End-to-End Optical Semantic Communication over a Nonlinear WDM Fiber Link

Hussein Jammal; Andrea Bianco; Cristina Rottondi

eess.SPSep 29, 2026

End-to-End Optical Semantic Communication over a Nonlinear WDM Fiber Link

Authors: Hussein Jammal, Andrea Bianco, Cristina Rottondi

Organizations: Department of Electronics and Telecommunications, Politecnico di Torino, Turin, Italy

Abstract

Emerging optical-network applications increasingly use received data for inference and control rather than exact source reproduction, creating an opportunity to trade bit-level fidelity for greater transmission reach and efficiency. We propose an end-to-end optical semantic communication system for joint image classification and reconstruction over a nonlinear wavelength-division multiplexed (WDM) fiber channel. The system maps each image directly into a fixed-length sequence of channel symbols that preserves task-relevant information, without explicit source compression or channel coding. Experiments on the MNIST dataset cover launch powers from -9 to +3 dBm, fiber lengths up to 800 km, and 16-, 64-, and 256-Quadrature Amplitude Modulation (QAM) formats. At 0 dBm, classification accuracy remains between 98.92% and 99.31% across all tested link lengths and modulation orders, while requiring fewer transmitted symbols than a Low-Density Parity-Check (LDPC)-coded JPEG baseline at every tested modulation order. These results show that semantic communication can simultaneously extend optical reach and reduce transmission resources by conveying only task-relevant information.

Figures & tables

Fig. 1 : Architecture and optimization flow of the proposed system.

Panel of Figure 2 — Fig. 2 : Classification accuracy (top) and SSIM (bottom) versus launch power at $M=16$ . Color indicates link length. The proposed system is shown on the left and the JPEG baselines on the right, where shading marks the performance range of the proposed system.

Symbol	Quantity	Value
$R_{\mathrm{sym}}$	Symbol rate	32 GBaud
$S_{\mathrm{ps}}$	Samples per symbol	16
$N_{\mathrm{RRC}}$	RRC filter length	1024 taps
$\beta$	RRC roll-off factor	0.01
$N_{\mathrm{ch}}$	Number of WDM channels	11
$\Delta f$	WDM channel spacing	37.5 GHz

TABLE I : Optical waveform and fiber parameters.

Panel of Figure 4 — Fig. 3 : Classification accuracy (a) and SSIM (b) versus link length at nominal launch power of $0$ dBm. Color indicates QAM order $M\in\{16,64,256\}$ , while marker and line style indicate the system variant.

Scheme	Symbols/image	Transmission time [ns]
Semantic system	128/128/128	4.0/4.0/4.0
Uncoded JPEG	233/155/117	7.3/4.9/3.7
LDPC-coded JPEG	466/311/233	14.6/9.7/7.3

TABLE II : Transmission cost per image at 16/64/256-QAM.

Explore similar work

When Semantic Communication Meets Queueing: Cross-Layer Latency and Task Fidelity Optimization
May 6, 2026Yalin E. Sagduyu, Tugba ErpekSemantic Communication Age of Information
Nexcepta, Gaithersburg, MD, USA
End-to-End Quantum Semantic Communication with Variational Quantum Neural Networks
Aug 30, 2026Melek Krichen, Nikhitha Nunavath, Riccardo Bassoli +2Quantum Machine Learning
Polytechnique Montréal Montréal, Canada · TU Dresden Dresden, Germany
On the Rate-Distortion-Complexity Tradeoff for Semantic Communication
Feb 16, 2026Jingxuan Chai, Yong Xiao, Guangming ShiSemantic Communication Rate-Distortion Theory
the School of Artificial Intelligence at the Xidian University, Shaanxi 710071, China · the School of Electronic Information and Communications at the Huazhong University of Science and Technology, Wuhan 430074, China · the Pengcheng Laboratory, Shenzhen, Guangdong 518055, China +2

May 6, 2026J/K move · Enter open · S save

When Semantic Communication Meets Queueing: Cross-Layer Latency and Task Fidelity Optimization

Yalin E. Sagduyu, Tugba Erpek

Nexcepta, Gaithersburg, MD, USA

Semantic Communication Age of Information Rate-Distortion Optimization Wireless Resource Allocation

Open paper AskPDF arXiv

Semantic communication (SemCom) with learned encoder-decoder architectures enables end-to-end learning of compact task-oriented representations optimized for the wireless channel, reducing channel resources needed to convey task-relevant information and improving spectrum efficiency. This paper studies semantic image transmission over block Rayleigh fading with AWGN using a multi-task semantic autoencoder that jointly reconstructs images and predicts labels from the received waveform. The latent dimension (complex channel uses per source sample) serves as a cross-layer control variable governing semantic fidelity and channel resource usage. We characterize the resulting latency-task fidelity tradeoff: larger latent representations improve inference accuracy but increase service time, channel uses, and queueing delay. Building on this insight, we develop online semantic-rate controllers that adapt the latent dimension per update under a long-term semantic error constraint. A queue-aware drift-plus-penalty policy minimizes delay subject to an average semantic error cap, while a complementary age-aware policy minimizes time-average Age of Information (AoI). By adapting the semantic rate to congestion and fidelity requirements, the proposed framework improves spectrum utilization and enables timely semantic updates with significantly lower delay and AoI than fixed-rate baselines.

Abstract

Explore similar work

When Semantic Communication Meets Queueing: Cross-Layer Latency and Task Fidelity Optimization

End-to-End Quantum Semantic Communication with Variational Quantum Neural Networks

On the Rate-Distortion-Complexity Tradeoff for Semantic Communication