Shuangyi Yan

Verified

Shuangyi verified their affiliation via an institutional email.

Learn more

Verified

Shuangyi verified their affiliation via an institutional email.

Learn more

• PhD
• Senior Lecturer at University of Bristol

We demonstrated an ANN-based multiple-channel QoT predictor over a 563.4-km field-trial testbed. The proposed ANN-based approach predicted Q-factors of the planning unestablished channels and the existed channels with maximum error of 0.06dB. The predictions provide important prior-information for dynamic reconfigurations of low-margin optical netw...

To support the emerging 5G applications and the 5G bearer networks, optical networks, as the critical infrastructure, are continuously evolving to be more dynamic and automatic. The vision of future autonomous networks with low link margins requires precise estimation/prediction of the quality of transmission (QoT) of optical links. Machine learnin...

This paper proposed and evaluated a programmable ROADM system for MCF-based SDM/WDM networks. The proposed ROADM system employing both bypass connection and Route-and-Select wavelength switching enables adaptable virtual topology in optical networks by dynamically configuring bypass connection cores. The simulation results confirmed this ROADM syst...

In this paper, we propose a deep reinforcement learning (DRL)-based algorithm to generate policies of Baseband Function (BBF) placement and routing. In order to explore the performance of the proposed algorithm in practical systems, the online scenario with the completely random requests is used in the simulation considering C-RAN and NG-RAN archit...

The management of wavelength routed optical mesh networks is complex with many potential light path routes and numerous physical layer impairments to transmission performance. This complexity can be reduced by applying the ideas of abstraction from computer science where different equipment is described in the same basic terms. The noise-to-signal...

In recent years, machine-learning (ML) applications have generated considerable interest and shown great potential in optimizing optical network management, such as quality of transmission estimation, traffic prediction, and resource allocation. However, these applications often require large datasets for training, inference, and updating, while ne...

The disaggregated architecture of advanced Radio Access Networks (RANs) with diverse X-haul latencies, in conjunction with resource-limited multi-access edge computing networks, presents significant challenges in designing a general model in placing baseband and user plane functions to accommodate versatile 5G services. This paper proposes a novel...

This paper proposes and validates a PTP-synchronized 8.4ns optical switching with a 100ns jitter at the switching edges. This approach is adopted and demonstrated for instant network recovery within 2.7ms and scheduled network recovery.

In the ultra-low inter-core crosstalk working zone of terrestrial multi-band and multi-core fiber (MCF) elastic optical networks (EONs), the ICXT in all channels of all cores remains below the ICXT threshold of the highest modulation format level (64QAM) for long-haul distances (10000 km). This paper analyzes the performance of this type of MCF in...

Both multi-band and space division multiplexing (SDM) independently represent cost-effective approaches for next-generation optical backbone networks, particularly as data exchange between core data centers reaches the petabit-per-second scale. This paper focuses on different strategies for implementing band and SDM elastic optical network (BSDM EO...

The industrial landscape is rapidly evolving with the advent of 6G applications, which demand massive connectivity, high computational capacity, and ultra-low latency. These requirements present new challenges, which can no longer be efficiently addressed by conventional strategies. In response, this article underscores the transformative potential...

We propose a DRL-assisted approach for service provisioning in multi-band elastic optical networks. Our simulation environment uses an accurate QoT estimator based on the GN/EGN model. Results show that the proposed approach reduces request blocking by 50% compared with heuristics from the literature.

The escalating demands for network capacities catalyze the adoption of space division multiplexing (SDM) technologies. With continuous advances in multi-core fiber (MCF) fabrication, MCF-based SDM networks are positioned as a viable and promising solution to achieve higher transmission capacities in multi-dimensional optical networks. However, with...

Digital twin for dynamic optical networks is implemented using GNPy, network telemetry and databases. It enhances ML-based End-to-End QoT predictions in field trials by supporting model pre-training and minimizing data requirements through the AI engine.

Private wireless networks have become essential en-ablers for network use cases in enterprises. Emerging enterprise applications push private networks to be more complex in terms of operation and management. However, current private network managers are contending with the challenge of finding a strategy for a network solution that adequately fulfi...

