| Short name | Deliverable | Abstract |
|---|---|---|
| 6G-INTEGRATION-3-E5 | Innovations for the NTN integration with 3GPP networks | This document reviews the fundamentals of satellite communications and the latest advances in fault-tolerant onboard equipment, AI/ML-based applications in STIN, and advancements and deployments in Non-Terrestrial Networks (NTNs). Additionally, the document delves into the 3GPP Release 17 standard in the context of NTN and analyzes the state of the art in hardware fault tolerance strategies in the space segment, as well as the applications of AI/ML in optimizing the operation and performance of satellite communications and High-Altitude Pseudo-Satellites (HAPS). Finally, the document concludes with a brief summary of the contributions and the analyzed state of the art. |
| 6G-INTEGRATION-3-E6 | Enhanced innovations for the NTN integration with 3GPP networks | This document provides a summary of examples for enhanced innovations in the NTN integration with 3GPP networks. The main innovations overviewed have to do with the applicability of modern AI/ML algorithms to help modeling, solving and optimizing different aspects of NTNs built with both terrestrial equipment and on-air and space equipment. Main challenges in these types of networks have to do with dealing with high latency, the errors due to radiation and transmission impairments, Doppler shifts, mobility management, handovers between terrestrial and non-terrestrial networks, channel reliability. Multi-access Edge Computing (MEC) can provide caching and computing resources on board to provide near real-time applications to minimise latency. Also intelligent algorithms based on Reinforcement Learning and other AI/ML strategies can be used to optimise network performance from multiple sides: resource optimization, optimal routing, network slicing and mobility management. This document provides an overview of such strategies and algorithms toward a real integration of both terrestrial and non-terrestrial networks with current 5G deployments and emerging 6G networks. |
| 6G-INTEGRATION-3-E7 | Final innovations design for NTN integration with 3GPP networks | This document presents key innovations for integrating High Altitude Platform Systems (HAPS) into non-terrestrial networks (NTN) aligned with 3GPP and 6G standards. It highlights the application of advanced AI and machine learning algorithms, such as reinforcement learning, to optimize resource allocation, routing, network slicing, and mobility management in highly dynamic NTN environments.
The work introduces efficient frame transmission strategies for LEO satellites and HAPS, including “withhold scheduling,” which balances data loads across ground stations to improve throughput and latency. Deep reinforcement learning agents are developed for optimal routing, adapting to real-time changes in network topology and congestion. A modular drone platform equipped with edge computing and 5G connectivity is designed and deployed to validate these innovations in real-world NTN scenarios. The document also analyzes the stringent bandwidth, latency, and reliability requirements of emerging AR/VR applications, informing the design of MEC-enabled HAPS nodes for distributed caching and processing. A convergent NTN-6G architecture is proposed, integrating MEC at HAPS nodes to support seamless handovers and ultra-low latency. Conclusions emphasize the need for holistic co-design of algorithms, hardware, and standards, and identify future research directions in scalable AI, open interfaces, post-quantum security, and advanced materials for HAPS platforms. These contributions form a comprehensive roadmap for scalable, reliable, and high-performance NTN integration towards 6G. |
