Content for TR 28.915 Word version: 19.0.0
1 Scope p. 7
The present document will study scenarios which use NDT. For these scenarios, the study will identify issues, potential requirements and potential solutions, based on which the recommendations can be provided.
2 References p. 7
The following documents contain provisions which, through reference in this text, constitute provisions of the present document.
- References are either specific (identified by date of publication, edition number, version number, etc.) or non-specific.
- For a specific reference, subsequent revisions do not apply.
- For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same Release as the present document.
[1]
TR 21.905: "Vocabulary for 3GPP Specifications".
[2]
TS 28.104: " Management and orchestration; Management Data Analytics (MDA)".
[3]
"Emulation versus simulation: a case study of TCP-targeted denial of service attacks". Available at https://www.cs.purdue.edu/homes/fahmy/papers/tridentcom.pdf .
[4]
TM forum IG1307: "Digital twin for decision intelligence (DT4DI) Whitepaper".
[5]
ETSI GR ZSM 015: "Zero-touch network and Service Management (ZSM); Network Digital Twin".
[6]
TS 28.541: "Management and orchestration; 5G Network Resource Model (NRM); Stage 2 and stage 3".
[7]
TS 28.552: "Management and orchestration; 5G performance measurements".
[8]
TS 28.554: "Management and orchestration; 5G end to end Key Performance Indicators (KPI)".
[9]
TS 28.310: "Management and orchestration; Energy efficiency of 5G".
[10]
TS 28.622: "Telecommunication management; Generic Network Resource Model (NRM) Integration Reference Point (IRP); Information Service (IS)".
[11]
TS 28.532: "Management and orchestration; Generic management services".
[12]
TS 32.422: "Telecommunication management; Subscriber and equipment trace; Trace control and configuration management".
3 Definitions of terms, symbols and abbreviations p. 7
3.1 Terms p. 7
For the purposes of the present document, the terms given in TR 21.905 and the following apply. A term defined in the present document takes precedence over the definition of the same term, if any, in TR 21.905.
Network Digital Twin (NDT):
virtual replica of mobile network or part of one, that captures its attributes, behaviour and interactions.
3.2 Symbols p. 8
Void.
3.3 Abbreviations p. 8
4 Concept and background p. 8
4.1 General description p. 8
Digital twin technology provides robust support for emerging technologies by creating a comprehensive virtual mapping of the corresponding physical network process, utilizing models, operational history, and additional data.
3GPP already uses the Network Resource Model (NRM) to model the attributes of a mobile network. The concept of Network Digital Twin also adds the ability to model the behavior of a mobile network. This behavior is modelled by emulating or simulating a complete mobile network or limited aspects of a mobile network.
Network Digital Twin (NDT) may be used as a replica of a mobile network, in order to learn how an actual mobile network would behave in certain scenarios, without causing any changes to the actual mobile network. To provide meaningful results, the Network digital twin needs to emulate (or simulate) the behavior of the mobile network, so that the result of the operations on the virtual replica are a good approximation to similar operations on the actual network. The standardization for an NDT focuses on the implementation of independent aspects of a network.
Digital twin technology has potential scenarios in enhancing the 3GPP management system. For example, the NDT can help for efficient simulation of the network operation, the configuration and policy decided by the 3GPP management system can be verified before the deployment. By using this Network Digital Twin, the 3GPP management system can obtain verification results and optimize configurations, thereby avoiding failures in the actual network. This approach benefits the optimization of network management in the telecommunications industry, reduces the cost of study and development of new technologies, and shortens the study and development cycle of new technologies.
Existing automation capabilities include analytics services and decision-making capabilities with the assumption that the knowledge about network behavior is available within the automation functions. Accordingly, the unaddressed gap for network automation is the capabilities for modelling the behavior of the network. This knowledge on how the network behaves or will behave is the new value provided by digital twins that understands and models the behavior of the network.
