- 13:15 - 13:25: Reception and Welcome
- 13:25 - 14:05: Keynote
- A Framework for Multi-Agent Distributed Retrieval Augmented Generation Systems
- Ken Huang - ken.huang@distributedapps.ai - CISSP
- 14:05 - 14:45: Technical Session #1
- Incremental Federated Host Embeddings for Network Telescopes Traffic Analysis
- Kai Huang - kai.huang@polito.it - Italy, Politecnico di Torino
- Luca Gioacchini - luca.gioacchini@polito.it - Italy, Politecnico di Torino
- Marco Mellia - marco.mellia@polito.it - Italy, Politecnico di Torino
- Luca Vassio - luca.vassio@polito.it - Italy, Politecnico di Torino
- FLAMINGO: Adaptive and Resilient Federated Meta-Learning against Adversarial Attack
- Md Zarif Hossain - mdzarif.hossain@siu.edu - United States, Southern Illinois University, Carbondale
- Ahmed Imteaj - imteaj@cs.siu.edu - United States, Southern Illinois University, Carbondale
- Abdur R. Shahid - shahid@cs.siu.edu - United States, Southern Illinois University, Carbondale
- Incremental Federated Host Embeddings for Network Telescopes Traffic Analysis
- 14:45 - 15:15: Break
- 15:15 - 16:15: Technical Session #2
- SHELOB-FFL: addressing Systems HEterogeneity with LOcally Backpropagated Forward-Forward Learning
- Stefano Izzo - stefano.izzo@unina.it - Italy, Università degli Studi di Napoli Federico II
- Fabio Giampaolo - fabio.giampaolo@unina.it - Italy, Università degli Studi di Napoli Federico II
- Diletta Chiaro - diletta.chiaro@unina.it - Italy, Università degli Studi di Napoli Federico II
- Francesco Piccialli - francesco.piccialli@unina.it - Italy, Università degli Studi di Napoli Federico II
- LLM-Assisted System Testing for Microservices
- Mustafa Almutawa - mustafa.almutawa@kaust.edu.sa - Saudi Arabia, KAUST
- Qusai Ghabrah - qusai.ghabrah@kaust.edu.sa - Saudi Arabia, KAUST
- Marco Canini - marco@kaust.edu.sa - Saudi Arabia, KAUST
- Explainable Reinforcement Learning for Network Management via Surrogate Model
- Annalisa Navarro - annalisa.navarro@unina.it - Italy, Università degli Studi di Napoli Federico II
- Roberto Canonico - roberto.canonico@unina.it - Italy, Università degli Studi di Napoli Federico II
- Alessio Botta - alessio.botta@unina.it - Italy, Università degli Studi di Napoli Federico II
- SHELOB-FFL: addressing Systems HEterogeneity with LOcally Backpropagated Forward-Forward Learning
- 16:15 - 16:25: Closing Remarks
Aim and Scope
The IEEE AI-DCS Workshop aims at the investigation of research results and at the systematic discussion of challenges at the intersection of Artificial Intelligence and Machine Learning (AI/ML) with Distributed Computing Systems. As distributed computing permeates our daily lives through mobile, industrial IoT, cloud, edge computing, blockchain, and 5G and the next 6G networks, the demand for high-performance and robust AI/ML solutions is unprecedented. However, the technical challenges in ensuring the predictability, trustworthiness, security, and reliability of these solutions are at the forefront. As a consequence, the workshop provides actionable insights into the novel adoption of generative, incremental, adversarial, and explainable AI/ML techniques for the management and security of current and future Distributed Computing Systems.
More specifically, the main goal of the workshop is (but not limited) to face the technical issues concerning such four cutting-edge pillars: (i) the harnessing of generative models for data synthesis, content generation, and their integration into distributed workflows; (ii) the incremental adaptation of AI/ML models to new data without retraining from scratch (e.g., continuous/online learning and transfer learning) and the challenges for its practical realization in the distributed computing landscape; (iii) the securing of AI/ML models against adversarial threats, considering the implications within the distributed infrastructure and practical defenses against adversarial attacks; (iv) the integration of techniques to face the lack of transparency and interpretability in AI/ML models and then understanding and trusting their decisions to support the management of critical distributed systems.
