Security of federated learning in 6G era: A review on conceptual techniques and software platforms used for research and analysis

Abstract

Federated Learning (FL) is an emerging Artificial Intelligence (AI) paradigm enabling multiple parties to train a model collaboratively without sharing their data. With the upcoming Sixth Generation (6 G) era, FL is expected to adopt a more prevalent role as a potential solution to overcome the challenges of data privacy, security and scalability in distributed and heterogeneous systems. Presently, research works in the security domain of FL in 6 G communication are widely pursued. However, the outcome of research efforts is critically dependent upon the concepts and platforms used during analysis and evaluation. Therefore, after an overview of FL in 6 G networks, this study highlights the requirements of analysis for the security of FL in distributed and heterogeneously involved multiple entities in 6 G networks. This study comprehensively identifies and reviews the potential Conceptual Techniques and Software Platforms for analysis and evaluation in security-related areas of FL in 6 G communication. Further, this study highlights major challenges faced during the analysis of the security of FL in 6G. Finally, this review deliberates upon the potential open research issues that can be pursued using the identified techniques and platforms.