Forensic Auditing Of Semantic Communication For 6g Networks

Document Type

Article

Source of Publication

Ieee Communications Standards Magazine

Publication Date

1-1-2026

Abstract

Semantic communication is a promising candidate for enabling native AI capabilities in 6G networks. It has the potential to reduce latency and bandwidth consumption in machine-to-machine communication by conveying semantic meaning rather than raw data. However, this paradigm shift also poses significant challenges for post-incident forensic analysis. Operators can no longer rely on traditional packet captures to determine what was actually said, which models were used, the prevailing channel conditions, or whether the receiver had granted consent. To address these challenges, we propose a forensic auditing architecture that provides engineers and researchers with a reliable framework for investigating incidents in semantic communication systems. The proposed design involves lightweight signed witnesses at each encoder, decoder, and channel monitor. These records are organized as per-agent hash chains anchored in a transparent public log for integrity and traceability. We demonstrate its efficacy through a vehicle-to-everything (V2X) case study in which the witness layer reconstructs attack provenance that traditional channel-and signature-based baselines miss. Evaluation results obtained in a controlled software-emulation environment indicate that the witness layer consumes less than 1% of a CPU core at 100 records per second and approximately 300 bytes per event, suggesting that the proposed architecture is sufficiently lightweight for practical deployment in operational environments.

ISSN

2471-2825

Publisher

Institute of Electrical and Electronics Engineers (IEEE)

Disciplines

Computer Engineering

Keywords

Recording, Decoding, Modeling, Architecture, Computer architecture, Signal detection, Vectors, Vehicles, Head, Semantic communication

Indexed in Scopus

no

Open Access

no

Link to Full Text

Share

COinS