Important Dates
- Workshop paper submission deadline (extended)
April 14th, 2024April 25th, 2024 (firm) - Notification of workshop paper acceptance
May 5th, 2024 - Submission of camera-ready workshop papers due
May 10th, 2024
Keynotes
A Framework for Multi-Agent Distributed Retrieval Augmented Generation Systems
Organizers
General Chairs
Publicity Chair
Web Chair
Technical Program Committee
- Anat Bremler-Barr, Tel-Aviv University, Israel
- Walter Cerroni, Università di Bologna, Italy
- Haiming Chen, Chinese Academy of Sciences, China
- Tomáš Čejka, Faculty of Information Technology CTU in Prague, Czech Republic
- Domenico Ciuonzo, Università di Napoli Federico II, Italy
- Claudio Fiandrino, IMDEA Networks Institute, Spain
- Danilo Giordano, Politecnico di Torino, Italy
- David Hay, The Hebrew University of Jerusalem, Israel
- Noam Koenigstein, Tel-Aviv University, Israel
- Jonatan Krolikowski, Huawei Technologies, France
- Catalin Meirosu, Ericsson, Sweden
- Marco Mellia, Politecnico di Torino, Italy
- Pham Tran Anh Quang, Huawei Technologies, France
- Solange Rito Lima, University of Minho, Portugal
- Kamal Singh, Telecom Saint Etienne, France
- Giancarlo Sperlì, Università di Napoli Federico II, Italy
- José Suárez-Varela, Telefonica Research, Spain
- Noa Zilberman, University of Oxford, United Kingdom
Call for Papers
AI-DCS aims at the investigation of research results and at the systematic discussion of challenges at the intersection of Artificial Intelligence and Machine Learning (AI/ML) with Distributed Computing Systems.
AI-DCS seeks original, completed, and unpublished work not currently under review by any other journal/magazine/conference. Topics of interest include, but are not limited to:
- Generative AI/ML Models in Distributed Systems
- Generative AI for efficient management and monitoring of network resources
- Automatic network configuration with Generative AI
- Generative AI for Traffic Engineering
- Generative AI for improving network security
- Prompt Engineering for using Large Language Models (LLMs) in distributed systems
- Strategies for training generative models across distributed nodes
- Efficient deployment of generative models in distributed environments
- Load balancing for generative model inference
- Automatic generation of diverse datasets in distributed environments (e.g., industrial IoT, mobile, vehicular, cloud computing, and edge computing)
- Incremental Learning in Distributed Systems
- AI/ML for handling dynamic data sources and network conditions
- Federated transfer-learning
- Adapting pre-trained models to distributed environments via transfer-learning
- Knowledge transfer between IoT devices
- Training meta-learning models in distributed environments
- Implementation of continuous learning algorithms in a decentralized fashion
- Edge-to-cloud communication for model updates
- Resource-efficient continual learning in IoT and edge devices
- Adversarial Learning in Distributed Systems
- Adversarial threats in federated learning setups
- Privacy-preserving training strategies in distributed adversarial environments
- Secure AI/ML model deployment in distributed systems
- Adversarial defense mechanisms in distributed environments
- Trade-offs between security and model performance in decentralized systems
- Threat models for distributed applications based on AI/ML
- Explainable AI in Distributed Systems
- Fairness, accountability, and transparency in AI/ML for networking
- Explainable AI techniques for distributed models
- Reliability of AI/ML methods in critical distributed applications
- Explainable machine learning models for network performance optimization
- Interpretability in AI/ML-based network traffic analysis and management tools
- Evaluation methods for explainable AI/ML in distributed systems
- Human-in-the-loop distributed systems
- General
- AI/ML and its applications in distributed systems
- AI/ML and its applications to industrial IoT systems
- AI/ML and its applications to cloud and edge computing
- AI/ML and its applications to blockchain
- AI/ML and its applications for securing Distributed Computing Systems
- AI/ML for network anomaly and misuse detection
- ML and DL approaches for network traffic analysis and management
Submission Guidelines
Authors are required to submit fully formatted, original papers (in PDF format) via Easy Chair through >>this link<<.
All workshop papers are limited to no more than 6 pages, including references, in the IEEE format aligned with the IEEE ICDCS 2024 main conference guidelines (ICDCS24 CfP). Each submission must be written in English, accompanied by a 75 to 200 words abstract that clearly outlines the scope and contributions of the paper. Papers exceeding 6 pages will not be accepted by Easy Chair. At least one author of each accepted paper is required to register to the workshop.
Accepted and presented papers will be published in the ICDCS Workshops proceedings and submitted to IEEE Xplore as well as other Abstracting and Indexing (A&I) databases. IEEE reserves the right to exclude a paper from distribution after the conference, including IEEE Xplore® Digital Library if the paper is not presented by the author at the conference.
Program
23 July 2024
Venue
The joint ICDCS 2024 conference will be held in Jersey City, New Jersey (USA). Please refer to the Conference Venue and Accommodation on the ICDCS 2024 website for any participating information including conference venue and travel.