6G-PFIP: Privacy-Preserving and Personalized Federated Intelligent Paging for 6G Wireless Networks

Authors:
Publisher website
Google Scholar
Abstract
Paging is anticipated to play a pivotal role in realizing the 6G vision for the IoE (Internet of Everything) and uRLLC (ultra-reliable low-latency communications) by enhancing service quality and improving resource efficiency. Nevertheless, as 6G architectures evolve toward large-scale, distributed architectures, existing rule-driven, centralized paging mechanisms face severe challenges, including inefficiencies due to massive mobile-device access, incompatibility with distributed network architectures, and significant privacy concerns. To address these challenges, this paper proposes a privacy-preserving and personalized federated intelligent paging mechanism for 6G Networks (6G-PFIP). 6G-PFIP introduces a novel intelligent decision model to enhance paging efficiency for large-scale user equipment (UE). A personalized hierarchical federated paging architecture (HFPA) is further constructed to improve architectural compatibility and privacy by decomposing the centralized paging process into distributed layers. HFPA effectively addresses the challenge of Non-IID user data in 6G by designing region-level and local-level personalization and fairness-aware aggregation strategies. Furthermore, HFPA integrates local differential privacy (LDP) on demand to perturb shared model parameters, thereby mitigating the risk of privacy leakage. Experiments on a 6G Telecom Dataset show that 6G-PFIP achieves significant signaling resource savings (71.33%) and delay reductions (34.78%), while rigorous analysis confirms its privacy guarantees.
Funding Information
  • National Science and Technology Major Project of China (2025ZD1303100)

This publication has 44 references indexed in Scilit: