Transformer-Based Prognostics: Enhancing Network Availability by Improved Monitoring of Optical Fiber Amplifiers

Dominic Schneider; Lutz Rapp; Christoph Ament

eess.SPMar 30, 2026

Transformer-Based Prognostics: Enhancing Network Availability by Improved Monitoring of Optical Fiber Amplifiers

Authors: Dominic Schneider, Lutz Rapp, Christoph Ament

Organizations: Advanced Technology, Adtran Networks SE, 98617 Meiningen, Germany · Faculty of Applied Computer Science, University of Augsburg, 86159 Augsburg, Germany

Abstract

We enhance optical network availability and reliability through a lightweight transformer model that predicts optical fiber amplifier lifetime from condition-based monitoring data, enabling real-time, edge-level predictive maintenance and advancing deployable AI for autonomous network operation.

Figures & tables

Fig. 1: Proposed transformer architecture SLAT for predicting the RUL of OFAs, derived from multivariate time series data using sliding time window technique.

Fig. 2: Data acquisition setup, illustrating the collection of multivariate time series data from OFAs, with induced degradation scenarios in the shown EDFA architecture.

Table 3

Fig. 3: RTF trajectories for representative test samples of each degradation scenario: (a) pump laser, (b) power detector, (c) variable optical attenuator, and (d) passive components.

RUL Estimation

Remaining useful life estimation predicts how long machines, batteries, and other assets can operate before failure or end of life.

See all 54 papers

APEX: A Network-Native Time-Series Foundation Model for Forecasting and Anomaly Detection for Wireless Edge Operations
Jun 10, 2026Swadhin Pradhan, Niloo Bahadori, Peiman AminiWireless Communications Wireless Foundation Models
Cisco Systems, USA
HAMON: Passive Optical Sequence Mixing for Long-Horizon Forecasting
Jun 15, 2026Alper YıldırımLong-Term Time Series Forecasting Diffractive Neural Networks
Lightweight Transformer Models for On-Device Fault Detection: A Benchmark Study on Resource-Constrained Deployment
Jun 23, 2026Disha PatelEfficient Transformer Inference Adaptive Inference
Department of Computer Science California State University, Fullerton

Jun 10, 2026J/K move · Enter open · S save

APEX: A Network-Native Time-Series Foundation Model for Forecasting and Anomaly Detection for Wireless Edge Operations

Swadhin Pradhan, Niloo Bahadori, Peiman Amini

Cisco Systems, USA

Wireless Communications Wireless Foundation Models Time-Series Foundation Models Multivariate Time Series Forecasting Time-Series Anomaly Detection Time Series Transformers+1

Open paper AskPDF arXiv

Generic time-series foundation models transfer poorly to wireless network telemetry whose signals are bursty, zero-inflated, and coupled across protocol layers. We present APEX, a network-native, decoder-only transformer for forecasting enterprise AP telemetry, and evaluate it on DHCP degradation as a representative network task. APEX is pre-trained on 10-channel multivariate telemetry from ~4,500 production wireless networks (~100K AP time series, 34 metrics per AP), and is available as APEX-Large (269M, cloud) and APEX-Edge (10.5M, edge). On a 192-step (4-day) DHCP degradation benchmark, APEX-Large reduces MAE by 18% over the strongest foundation-model baseline (Toto) and 38% over SARIMA, with anomaly-detection F1 = 0.93, while APEX-Edge enables sub-second, privacy-preserving inference on AP-class edge hardware. These results suggest network-native pre-training is a practical foundation for proactive wireless operations.

Abstract

RUL Estimation

Explore similar work

APEX: A Network-Native Time-Series Foundation Model for Forecasting and Anomaly Detection for Wireless Edge Operations

HAMON: Passive Optical Sequence Mixing for Long-Horizon Forecasting

Lightweight Transformer Models for On-Device Fault Detection: A Benchmark Study on Resource-Constrained Deployment