p. 4

B. DT of the next-generation of networks

• DT-based network design and operation can benefit mostly in 6G, whose KPIs and AI synergy are described as supporting the overhead of a network’s own DT.The passage contrasts this with 5G, which it says cannot handle that overhead while supporting high-throughput or ultra-delay-sensitive services.Source: B. DT of the next-generation of networks · p. 4

• Wider DT adoption may create capacity demands and scenarios beyond 5G, while DT communication must support ultra-reliable, real-time or application-dependent semi-real-time, high-QoS exchange.Source: §IV · p. 4
• 99.999% reliability corresponds to five minutes of yearly downtime, compared with 5G’s 99.9% reliability and eight-and-a-half hours, making 6G more advantageous for massive-scale industrial IoT facilitated by DTs.The passage cites General Electric’s reported $1.6B saved through early detection of industrial component failures.Source: §IV · p. 4
• A mega-DT for an industrial IoT factory distributes component twins across cloud and edge servers, with 6G described as supporting millisecond latency, 100 GB/s data rate, and 99.999% reliability.The figure description identifies red dotted lines as logical bidirectional connections between the physical twin and its digital twins.Source: §IV · pp. 4–5
• Distributed federated AI deployments may require 99.999% reliability and one-millisecond latency, beyond the small-scale federated AI implementations already using 5G.The passage describes distributing neural-network components across edge nodes and cloud processing resources.Source: §IV · p. 5
• Modular DTs distributed across cloud or edge servers create challenges for seamless communication, distributed AI computation, and data security and integrity; blockchain transactions require 10,000 transactions per second and millisecond latency, which the passage says requires 6G-level QoS.Source: §IV · p. 5

A. DT Ownership Issues

• DT ownership is challenging because the physical entity and DT platform may have different owners, creating technical, financial, and legal issues.The passage describes an appliance owner who may rent cloud, fog, or edge services to install and maintain the DT.Source: A. DT Ownership Issues · p. 5
• Ownership scenarios for process or system DTs with components owned by multiple entities remain open for further research, especially in industrial 6G settings.Source: A. DT Ownership Issues · p. 6

B. Ultra-Low-Latency and Reliable Communication between DT and PT

• Most Digital Twin data and analytics may reside in the cloud, though edge deployment can suit critical cases such as remote surgery.Real-time, reliable exchange with the physical twin is a necessary condition for a Digital Twin system.Source: B. Ultra-Low-Latency and Reliable… · p. 6
• Further breakthroughs across all network protocol layers are still needed to meet strict latency and reliability requirements.Prior studies report relevant technologies and algorithms, but the text says more work remains.Source: B. Ultra-Low-Latency and Reliable… · p. 6

C. Federated DT in the Cloud/Edge

• Cloud/edge distribution or replication can address hosting limits and connectivity failures that may bottleneck a Digital Twin kept in one place.The text anticipates moving much computation and storage to the edge while hosting backend solutions across multiple cloud and/or edge data centers.Source: C. Federated DT in the Cloud/Edge · p. 6
• Federated Digital Twins coordinate data exchange and AI-model training across components, but synchronized collaborative AI remains an open research area.The text describes this coordination across network nodes as challenging.Source: C. Federated DT in the Cloud/Edge · p. 6

D. DT of an Entire Network

• A composite network Digital Twin could intertwine twins of physical components with software-based network components for management, upgrades, troubleshooting, and pre-deployment testing.The paper notes that most network components apart from physical infrastructure can be implemented as cloud-native software.Source: D. DT of an Entire Network · p. 6
• Figure 4 highlights AI-based analytics for network monitoring and troubleshooting, alongside smart-contract ownership issues hosted in a Blockchain.The figure is presented as a vision for enabling a Digital Twin of an entire network.Source: D. DT of an Entire Network · p. 6, Figure 4 · p. 7

E. Experimental Investigation of DTs

• A documented LTE setup using Amarisoft LTE 100 eNodeB, srsUE, and a generic RF front end is proposed as a starting point for network-Digital-Twin experiments.The setup switched the network on and off, streamed YouTube video, and collected and plotted performance data in real time; a GUI and real-time connections between twin and physical representations would still be needed.Source: E. Experimental Investigation of DTs · p. 6

VI. CONCLUSIONS

• 6G can facilitate Digital Twin realization and adoption across industries by providing the required reliability and speed.Source: §VI · p. 7
• AI-integrated Digital Twins can facilitate 6G network design, deployment, and operation, with potential impact on network resilience.Source: §VI · p. 7
• Demanding Digital Twin applications in aerospace, Industry 4.0, and healthcare could be a major driver toward 6G development.The paper also highlights potential Digital-Twin-related research directions.Source: §VI · p. 7