Workshop Date: Friday, May 10th, 2024
Program
The workshops chairs welcoming are participants and presenters to the GAIN 2024 workshop at NOMS24. We are looking forward to meet you in Seoul. We have compiled the following program:
09:00 Welcome
09:05 Paper 1: Intent Assurance using LLMs guided by Intent Drift
(K. Dzeparoska, A. Tizghadam , A. Leon-Garcia, University of Toronto, Canada)
09:35 Paper 2: Utilising Generative AI for Test Data Generation - use-cases for IoT and 5G Core Signalling
(T. Tothfalusi, AITIA Int. l Inc., Hungary, Z. Csiszar, P. Varga, Budapest Uni. of Tech. and Econ., HU.)
10:05 Paper 3: GAN Enhanced Vertical Federated Learning System for WHAR with non-IID Data
(C. Lee , S. Cho, H. Park, J. Park, S. Lee, Yonsei University, South Korea)
10:35 Coffee Break
11:00 Paper 4: S-Witch: Switch Configuration Assistant with LLM and Prompt Engineering
(E.Jeong , H. Kim, S. Nam, Pohang Uni. of Sci. and Tec., South Korea), J. Yoo, J. W. Hong,
POSTECH, South Korea)
11:30 Paper 5: Impact of Graph-to-Sequence Conversion Methods on the Accuracy of Graph Generation for Network Simulations
(K. Yasuda , S. Tsugawa, K. Watabe, Nagaoka University of Technology, Japan)
12:00 Community Panel
12:25 Wrap-up and Farewell
Presenters Information: each presentation slot is assigned 25min for presentation and 5min for Q&A.
CALL FOR WORKSHOP PAPERS
The GAIN workshop aims to systematically discuss and investigate the use of Generative AI in network and service management. Generative AI is an exciting technology that has recently drawn tremendous amounts of attention. Large Language Models (LLM), as part of Generative AI, provided answers that amazed domain experts. Generative AI, however, is expected to be more than a hype. It builds on the profound theory and concepts of generative statistical models that can identify and exploit complex statistical relationships in large data amounts, such as joint probability distributions.
Network Management is traditionally data-driven, particularly in large-scale and public communication networks. For example, these systems produce large error logs. Moreover, the service quality is typically constantly monitored and reported, leading to large data sets. In addition, modern communication networks are highly complex in their structure and nowadays don’t only represent physical relationships but also often virtual ones, like slices in SDN, network function virtualization, disaggregated RAN (OpenRAN), Cloud-native network operation, MEC, etc. Furthermore, network management is typically costly, which is often caused by a low degree of automation. The use of data-based AI and, particularly, the application of Generative AI might enable overcoming these impasses in modern networks.
GAIN will bring together academic researchers from various disciplines (communication networks, data science, operational research) and practitioners from industry. The workshop welcomes scientific papers as well as industrial use case papers. The considered topics in Generative AI for network management are initially structured along the well-accepted FCAPS model for network and network service management. However, the workshop is not limited to these topics, and novel contributions on using Generative AI in the operation and management of networks and services are welcome.
Predictive Maintenance: Using generative AI for proactive network management
Network Troubleshooting using Generative AI, incl. root cause analysis and resolution
Monitoring using Generative AI: Using generative AI for efficient monitoring of network resources
Configuration Management
Network Configuration Automation with Generative AI
Automated Network Design and Deployment using Generative AI
Generative AI for Traffic Management: Optimizing network traffic engineering through AI
Accounting
Ethical Considerations: Addressing Privacy and Security Concerns in AI-Based Network Management
Ensuring fairness between network users using generative AI for optimal resource allocation
Network Optimization with AI: Leveraging generative algorithms for network efficiency
Dynamic Resource Allocation: Leveraging generative AI for efficient network resource management
Efficient Network Data Analysis using Generative A
Security
AI-Based Security Protocols: Developing next-generation network security strategies
Anomaly Detection and Response: Utilizing generative AI for enhanced network security
Use Cases
Generative AI for management of IoT, Wireless/RAN, or Core, and Cloud-to-Edge networking
Prompt Engineering for Network Management Using LLMs
Robustness and Reliability of Generative AI for net. management (incl. benchmarks and datasets)
Scalability Orchestration, Testing and Validation of Generative AI for Network Management
Submission site: https://jems3.sbc.org.br/noms_gain2024
Paper Submission Deadline: Jan. 19, 2024
Feb. 2, 2024
(extended)
Notification of Acceptance: Mar. 1, 2024
Final Camera Ready: Mar. 15, 2024
Alberto Leon-Garcia, Univ. of Toronto, CA
(alberto.leongarcia@utoronto.ca)
Pal Varga, Budapest Univ. of Technology and Economics, HU
(pvarga@tmit.bme.hu)
Kurt Tutschku, Blekinge Inst. of Technology, SE
(ktt@bth.se)
Authors are invited to submit original contributions to the workshop that are written in English and that have not been published or submitted for publication elsewhere. Workshop papers must be submitted as PDFs using the IEEE conference double-column format style. (Style templates can be found here: https://www.ieee.org/conferences/publishing/templates.html. ) A NOMS workshop paper must not exceed six pages (excluding references) for full paper submissions and not exceed four pages (excluding references) for short paper submissions.
All submitted papers will be peer-reviewed. For accepted papers, at least one author is expected to register and present the paper in person at the workshop. Accepted and presented papers will be published in the conference proceedings and submitted to IEEE Xplore.
