In a federated learning paradigm, clients obtain a global model and train it locally using their own data. Subsequently, every client transmits its computed model updates to the central server, and the model updates are combined to produce a fresh version of the global model. In contrast to conventional centralised machine learning, federated learning does not necessitate the collection of all data in one place. Rather, it lessens the requirement for large scale data transfer and permits the training of models on a variety of data sets while protecting data privacy. By automating and improving network operations and management, integrating AI into 6G technology greatly accelerates its implementation. This synergy improves the performance and dependability of 6G networks by allowing predictive analytics to detect and address network problems before they cause service disruptions. Since federated learning can train a single model on data from several sources while ensuring data privacy and security, it offers a wide range of applications, including financial services, smart health, sales, and many more. Federated learning, which makes use of local computing capacity and distributes training data among devices, has an advantage over centralised learning in terms of cost savings and privacy issues.
Additionally, it protects data privacy by using privacy preserving techniques for sharing, transferring, and storing data. Federated learning, then, has the potential to reduce a number of systemic privacy issues and expenses that arise from conventional, centralised machine learning techniques and centralised models. The convergence of telecommunications, in which 6G functions as a network of networks that combines many access technologies to provide global coverage that is omnipresent. the merging of real and virtual environments as the sector heads toward the metaverse. 6G uses Time Sensitive Networking to provide incredibly low latencies and consistent data flow. With the help of this technology, vital and time sensitive applications like robotics, telemedicine, and driverless cars may run smoothly and without interruption in real time. Multiple clients can share the trained model without disclosing their labelled data thanks to supervised federated learning. Adversarial attacks and data poisoning are two security concerns that might skew model predictions. Furthermore, recent research has demonstrated that private raw data restoration is still feasible. Federated learning uses remote data sharing among participants to jointly train a single deep learning model and refine it iteratively, much like in a team report or presentation. Every party downloads the model, which is often a foundation model that has already been trained, from a cloud datacenter.
The supervised learning problem, where data are presumed to be completely labelled, is the primary focus of existing federated learning algorithms. The sixth generation of cellular technology standards for mobile networks are collectively referred to as 6G. 6G will expand on 5G by using higher radio frequencies to deliver faster microsecond speeds, greater bandwidth, and reduced latency. Despite the fact that 5G has just recently been released, the quick advancement of new technology has already surpassed it. Contributions are invited from a range of disciplines and perspectives, including, but not restricted to: Convergence of Federated Learning for Big Data Analytics in High-Performance 6G Networks.
Topics include but are not limited to:
Platform for heterogeneous computing for collaborative content caching towards 6G networks based on federated learning.
Federated learning and edge computing combined for 6G network based omnipresent intelligence.
An enhanced federated learning method for 6G systems powered by cyber twins that preserves privacy.
Air ground 6G networks such as lightweight digital twin and federated learning with distributed incentive.
Internet of cars powered by 6G such as Two layer federated learning with heterogeneous model aggregation.
Future mobile edge devices with dynamic batch sizes enabled time efficient federated learning.
A federated learning approach in 6G networks that uses d2d assistance and an incentive mechanism.
A flexible reward system and reputation system for 6G federated learning that saves energy.
Investigation of Terahertz Wave Distribution and Connectivity Building for Federated Teaching in 6G Wireless Network.
Multi objective optimization using decomposition and Meta DRL for asynchronous federated learning in 6G satellite systems.
A softwarized and federated intrusion detection system for the 6G network enormous internet of things.
A high compression method for communication efficient federated learning in massively parallel networks of interconnected devices.
25th December, 2024: Submission Deadline 05th February, 2025: Authors Notification 10th May, 2025: Revised Version Submission 12th July, 2025: Final Decision Notification
About the Journal
Intelligent Data Analysis provides a forum for the examination of issues related to the research and applications of Artificial Intelligence techniques in data analysis across a variety of disciplines. These techniques include (but are not limited to): all areas of data visualization, data pre-processing (fusion, editing, transformation, filtering, sampling), data engineering, database mining techniques, tools and applications, use of domain knowledge in data analysis, big data applications, evolutionary algorithms, machine learning, neural nets, fuzzy logic, statistical pattern recognition, knowledge filtering, and post-processing. In particular, papers are preferred that discuss the development of new AI-related data analysis architectures, methodologies, and techniques and their applications to various domains.
