With the widespread deployment of Internet of Things, a wide range of modern data-intensive applications are emerging to enrich our daily life. Empowered by the rise of advanced machine learning (ML) techniques, intelligent decisions based on sensing data significantly improve the application performance and give rise to a large number of innovative applications and intelligent services, which will become a crucial component for future communications and networking, e.g., 5G, 6G, and beyond. It is expected that the ML technologies will be adopted in many fields, and the ubiquitous artificial intelligence (AI) will empower promising future communications and networking. However, user data are usually stored in the end devices, which leads to higher-level requirements in aspects of resources, security, and privacy protection when employing ML techniques. In particular, most of the traditional learning techniques for wireless networks perform centralized data aggregation and inference operations on a single data center, which are suffering from critical security challenges, e.g., single point of failure. To this end, distributed ML solutions as an important direction for learning techniques enable the wireless devices to collaboratively build a shared learning model with training their collected data locally. These solutions have high potential to enable ubiquitous, efficient, and secure AI for future communications and networking.
The goal of this workshop is to bring together researchers and practitioners interested in distributed machine learning to understand the topic, identify technical challenges, and discuss potential solutions. Topics of interest include but are not limited to the following:
- Novel concept, theory, principles, and algorithms on the convergence of distributed ML and future communication or network techniques
- Intelligent radio resource management for future communications and networking
- Distributed ML for intelligent signal processing, e.g., signal detection
- Over-the-air access for future communications and networking
- Energy efficiency of implementing distributed ML over future wireless communications and networking
- Ultra-low latency distributed ML for latency-sensitive future communications and networking
- Data analytics driven wireless communications with distributed ML
- Multi-agent reinforcement learning for intelligent network control and optimization
- Network architectures, and communication protocols for distributed ML
- Privacy and security issues of distributed ML for future communications and networking, e.g., physical layer security.
- Distributed ML for mobile user behaviour analysis and inference
- Distributed ML for emerging applications, e.g., vehicle to everything (V2X), UAV-enabled communication, Internet of Things, intelligent reflecting surface (IRS), Massive MIMO, virtual reality (VR), and augmented reality (AR)
- Wireless network optimization for improving performance of distributed ML
- Communication-efficient distributed ML for future communications and networking
- Emerging theories and techniques such as age of information, blockchain, and edge computing for distributed ML
The workshop will feature 2 keynote speeches given by world leading researchers in the field. The workshop accepts only original and previously unpublished papers. All submissions must be formatted in standard IEEE camera-ready format (double-column, 10pt font). The maximum number of printed pages is six including figures without incurring additional page charges.
Prof. Osvaldo Simeone (IEEE Fellow), King's College London, UK
Prof. Chau Yuen (IEEE Fellow), Singapore University of Technology and Design, Singapore
- Deadline for Workshop Paper Submission: 31 December 2021
- Acceptance/Rejection Announcement: 31 January 2022
- Final Workshop Papers Due: 25 February 2022
EDAS submission link: https://edas.info/newPaper.php?c=28610&track=109165