Network function virtualization (NFV) uses virtual network functions (VNFs) and VNF service chains (VNF-SCs) as a flexible and cost-efficient way of providing network services. This new backbone plays a vital role in dealing with the outbreak of traffic patterns generated by diverse services. Elastic optical networks (EONs) can be used for maximum...

In optical networks, failure localization is essential to stable operation and service restoration. Several approaches have been presented to achieve accurate failure localization of nodes and inter-nodes. However, due to increasing traffic and demand for flexibility, the reconfigurable optical add/drop multiplexer (ROADM) is evolving towards a mul...

Cascaded ANN-based link penalty models are developed and demonstrated for QoT predictions over a 986-km field trial testbed, with precision of ± 0.16 dB. Co-training of ANN models allows network-level QoT prediction feasible.

We implement an LSTM-based algorithm to predict and analyse transmission performance and detect anomalies. Cross-validation of the model over two experimental datasets shows high precision of up to 96% for R2 to predict short-term variations.

Intent driven networking holds the promise of simplifying network operations by allowing operators to use declarative, instead of imperative, interfaces. Adoption of this technology for 5G and beyond networks is however still in its infancy, where the required architectures, platforms, interfaces and algorithms are still being discussed. In this wo...

Open Radio Access Network (Open RAN) has gained tremendous attention from industry and academia with decentralized baseband functions across multiple processing units located at different places. However, the ever-expanding scope of RANs, along with fluctuations in resource utilization across different locations and timeframes, necessitates the imp...

Recently open radio access network (Open RAN), which splits baseband functions into multiple process units at different locations has received considerable attentions from both industries and academia with the potential to enable a fully disaggregated RAN with more flexibility in delivering energy-saving and latency-sensitive applications. However,...

As a recent addition to the emerging 2D bismuth oxychalcogenides (Bi2O2X, where X = S, Se, and Te), atomically‐thin bismuth oxytellurium (Bi2O2Te) exhibits unique thermoelectric and photoelectric properties. In this study, high‐quality Bi2O2Te nanosheets (NSs)‐based saturable absorber (SA) with a modulation depth of 8.2% is developed using the liqu...

The emerging mobile services have imposed several new challenges on the radio access network (RAN), which stimulates its evolution from cloud-RAN to next-generation RAN (NG-RAN). In NG-RAN, baseband functions are redefined as the radio unit (RU), distributed unit (DU), and central unit (CU), which are, respectively, connected by optical front/mid/b...

Multi-access edge computing (MEC) has been regarded as one of the essential technologies for mobile networks, by providing computing resources and services close to users, thereby, avoiding extra energy consumption and fitting the low-latency ultra-reliable requirements for emerging 5G applications. Task offloading policy plays a pivotal role in ha...

The adoption of private 5G networks or Non-Public Networks (NPN) by industry verticals is igniting a digital transformation across various sectors and also leading to industry 4.0. This impetus comes from the integration of private wireless networks with 5G capabilities. Currently, a range of innovative applications and use cases are emerging and r...

This white paper on AI and ML as enablers of beyond 5G (B5G) networks is based on contributions from 5G PPP projects that research, implement and validate 5G and B5G network systems. The white paper introduces the main relevant mechanisms in Artificial Intelligence (AI) and Machine Learning (ML), currently investigated and exploited for 5G and B5G...

A SDM/WDM ROADM is proposed with low port-count WSSs. Fiber-core bypassing reduces the number of and port-count of WSSs in the implementation. The design requires less hardware without compromising on network performance with the developed routing core and wavelength assignment algorithm. © 2020 The Author(s) OCIS codes: (060.4251) Networks, assign...

The emerging 5G network will bring a huge amount of network traffic with big variations to optical transport networks. Software-defined optical networks and network function virtualization contribute to the vision for future programmable, disaggregated, and dynamic optical networks. Future optical networks will be more dynamic in network functions...

5G network demands massive infrastructure deployment to meet its requirements. The most cost-effective deployment solution is now a challenge. This paper identifies a cost implementation strategy for 5G by reformulating existing cost models. It analyses three geo-type scenarios and calculates the total cost of ownership (TCO) after estimating the C...

