Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Interference Detection in Spectrum-Blind Multi-User Optical Spectrum as a Service

Agastya Raj, Daniel C. Kilper, Marco Ruffini

TL;DR

The paper addresses interference detection and attribution in spectrum-blind OSaaS deployments, where operators rely only on coarse power measurements and end-to-end channel performance. It introduces a two-branch 1D-CNN with sinusoidal positional encodings to leverage the spectral and network-position structure of OSaaS data, enabling attribution of interference to the responsible user. Validated on a 190 km Open Ireland testbed with three OSaaS users, the approach achieves 90.3% accuracy in identifying the interfering source, with per-user F1-scores as high as 87–94% and robustness against three interference types, including rogue OOK channels. The method supports proactive mitigation, improves QoS, and offers a scalable path toward production deployment with potential transfer learning for varying user counts.

Abstract

With the growing demand for high-bandwidth, low-latency applications, Optical Spectrum as a Service (OSaaS) is of interest for flexible bandwidth allocation within Elastic Optical Networks (EONs) and Open Line Systems (OLS). While OSaaS facilitates transparent connectivity and resource sharing among users, it raises concerns over potential network vulnerabilities due to shared fiber access and inter-channel interference, such as fiber non-linearity and amplifier based crosstalk. These challenges are exacerbated in multi-user environments, complicating the identification and localization of service interferences. To reduce system disruptions and system repair costs, it is beneficial to detect and identify such interferences timely. Addressing these challenges, this paper introduces a Machine Learning (ML) based architecture for network operators to detect and attribute interferences to specific OSaaS users while blind to the users' internal spectrum details. Our methodology leverages available coarse power measurements and operator channel performance data, bypassing the need for internal user information of wide-band shared spectra. Experimental studies conducted on a 190 km optical line system in the Open Ireland testbed, with three OSaaS users demonstrate the model's capability to accurately classify the source of interferences, achieving a classification accuracy of 90.3%.

Interference Detection in Spectrum-Blind Multi-User Optical Spectrum as a Service

TL;DR

The paper addresses interference detection and attribution in spectrum-blind OSaaS deployments, where operators rely only on coarse power measurements and end-to-end channel performance. It introduces a two-branch 1D-CNN with sinusoidal positional encodings to leverage the spectral and network-position structure of OSaaS data, enabling attribution of interference to the responsible user. Validated on a 190 km Open Ireland testbed with three OSaaS users, the approach achieves 90.3% accuracy in identifying the interfering source, with per-user F1-scores as high as 87–94% and robustness against three interference types, including rogue OOK channels. The method supports proactive mitigation, improves QoS, and offers a scalable path toward production deployment with potential transfer learning for varying user counts.

Abstract

With the growing demand for high-bandwidth, low-latency applications, Optical Spectrum as a Service (OSaaS) is of interest for flexible bandwidth allocation within Elastic Optical Networks (EONs) and Open Line Systems (OLS). While OSaaS facilitates transparent connectivity and resource sharing among users, it raises concerns over potential network vulnerabilities due to shared fiber access and inter-channel interference, such as fiber non-linearity and amplifier based crosstalk. These challenges are exacerbated in multi-user environments, complicating the identification and localization of service interferences. To reduce system disruptions and system repair costs, it is beneficial to detect and identify such interferences timely. Addressing these challenges, this paper introduces a Machine Learning (ML) based architecture for network operators to detect and attribute interferences to specific OSaaS users while blind to the users' internal spectrum details. Our methodology leverages available coarse power measurements and operator channel performance data, bypassing the need for internal user information of wide-band shared spectra. Experimental studies conducted on a 190 km optical line system in the Open Ireland testbed, with three OSaaS users demonstrate the model's capability to accurately classify the source of interferences, achieving a classification accuracy of 90.3%.

Paper Structure

This paper contains 13 sections, 5 equations, 10 figures, 3 tables.

Table of Contents

  1. Introduction
  2. OSaaS Experimental Setup and Configuration
  3. Open Ireland Testbed
  4. OSaaS Experimental Setup
  5. OSaaS Service Specifications
  6. Interferences introduced by OSaaS users
  7. Increase in power across the OSaaS window
  8. ADD/DROP interference
  9. Interference due to OOK Channels
  10. Data Collection
  11. Model Architecture
  12. Results and Discussion
  13. Conclusion

Figures (10)

  • Figure 1: Optical Domain of the Open Ireland Testbed, Dublin, Ireland. Additional fiber spools are connected through the Polatis optical switch.
  • Figure 2: Experimental setup implemented in Open Ireland testbed.
  • Figure 3: OSaaS spectrum configuration at steady state for three users.
  • Figure 4: Different types of interference in the OSaaS model.
  • Figure 5: Interference Type 1: Total power increase across the OSaaS User spectrum
  • ...and 5 more figures