TRANSNET at TOP Conference 2025
Now in its fourth year, the TOP Conference focuses on the development and application of optics and photonics in telecommunications and data centers, alongside presentations on photonic components and free-space optics.
The conference will be held on 25-26 February 2025, featuring significant contributions from TRANSNET researchers. Notably, we are excited to announce that Professor Robert Killey is an invited speaker at the conference, presenting "Numerical simulations and experimental assessments of ultra-wideband optical fibre communication systems". We are also proud to see that many friends and alumni of TRANSNET are involved in TOP Conference this year - particularly Mingming Tan and Lidia Galdino as Sub-Committee Topic Chairs, and Wladek Forysiak as a Conference Chair.
The full list of TRANSNET contributions to TOP Conference 2025 is shown below.
Invited Talks
Stream 1: Telecommunications I
Title: Numerical simulations and experimental assessments of ultra-wideband optical fibre communication systems
Speaker: Professor Robert Killey
Abstract: We present recent simulations and experiments assessing the performance of ultra-wideband (UWB) S+C+L band optical fibre transmission systems, with bandwidths exceeding 15 THz. The widely-used Gaussian noise (GN) model of nonlinear interference has been extended to take into account inter-channel stimulated Raman scattering and distributed Raman amplification, and we performed long-haul transmission experiments to confirm the accuracy of the GN model for UWB systems. Using the model to optimise system design, S+C+L band transmission experiments achieving 122.6 Tb/s over an unrepeatered 223 km single-mode fibre (SMF) link with bidirectional Raman amplification, and 202.3 Tb/s transmission over field-deployed SMF, were carried out.
Speaker: Professor Robert Killey
Abstract: We present recent simulations and experiments assessing the performance of ultra-wideband (UWB) S+C+L band optical fibre transmission systems, with bandwidths exceeding 15 THz. The widely-used Gaussian noise (GN) model of nonlinear interference has been extended to take into account inter-channel stimulated Raman scattering and distributed Raman amplification, and we performed long-haul transmission experiments to confirm the accuracy of the GN model for UWB systems. Using the model to optimise system design, S+C+L band transmission experiments achieving 122.6 Tb/s over an unrepeatered 223 km single-mode fibre (SMF) link with bidirectional Raman amplification, and 202.3 Tb/s transmission over field-deployed SMF, were carried out.
Date and Time: Tuesday 25 February 2025, 14:00 – 14:30 (GMT)
Location: Ashes Suite (1st floor)
Poster Sessions
Paper: Graph Intelligence for Benchmarking Optical Networks
Presenter: Akanksha Ahuja, University of Cambridge
Presenter: Akanksha Ahuja, University of Cambridge
Abstract: Topology Bench advances benchmarking in wavelength-routed optical networks. Our work has three key contributions. First, we provide the most extensive open-access dataset, comprising 105 georeferenced real-world core networks and 270,900 synthetic topologies. Second, we integrate structural, spatial, and spectral metrics to capture the graph-based characteristics of optical networks. Third, we provide a systematic framework to select representative sets of topologies based on unsupervised learning to organise real networks into objectively defined clusters. These three contributions provide three key benefits.
Paper: MPI-Penalty-Free S-Band Transmission over G.654.E-Compliant Fibres
Presenter: Romulo Aparecido, UCL
Presenter: Romulo Aparecido, UCL
Abstract: G.654.E-compliant fibres feature ultra-low loss and a large effective area (~125 μm² compared to the standard 80 μm²), reducing nonlinear interactions during propagation. However, their maximum allowed cable cutoff wavelength of 1530 nm raises concerns for S-band transmission (1460–1530 nm) and potential multipath interference (MPI). For the first time, we experimentally demonstrated MPI-penalty-free S-band transmission over G.654.E fibres, including Corning® Vascade® EX2500 and Corning® TXF® fibres, by comparing their performance to a variable optical attenuator (VOA) with equivalent loss. Launch power was optimised for maximum SNR in the linear regime, limiting impairments to transceiver noise, ASE noise, chromatic dispersion (fully compensated digitally), and MPI. If present, MPI would introduce an SNR difference between fibre and VOA transmission. We transmitted 22 channels (1474–1525 nm) over 80 km and 160 km for 16 GBaud and 64 GBaud rates. SNR differences between fibre and VOA were below 0.4 dB, confirming penalty-free S-band operation.
