A5020183635- Optical Network Technologies
- Advanced Optical Network Technologies
- Advanced Photonic Communication Systems
- Semiconductor Lasers and Optical Devices
- Advanced MIMO Systems Optimization
- Wireless Communication Networks Research
- Cooperative Communication and Network Coding
- Advanced Wireless Communication Techniques
- Photonic and Optical Devices
- Software-Defined Networks and 5G
- Anomaly Detection Techniques and Applications
- Network Security and Intrusion Detection
- Power Line Communications and Noise
- Spectroscopy Techniques in Biomedical and Chemical Research
- Advanced Adaptive Filtering Techniques
- Advanced Wireless Network Optimization
- Advanced Fiber Laser Technologies
- Error Correcting Code Techniques
- Neural Networks and Reservoir Computing
- Wireless Communication Security Techniques
- Remote Sensing in Agriculture
- Full-Duplex Wireless Communications
- Advanced Signal Processing Techniques
- Advanced Optical Sensing Technologies
- Retinal Imaging and Analysis
Nokia (France)
2016-2025
Nokia (United States)
2013-2024
Nokia (Finland)
2017-2019
University of Peloponnese
2015
Orange (France)
2007-2009
In this work we detail the strategies adopted in European research project IDEALIST to overcome predicted data plane capacity crunch optical networks. order for core and metropolitan telecommunication systems be able catch up with Internet traffic, which keeps growing exponentially, exploit elastic networks paradigm its astounding characteristics: flexible bandwidth allocation reach tailoring through adaptive line rate, modulation formats, spectral efficiency. We emphasize novelties stemming...
Operators' network management continuously measures health by collecting data from the deployed devices; is used mainly for performance reporting and diagnosing problems after failures, as well human capacity planners to predict future traffic growth. Typically, these tools are generally reactive require significant effort skills operate effectively. As optical networks evolve fulfil highly flexible connectivity dynamicity requirements, supporting ultra-low latency services, they must also...
Optical networks have historically been designed to be operated statically. Connections are overprovisioned so that they remain uninterrupted over several (e.g., 10) years, using high physical-layer margins cover the evolution of physical conditions and modeling uncertainties. As a first step, we can increase efficiency without sacrificing network reliability by removing uncertainties reducing long-term margins, observing adjusting them at intermediate periods. This requires certain...
We introduce an innovative vision transformer approach to identify and precisely locate high-risk events, including fiber cut precursors, in state-of-polarization derived spectrograms. Our method achieves impressive 97% diagnostic accuracy precise temporal localization (6-ms- RMSE).
Worldwide operator deployment of high-speed 100G coherent optical networks is currently underway. To ensure a competitive solution offering significant performance improvements to cope with the ever-increasing traffic demand, novel network concept has been proposed for improved resource utilization based on "elasticity"; specifically, ability make number previously fixed transmission parameters tunable, example data rate or channel spacing. The benefits are numerous, including increased...
We develop an algorithm extension for a coherent receiver, coupled with machine learning to monitor mechanical stress optical fiber, recognizing fiber breaks before they occur. demonstrate event classification 95% accuracy over real-time PDM-QPSK testbed.
Autonomic optical transmission and networking requires machine learning (ML) models to be trained with large datasets. However, the availability of enough real data produce accurate ML is rarely ensured since new equipment techniques are continuously being deployed in network. One option generate from simulations lab experiments, but such could not cover whole features space would translate into inaccuracies models. In this paper, we propose an ML-based algorithm life cycle facilitate...
Considering flexible technologies available nowadays, operating optical networks much closer to their physical capacities is very tempting but necessarily requires efficient network automation. To achieve this, the two main challenges are handling failures, and accurately predicting performance in dynamic environments. We experimentally demonstrate ability of ORCHESTRA solution for early detection localization preventively mitigate impact, thus guarantee smooth operation. Then, leveraging...
We demonstrate the power profile estimation over a deployed 10,000-km submarine link using digital processing at receiver. experimentally show that we estimate span lengths with 0.49km uncertainty and locate multiple losses.
We present an additive noise model for the signal-to-noise ratio prediction in optical systems employing low-resolution (4 bits and below) digital-to-analog analog-to-digital converters. Firstly, expected ASIC power consumption saving by converters rather than their high-resolution version (8 bits) is assessed found to be up 20% a 4-bit physical resolution. Secondly, assess achievable data rate with low-resolution, we model, which relies on two steps. First, mean square error at quantizer...
