Tasks 5.2 and 5.3 Progress and Future Activities
Objectives related to tasks 5.2 and 5.3:
The objective of this task is twofold.
Hardware Platform Implementation: The project has moved from conceptual design to concrete implementation with Kubernetes orchestration.
AI-Powered Orchestration Framework: The team developed an integrated system combining traffic prediction using LSTM and ARIMA models with dynamic VNF placement. Their testing shows LSTM models outperform ARIMA for CPU usage prediction, achieving better accuracy across multiple metrics (MAPE, MAE, RMSE).
Dynamic Resource Management: The orchestrator can proactively allocate resources based on ML predictions, with capabilities for real-time network monitoring and anomaly detection. The system aims to reconfigure resources within minutes during network disruptions.
Security Framework: A three-phase security mechanism was identified: proactive detection before deployment (currently operational), real-time reaction during runtime, and threat mitigation (planned for phase 2). Only first phase has been implemented.
Comprehensive State-of-the-Art Analysis: The document includes extensive coverage of VNF orchestration challenges, reviewing 60+ academic sources and comparing existing solutions like Kubernetes, OpenStack, and ONAP for NTN environments.
NTN-Specific Challenge Identification: The team identified critical challenges including resource constraints in satellite systems, network heterogeneity across orbital layers (LEO, MEO, GEO), and dynamic topology management due to satellite mobility.
Phase 2 Planning: A structured roadmap for the second project period includes algorithm development, testing/optimization, deployment evaluation, and comprehensive reporting.
While the technical progress is substantial, several limitations emerge:
Version 3.0 presents a completely rewritten executive summary that:
> Other sections are still unsatisfactory. For example the cybersecurity section, which is very limited in detail by comparison with the "comprehensive analysis" that is targeted.
Answer: As described in the DoW of the project, Task 5.3 is related to secure communications between the VNFs. We do not address in this task the security of NTN links.
> T5.3: Cybersecurity: Provide examples of the use of this toolkit, how it is configured, the micro-service graphs, its detection and reporting of specific vulnerabilities.
This toolkit has been only adapted to be used in our platform and has not been designed, implemented, and evaluated in the scope of 6G-NTN. Thus, we do not estimate relevant to provide this information. However, the paper describing this work has been cited and contains all required information.
> Refer to the existing state of the art established in 3GPP Rel.18 to ensure that standardised requirements and features are adopted.
We have decided to adopt a pragmatic approach. Our solution relies on leveraging an existing 5G core infrastructure, which, while not perfect, serves as a robust foundation for our VNF orchestration platform. Rather than seeking an idealized solution that meets all requirements, we are using available open-source software to drive forward with our objectives. This approach is necessitated by the complexity and scope of R18, which poses significant implementation challenges that are more feasible in a controlled test environment.
> Note that STR-FUNC-05 and STR-FUNC-20 suggest that edge computing capabilities are required. Thus they cannot be excluded from the orchestration platform design and its functionality, even if, as stated they will not be implemented.
Most of the requirements are covered especially the first ones. However, STR-FUNC-11, for instance, is too specific to be considered here.
> - D4.2 describes the functions 6G-NTN RIC, also driven by AI and also managing use of system resources. These must be orchestrated themselves. RIC (radio) resource management decisions could potentially conflict with service management decisions concerning the orchestration of VNFs.
An orchestrator can take into account such a RIC. But for now, we first focus on orchestration of different VNFs depending on what we can achieve. The cohabitation between 2 intelligent components will not be considered in 6G-NTN.
> Additional to the further detail requested about the platform and its realisation, describe the deployment (planned or actual) of the 5G Core and any functions related to the NTN air/space-borne RAN.
Integration of 3GPP NG RAN functions are not planned to be integrated into the platform.
> Provide additional analysis of APIs relevant to orchestration and service management, including security mechanisms, especially referencing relevant work of 3GPP, ORAN Alliance and ETSI.
We do not plan to implement these specific functions in our core network. Our plan is to use a containerized core network and demonstrate the feasibility of VNF placement.
> The first part of this section is only about one aspect of Kubernetes deployment security and should be separated out into a section on its own with a new title.
This section is all about securing VNFs deployment and we focused on threats related to Kubernetes deployments. We have adapted an already existing tool developed by TH-SIX to be used as part of the platform.
> As noted before, there is a substantial state of the art for all of the topics mentioned in this section. The topics are presented as planned work but the report would benefit from references to the security challenges described by other experts. Security of integrated TN and NTN systems must also be addressed.
In this section, we have described what are the challenges related to securing 6G NTN links but we do not plan to address these challenges in the scope of this project.
> Provide state of the art analysis and references for the security topics mentioned in new section "SECURING 6G NON-TERRESTRIAL NETWORKS", e.g. IEEE Future Networks NGNR roadmaps, NextG Alliance, ORAN Alliance, and 3GPP, addressing system, API and orchestration security.
As described in the DoW of the project, Task 5.3 is related to secure communications between the VNFs. We do not address in this task the security of NTN links.