Paper: Performance evaluation of a polarization insensitive Mach-Zehnder fiber parametric amplifier with 38 channel transmission
Presenter: Mariia Bastamova
Presenter: Mariia Bastamova
Abstract: We demonstrate the record reach and capacity for multi-span WDM transmission when using a fiber optical parametric amplifier (FOPA) as an in-line amplifier. The performance enhancement was achieved through the adoption of the Mach-Zehnder (MZ) FOPA architecture, which offers an improved noise figure and reduced nonlinear crosstalk as compared to earlier FOPA designs. Furthermore, we optimized pump phase modulation waveform employed to mitigate stimulated Brillouin scattering (SBS), and thus reduced signal phase noise induced by the pump phase modulation. As a result, we successfully transmitted 30x PM-16QAM channels with 50 GHz spacing at per-channel data rate of 200 Gbit/s over a distance of 540 km or seven spans.
Paper: Novel Optimisation Problems Enabled by Fast Coherent GN Model Network Simulation
Presenter: Michael Doherty
Presenter: Michael Doherty
Abstract: Simulators for optical networks are designed to provide accurate estimates of network performance under varying traffic conditions. This allows for optimized allocation of network resources, e.g. routing, bandwidth, and launch powers. However, existing simulators either provide accurate physical layer models that are slow to compute (e.g. GNPy) or use a basic physical layer model to reduce computation time (e.g. optical-rl-gym). We present our network simulator, XLRON-GN, that provides accurate per channel GSNR estimation through the closed-form ISRS GN model and GPU-based computation, accurate to 0.2dB over C+L bands. It provides sufficiently high throughput (>1000 new lightpath connections per second) to enable new optimization problem settings. We demonstrate per-channel launch power optimization in a dynamic C+L elastic inter-DC network. This is the first time launch power has been optimized for all paths in a network. We apply reinforcement learning to the problem and benchmark other optimisation techniques, demonstrating significant capacity increase.
Paper: Continuous Wave Fiber Optical Parametric Amplifier Tuneable across ~590 nm Range with Gain of up to 52 dB
Presenter: Vladimir Gordienko
Presenter: Vladimir Gordienko
Abstract: We demonstrate a continuous wave (CW) fibre optical parametric amplifier (FOPA) tuneable across the 70 THz wide range between ~1310 nm and ~1900 nm. The FOPA provides a narrow gain band tuneable across the specified wavelength range by adjusting the pump wavelength between 1548 nm and 1563 nm. This allows to amplify a single channel or a CW, perform wavelength conversion or to devise a broadly tuneable CW fibre optical parametric oscillator operating anywhere within this range. In this demonstration, first, we observe the FOPA amplified spontaneous emission with central wavelength adjustable between 1311 nm and 1892 nm. Then, we employ this FOPA to amplify a CW probe adjusted between 1308 nm and 1462 nm with gain in the range of 22-47 dB. The amplified probe is also wavelength converted to the range between 1675 nm and 1897 nm with gain in the range of 21-52 dB.
Paper: Towards minimising quantum operations for low-latency long-distance multipartite entanglement generation via Q-learning
Presenter: Anuj Gore
Abstract: Multipartite entanglement distribution is key for applications such as multi-party QKD or distributed quantum
computing. Previous work has shown that latency in distributing multipartite states in grids is strongly dominated by classical communications. The number of times the control plane instructs the nodes to perform entanglement and swapping operations drives this dominance. In this work, we decrease the latency due to classical communication by training a Q-learning agent to minimise the number of quantum operations required to generate the multipartite state.
