top of page

Optical Network Tomography: The Orange–Nokia Trial Explained

Oct 30
3 min read

On September 10, 2026, Orange France and Nokia unveiled a real-world trial of tomography applied to an optical transport network. The solution leverages algorithms developed by Nokia Bell Labs, utilizing measurements already available within existing transport equipment. The core concept is to reconstruct the state of network links more effectively, without the need to install intrusive probes everywhere. This trial represents a significant technological achievement at the intersection of networking, data processing, and automation.



Why Visibility in Optical Networks Is Often Incomplete


An optical transport network carries multiple wavelength channels and traverses a wide array of equipment, including fibers, amplifiers, and switching elements. When performance degrades, the root cause could lie anywhere across various segments—whether due to a physical fault, an amplifier drifting out of spec, a misconfiguration, or a combination of factors. Traditional methods rely on measurements taken at specific points, supplemented by targeted investigations. However, physically instrumenting every single link with additional probes is not always practical. The challenge, therefore, lies in extracting more information from existing equipment by identifying the relationships between observed data and the actual state of individual network segments.


Network Tomography as an Inference Problem


Much like image reconstruction, network tomography does not measure every local property directly. Instead, it combines various indicators to estimate what is occurring within the infrastructure. According to Nokia, its algorithms process thousands of data points from optical network equipment to provide near real-time visibility—link by link—across the network. The value of such an approach depends on the quality and synchronization of measurements, the modeling of optical components, and the ability to distinguish between various anomalies that might produce similar signatures. This is a problem of observability and inference, involving signal analysis and—depending on the implementation—potentially statistical or machine-learning techniques. The press release does not provide full details of the models; therefore, one should not assume the use of any specific algorithm not explicitly described therein.


What results were announced?


According to Orange and Nokia, the test demonstrated the detection of fiber degradation, amplification issues, and configuration errors that could impair service. This enhanced visibility can enable interventions before contractual thresholds are breached and allow for a more precise assessment of the usable capacity of network links. Both companies also highlight the potential for improved spectral efficiency and the benefit of reducing the need for on-site interventions by leveraging existing infrastructure. These findings stem from a real-world trial; however, the public statement does not provide detailed quantitative performance metrics—such as false-alarm rates, sensitivity to various types of faults, or average gains in availability.


From diagnostics to automated networks


True network automation requires a sequence of observation, diagnosis, forecasting, and—in some cases—controlled corrective action. Tomography primarily contributes to the first part of this chain: generating a more reliable representation of the network's state. The subsequent step involves determining when an alert is sufficiently robust to warrant reconfiguration, and verifying that the resulting action does not exacerbate the situation. Key challenges include the explainability of the diagnosis, the management of topology changes, integration with operational tools, and cybersecurity. In a national-scale network, widespread adoption will depend on compatibility with various generations of equipment and the quality of available data.


Technology Monitoring and Positioning


This experiment highlights a research area that extends beyond telecommunications alone: ​​the reconstruction of hidden states based on partial observations. To keep pace with this field, it is necessary to map out telemetry solutions, optical performance models, diagnostic algorithms, and key players in photonic transport. A combined analysis of patents, publications, and network trials makes it possible to identify robust approaches and distinguish between a localized proof-of-concept and a solution capable of large-scale deployment. GraphMyTech can facilitate this monitoring for R&D and strategy teams involved in infrastructure transformation.


 
 
 

Comments


Commenting on this post isn't available anymore. Contact the site owner for more info.

22 rue de Bagnolet,

75020 Paris, France

  • LinkedIn

Privacy & Cookies Policy

bottom of page