In the context of future intelligent optical networks, dedicated learning techniques can be employed to monitor physical system parameters with a guaranteed accuracy. this work, we investigate method that establishes link between input parameter uncertainties and overall performance uncertainty. To end, neglecting stochastic effects focusing on simplified Gaussian noise model version, employ uncertainty propagation evaluate from uncertainties, propose simple way margins. With method, as...
We propose a reliable and easy-to-deploy method for localizing failures in optical networks with low added complexity. It essentially relies on the aggregation of raw monitoring data from coherent transponders; readily accessible through streaming telemetry modern networks. first verify reliability our algorithm extensive simulations over European topology under realistic conditions. Then, we apply it to field backbone network carrying live traffic. Here, show how failure localization allows...
Varying the symbol rate is an alternative or complementary approach to varying modulation format channel spacing turn optical networks into elastic networks. We propose allocate just-enough bandwidth for each connection by adjusting such that penalty originating from long cascades of filters contained. This helps reduce overprovisioning lightpaths where full capacity not needed, (i) eliminating unnecessary regenerators and (ii) reducing power consumption terminals, when clock electronics...
Located at the meeting point between telecom operators and over-the-top service providers, metro networks are particularly well suited for introduction of radical acceleration dynamics in optical networks, leveraging elastic building blocks such as transponders nodes. In this paper, we review innovative solutions which could be used to address some challenges short-medium term (e.g., 2-5 years from now). particular, discuss most valuable application scenarios networking. We then highlight a...
We propose a new and cost-efficient method based on receiver-side spectrum monitoring to mitigate filter impairments. experimentally demonstrated up 3dB-reduction in penalties for 32GBd PDM-QPSK signals.
The analysis of a public performance monitoring dataset reveals that reducing margins necessarily amounts to trade availability off for capacity. After carefully establishing this fact, we propose data-driven dynamic rate adaptation as pragmatic solution optimally reclaim in deployed networks. Based on the dataset, show overall capacity monitored network could have been increased by 106% while maintaining over 99.99% 95% connections. Beyond such promising results, our work provides...
We demonstrate a real-time Elastic Interface for future flex-grid networks with software-defined symbol rate transmission. Using PDM-QPSK modulation, live experiments show line adaptation from 10.7 to 107Gbit/s sub-millisecond reconfiguration time.
We experimentally show that noise distribution with regards to filters impacts the QoT by several dB. thus propose a new prediction method achieving up 1.7dB margin reduction for links high hop counts.
We demonstrate optical channel monitoring capabilities executed as SDN applications. To guarantee Quality of Transmission, diagnostic is performed by dynamically selecting the list parameters to be monitored and adjusting their polling rates.
Elastic optical networks generally rely on protection and restoration mechanisms to prevent the detrimental effect of fiber cuts. In particular, we investigate in flexgrid where symbol rate is varied as a complementary approach modulation format. We propose novel scenario, whereby premium traffic survives cuts with just-enough bandwidth, which call "quality-of-service-aware protection" (QoS-aware protection). The proposed mechanism exploits variable baud transponder order reduce spectrum...
The operation of multi-domain and multi-vendor EONs can be achieved by interoperable Sliceable Bandwidth Variable Transponders, a GMPLS/BGP-LS-based control plane planning tool. This paper reports the first full demonstration validation this end-to-end architecture.
The operation of multidomain and multivendor EONs can be achieved by interoperable sliceable bandwidth variable transponders (S-BVTs), a GMPLS/BGP-LS-based control plane, planning tool. plane is extended to include the S-BVTs elastic cross connects, which combine large port-count fiber-switch (optical backplane) bandwidth-variable wavelength-selective switches, enabling end-to-end provisioning recovery network services. A multipartner testbed built demonstrate validate proposed architecture....
Monitoring data from deployed backbone networks shows that many lightpaths present large fluctuations of their quality transmission (QoT), beyond 1 dB in range. In this paper, we investigate the hypothesis evidenced stem anomalies polarization-dependent loss (PDL), equivalent to one PDL element. We eventually propose and experimentally test a closed-loop automation designed stabilize QoT at maximal level for impaired by such anomaly.
Monitoring the state of polarization (SOP) is crucial for tracking vibrations or disturbances in vicinity optical fibers, such as precursors to fiber cuts. While SOP data are valuable machine learning (ML) models identifying vibrations, acquiring a sufficient amount presents significant challenge. To overcome this hurdle, we introduce an innovative transfer framework designed identification (events) when confronted with limited data. Our methodology leverages pre-trained convolutional neural...