Presenter: Anuj Gore
Abstract: Multipartite entanglement distribution is key for applications such as multi-party QKD or distributed quantum
computing. Previous work has shown that latency in distributing multipartite states in grids is strongly dominated by classical communications. The number of times the control plane instructs the nodes to perform entanglement and swapping operations drives this dominance. In this work, we decrease the latency due to classical communication by training a Q-learning agent to minimise the number of quantum operations required to generate the multipartite state.
Paper: Impact of topology on multipartite entanglement distribution protocols in quantum networks
Presenter: Jazz Ee Ze Ooi
Presenter: Jazz Ee Ze Ooi
Abstract: Quantum networks over regional or larger geographic areas will require entanglement distribution via repeaters before exchanging quantum information between users. Previous work on multipartite entanglement distribution protocols has shown that performance is dependent on network topology and repeater placement. However, the key topological properties impacting performance have not been identified yet. In this work, we study the performance of single-path and multi-path multipartite entanglement distribution protocols on 75 real network topologies. Results show that, based on the protocols’ entanglement distribution rate, networks can be partitioned into four distinctive clusters, with specific topological attributes correlated to each cluster. Furthermore, we demonstrate that network topology significantly impacts the repeater requirements of multi-path protocols. For example, to achieve 2x the distribution rate of a single-path protocol, one network cluster allows multi-path protocols to operate with less than half of network repeaters whilst another cluster requires all repeaters to be available.
Paper: High gain U-band discrete Raman amplifiers for U-band coherent transmission
Presenter: Dini Pratiwi
Presenter: Dini Pratiwi
Abstract: We present high gain incoherently pumped U-band discrete Raman amplifiers using three different Raman gain fibres: highly nonlinear fibre (HNLF), inverse dispersion fibre (IDF) and highly nonlinear dispersion shifted fibre (HNLDSF). The highest net gain of 22.3 dB at 1646 nm was achieved by HNLF, with noise figure of 4.2 dB. HNLDSF gave ~4.8 dB higher gain than IDF. These U-band discrete Raman amplifiers were used for C+L+U coherent transmission using 516×24.5GBd/s DP-64/256QAM channels achieving 123.46 Tbps decoded data rate covering ~125 nm bandwidth over 50 km standard single mode fibre (SSMF).
Paper: Enhancing GHZ State Distribution in Noisy Quantum Networks via Multi-Path Routing
Presenter: Evan Sutcliffe
Presenter: Evan Sutcliffe
Abstract: Multipartite entangled states, such as the Greenberger-Horne-Zeilinger (GHZ) states, are multi-qubit entangled states with several envisaged applications such as multi-party Quantum Key Distribution (QKD), distribution quantum sensing and distributed quantum computing. We consider, for the first time, the problem of using multi-path routing protocols to distribute GHZ states with both, higher rate and fidelity than single-path approaches. To do so, we: a) propose novel multi-path protocols aimed to maximise fidelity and b) optimise a memory cutoff time scheme for Bell states stored in non-ideal quantum memories. Monte Carlo simulation results show that multi-path protocols achieved up to a 9.5× higher distribution rate and distributes GHZ states with up to 30% higher fidelity compared to single path routing. These results highlight the benefits of multi-path routing techniques for multipartite state distribution in noisy quantum networks.
Paper: Quantum Annealing for resource allocation in optical networking
Presenter: Victor Virag
Presenter: Victor Virag
Abstract: We explore the application of quantum and hybrid optimization techniques to solve the Wavelength Allocation (WA) and Routing and Wavelength Allocation (RWA) problems in static WDM optical networks. For WA, we benchmark Quantum Annealing (QA), Hybrid Annealing (HA), and Simulated Annealing (SA) against the classical LDF heuristic and the exact solution. For RWA, we introduce and solve a novel QUBO formulation, focusing on minimizing link utilization and verifying solution optimality with brute-force methods. WA results show that the classical LDF heuristic consistently outperforms quantum and hybrid solvers in both computation time and solution quality, with the exact solution struggling on larger instances. For RWA, quantum and quantum-inspired solvers yield optimal solutions but require careful parameter tuning and does not scale well. Our key contributions include the first comparison of QA, HA, and SA for WA on real-world network topologies and the introduction of a novel QUBO formulation for static RWA.