Leader Selection and Follower Association for UE-Centric Distributed Learning in Future Wireless Networks

This paper focuses on UE-centric DL algorithms where UEs initiate requests to adapt AI/ML models for better performance in dynamic environment, e.g., locally refined AI/ML models among a set of headsets or smartphones. This new setup requires selecting a set of UEs as aggregators (here called leaders) and another set as followers, where all UEs update their models based on their local data, and followers share theirs with leaders for aggregation. From a networking perspective, the first question is how to select leaders and associate followers efficiently. This results in a high dimensional mixed integer programming problem and involves internal UE state information and state information among UEs, called external state information in this paper. To address this challenge, we introduce two new indices: a Leader Internal Index (LII), which is a function of the internal states of each device, demonstrating the willingness to be a leader such as battery life and AI/hardware accelerators, and a Leader eXternal Index (LXI), which is a function of external state information among UEs, such as trust, channel condition, and any aspect relevant for associating a follower with a leader. These two indices transform the highly complex leader selection and follower association problem into a better tractable formulation. More importantly, LIIs and LXIs allow to keep the internal and external state information of this problem inside of each device without compromising users’ privacy. We present the necessary constraints and objective function, propose a distributed algorithm, and explore the feasibility, challenges, and implementation within 5G and 6G networks. Our simulation results show that the performance of the distributed algorithm is only 8.9% lower compared to the optimal solution if the related threshold is adjusted efficiently. Also, the message-passing overhead of our approach is linear with the number of users while it effectively translates an NP-hard problem into a more scalable P-hard problem without significant loss in performance. All these elements demonstrate the scalability of our distributed proposed approach for a large number of UEs while introducing a new level of privacy, i.e., no need to share LXI with the network.