4.1.1 Scope of NDTs p. 8
4.2 Introduction and Overview p. 8
4.2.1 Digital Twins and Network Digital Twins p. 8 
A Digital twin is a representation of an object that models or simulates the characteristics and behaviors of the physical object. A digital twin can be created for any physical object, including any objects in communication networks. The digital twin may also be created for a group of objects, e.g. for the sets of network objects that form the RAN of a city.
Accordingly, a digital twin modelling an object of a communication network may be called Network Digital twin
4.2.2 Relations between digital twins and network automation functions p. 9
The NDT can be requested by the consumer to model a mobile network or part of one, support evaluating the corresponding impact, and return the report of the simulated impact generated by the NDT. The NDT may be used as a replica of a mobile network, which may synchronize data from the managed network for modeling. The digital twins provide modelling capabilities that are used by the network automation functions (e.g. MDA, SON, etc.) to accomplish their automation functionality. The related automation capabilities are provided by the network automation functions regardless of whether the digital twin models are integrated within or external to the network automation functions - see Figure 4.2.2-1.
4.2.3 Relation between emulation and simulation p. 9
4.2.3.1 Emulation p. 9
Emulation uses a system's actual algorithms or functions to mimic how a system will behave. For NDT, duplicates of the network functions and/or network management functions are executed in an NDT environment.
To emulate the behaviour of a mobile network, it is necessary to create an NDT environment which contains virtualized network equipment, network functions, network management functions, and all the configuration and status data for this equipment/functions. To measure the reaction to network traffic, the NDT environment also contains traffic generators.
4.2.3.2 Simulation p. 9
Simulation uses a mathematical model to mimic how a system will behave. For NDT, models of the behaviour of network functions and/or network management functions to mimic the behaviour of the overall mobile network (or part thereof).
To simulate the behaviour of a mobile network, it is necessary to create an NDT environment which combines the models of network equipment, network functions and/or network management functions, with the relevant configuration and status data for this equipment/functions. To measure the reaction to network traffic, the network traffic is also modelled.
4.2.3.3 Comparison of emulation and simulation p. 10
Emulation has the advantage of more accurate behaviour, especially in complex systems that are experiencing abnormal cases. Emulation also has the advantage that there is no need to create a mathematical model of the behaviour of each individual component. The vendor-provided software for each emulated component can be executed in the emulation environment and should produce the expected behaviour.
Emulation has the disadvantage that it is resource-expensive, because the emulation environment will require a similar amount of compute/storage/network resources as a real network. Therefore, the primary advantage of simulation is to reduce cost.
A major disadvantage of simulation is the need to create models of how each component will behave. The typical or expected behaviour of equipment or a function may be possible to model easily. But in extreme cases (such as overload or error), only the vendor knows exactly how the equipment or function will behave.
It may be possible to combine emulation and simulation to create an integrated solution. Because emulation and simulation are not mutually exclusive and each has value in different scenarios, NDT need to support both emulation and simulation.
4.3 Potential uses of NDTs p. 10
NDTs may support many use cases in network management and automation. In all the use cases, NDTs provide modelling capabilities that are then applied by the network management and automation functions or applications to achieve the desired outcomes.Use cases where NDT may provide support include:
- Use case 1: Network management RAN ES policy verification using NDT
- Use case 2: Signalling storm analysis
- Use case 3: Emergency preparedness (see clause 5.3).
- Use case 4: Network failure and risk prediction (see clause 5.4).
- Use case 5: NDT support to network automation
- Use case 6: Using NDT to generate ML training data
- Use case 7: Nested NDTs
- Use case 8: Visualization of network topology and traffic
- Use case 9: Configuration verification
- Use case10: Network issue inducement
- Use case 11: Measuring customer satisfaction with the network services
4.4 Illustrate the life-cycle management of an NDT p. 10
When the MnS consumer submits a request to create an NDT, the MnS producer who provides the NDT may create an NDT instance to fulfil or satisfy the specific scenarios (see clause 5). NDT LCM may include the following capabilities:
- Creation.
- Configuration.
- Activation
- Deactivation
- Re-configuration.
- Deletion
