|
NaNA2025 Keynote 1: |
|
Speaker:
Prof. Hui Li
Title:
AI Application Security and Privacy Computing in B5G Satellite-Terrestrial Integrated Network
Affiliation:
Xidian University, China
|
| Abstract:Satellite networks providing low-latency, high-bandwidth, and low-cost communication services have attracted extensive attention. With the development of B5G networks, AI/ML technologies have gradually become the core force to enhance the performance of satellite networks. Especially in the satellite core network, implementing AI technologies on satellites can significantly improve data processing efficiency. However, the data used in the AI training process on the satellite core network may lead to issues such as user privacy leakage. This report first introduces the typical architecture, network characteristics, standard progress, and security challenges of satellite networks, then discusses the application of AI technologies in satellite networks, proposes the requirements and necessity of deploying AI on satellites, and further discusses the importance of privacy computing technologies in ensuring data security and privacy protection in satellite AI applications.
|
Short bio:
Prof. Hui Li, received B.Sc. degree from Fudan University in 1990, M.Sc. and Ph.D. degrees from xidian University in 1993 and 1998. In 2009, he was with Department of ECE, University of Waterloo as a visiting scholar. Since 2005, he has been the professor in Xidian University, China. Now, he is the executive dean of School of Cyber Engineering. His research interests are in the areas of cryptography and security protocl, privacy computing, information theory and coding . He has published over 300 papers in academic journals and conferences. His Google Scholar H-index is 67.
|
NaNA2025 Keynote 2: |
|
Speaker:
Prof. Limei Peng
Title:
Adaptive and Scalable SFC Deployment in Dynamic SAGIN Architectures
Affiliation:
Kyungpook National University, South Korea
|
| Abstract:As the Space-Air-Ground Integrated Network (SAGIN) continues to evolve toward large-scale, dynamic, and service-oriented architectures, deploying Service Function Chains (SFCs) across domains and layers remains a key technical challenge. The dynamic topology, heterogeneous resources, and continuous integration of new entities demand adaptive, scalable, and generalizable orchestration strategies. In this talk, we present a unified framework for SFC deployment in TEG-modeled SAGIN, where the Time-Expanded Graph (TEG) captures both temporal and structural variations in the network. We first introduce a deep reinforcement learning (DRL)-based approach for real-time, latency-aware SFC placement in dynamic SAGIN environments. To further support cross-domain generalization and adaptation to new entities, we extend our solution with a Federated Meta-Learning framework, enabling distributed learning and fast policy adaptation without centralized retraining. Our experimental results demonstrate that the proposed methods significantly improve SFC success rate and responsiveness under varying topologies and domain shifts. We conclude with a discussion on the role of distributed and adaptive intelligence in enabling robust and scalable service delivery across future space-air-ground networks.
|
Short bio:
Prof. Limei Peng is with the School of Computer Science and Engineering, Kyungpook National University (KNU), Daegu, South Korea. Her research interests include edge computing, wireless networks, Internet of Things (IoT)/Internet of Vehicles (IoV), and 5G/6G communications.
|
|
2023-2025
All Rights Rreserved.
|
| |
