Semantic Communication Meets Heterogeneous Network: Emerging Trends, Opportunities, and Challenges
Guhan Zheng, Qiang Ni, Aryan Kaushik, Lixia Yang, Yushi Wang, Charilaos Zarakovitis
TL;DR
To address spectral efficiency and robustness in future networks, the paper surveys semantic communication (SemCom) in heterogeneous environments. It identifies semantic drift and network-wide codec updating as core challenges and analyzes how system, data, and model heterogeneity, plus personalized many-to-one requirements, complicate collaborative updates. It surveys main updating methods—federated learning, split learning, transfer learning, and hybrids—and proposes a heterogeneity-aware updating scheme that anchors all representations to a shared SKB while updating only a lightweight global adapter. It demonstrates, via a case study, that the approach reduces overhead and improves convergence, and discusses open topics including discrimination, fairness, privacy, dynamics, and incentives for practical deployment.
Abstract
Recent developments in machine learning (ML) techniques enable users to extract, transmit, and reproduce information semantics via ML-based semantic communication (SemCom). This significantly increases network spectral efficiency and transmission robustness. In the network, the semantic encoders and decoders among various users, based on ML, however, require collaborative updating according to new transmission tasks. The various heterogeneous characteristics of most networks in turn introduce emerging but unique challenges for semantic codec updating that are different from other general ML model updating. In this article, we first overview the key components of the SemCom system. We then discuss the unique challenges associated with semantic codec updates in heterogeneous networks. Accordingly, we point out a potential framework and discuss the pros and cons thereof. Finally, several future research directions are also discussed.