Semantic Communication in Dynamic Channel Scenarios: Collaborative Optimization of Dual-Pipeline Joint Source-Channel Coding and Personalized Federated Learning

2025-03-18

AI Summary

To address the poor generalization and difficulty in achieving personalization in semantic communication—caused by combinatorial explosion of channel state information (CSI) and data in multi-user dynamic channel environments—this paper proposes a CSI-aware dual-pipeline joint source-channel coding (JSCC) architecture. The architecture integrates CSI-driven adaptive semantic encoding/decoding with personalized federated learning and introduces a zero-optimization-gap method for solving non-convex loss functions, enabling concurrent global robustness and local personalization optimization. Experiments across diverse SNR distributions and benchmark datasets demonstrate that the proposed approach reduces average semantic distortion by 23.7% compared to baseline methods, accelerates convergence by 41%, and significantly improves both semantic fidelity and transmission efficiency.

Technology Category

• Search and Optimization: Learning to Search
• Constraint Satisfaction and Optimization: Distributed CSP/Optimization
• Natural Language Processing: Learning & Optimization for NLP

Application Category

• User Modeling, Personalization and Recommendation: Federated recommendation systems and personalization
• Semantics and Knowledge: Methods to enhance, augment, integrate or synergize semantic models such as knowledge graphs and LLMs
• Search and Retrieval-Augmented AI: Personalized, context-aware and across-device search

Abstract

Semantic communication is designed to tackle issues like bandwidth constraints and high latency in communication systems. However, in complex network topologies with multiple users, the enormous combinations of client data and channel state information (CSI) pose significant challenges for existing semantic communication architectures. To improve the generalization ability of semantic communication models in complex scenarios while meeting the personalized needs of each user in their local environments, we propose a novel personalized federated learning framework with dual-pipeline joint source-channel coding based on channel awareness model (PFL-DPJSCCA). Within this framework, we present a method that achieves zero optimization gap for non-convex loss functions. Experiments conducted under varying SNR distributions validate the outstanding performance of our framework across diverse datasets.

Problem

• Addresses bandwidth constraints and high latency in communication systems.
• Solves challenges from client data and channel state combinations.
• Improves generalization and personalization in semantic communication models.

Innovation

• Dual-pipeline joint source-channel coding
• Personalized federated learning framework
• Zero optimization gap for non-convex loss