In this paper, artificial neural network (ANN)-based multi-channel Q-factor prediction is investigated with real-time network operation and configuration information over a 563.4-km field-trial testbed. A unified ANN-based regression model is proposed and implemented to predict Q-factors of all the channels simultaneously. A scenario generator is d...

We adopted K-means clustering to efficiently partition the subcarriers to reduce the complexity of PS-QAM on FBMC/OQAM system using KK receiver. The net data rate of 100 Gb/s is achieved after 125 km transmission.

This paper demonstrates the field-trial validation for a novel machine learning-assisted lightpath abstraction strategy in multi-domain optical network scenarios. The proposed abstraction framework shows high accuracy for dynamic optical networks with 0.44 dB estimation error.

We present a novel approach for Quality of Transmission estimation using hybrid modelling and transfer-learning. Our method reduces the training data requirement by 80% while obtaining an MSE of 0.27dB. The approach facilitates a streamlined ML life-cycle for data collection, training and deployment.

5G network demands massive infrastructure deployment to meet its requirements. The most cost-effective deployment solution is now a challenge. This paper identifies a cost implementation strategy for 5G by reformulating existing cost models. It analyses three geo-type scenarios and calculates the total cost of ownership (TCO) after estimating the C...

We compare three independent impairment abstractions of an installed fibre infrastructure. Abstractions agreed to within 1.3dB despite being obtained from different nodes using different terminal equipment. Validation using a DWDM virtual topology was within 1.4dB.

The paper proposes a deep reinforcement learning (DRL) based policy for BBU placement and routing in C-RAN. The simulation results show DRL-based policy reaches the near-optimal performance with a significantly reduced computing time.

In this paper, we reported machine-learning based network dynamic abstraction over a field-trial testbed. The implemented network-scale NCMDB allows the ML-based quality-of-transmission predictor abstract dynamic link parameters for further network planning.

In current dynamic optical networks with cascaded filters and amplifiers, the optical signal-to-noise ratio (OSNR) can vary significantly from channel to channel. Under such uncertainty, OSNR prediction for unestablished channels becomes indispensable but remains a big challenge. For protective network planning purposes such as margin threshold set...

We demonstrated, for the first time, a machine-learning method to assist the coexistence between quantum and classical communication channels. Software-defined networking was used to successfully enable the key generation and transmission over a city and campus network.

A coordinated optical power optimisation strategy over per-fibre-span and ROADMs is proposed to increase the network throughput. The simulation shows the strategy can improve network throughput by 15%. The experiment demonstration verifies 0.7 dB SNR improvement per lightpath.

Virtual data center (VDC) solutions provide an environment that is able to quickly scale up, and where virtual machines and network resources can be quickly added on-demand through self-service procedures. VDC providers must support multiple simultaneous tenants with isolated networks on the same physical substrate. The provider must make efficient...

On the basis of profound understanding of data center (DC) traffic demands and optical switching technologies, we present a hybrid optical network design for future data center network (DCN). Such design integrates optical circuit switching (OCS) and optical packet switching (OPS) schemes via hybrid Top-of-the-Rack (ToR) switches which provide flex...

We demonstrate an intelligent monitoring on-demand switching strategy at network nodes based on Bayesian optimization. It is shown that our proposed method achieves identical monitoring capability as complete system exploration while saving a lot of data.

We successfully demonstrated a novel performance learning method using monitoring and Gaussian process. After 436km dark fiber transmission the model captures most of the test data with 0.7dB mean error and enables robust QoT predictor.

We review recent advancements in QKD networking when applying machine learning techniques to optimize transmission of co-existing quantum and classical channels in metro networks, while we explore potential application in IoT to enhance device autonomy.

e demonstrate a 26 Gbaud real-time quick-reconfigurable 16QAM/QPSK-adaptable trans- mitter. The modulation format can be switched in less than a second by an RMAT agent. The FPGA- driven reconfigurable transmitter can work as a generic edge-node interface for traffic aggregation.

An SDN based network planning framework utilizing machine-learning techniques, network configuration and monitoring database is implemented over an optical field-trial testbed comprised of 436.4-km fibre. Adaption of the spectral efficiency utilising probabilistic-shaping BVT based on link performance prediction is demonstrated.

Optical transport networks with an expanding variety and volume of network data services challenge network providers’ ability to provide high-quality service assurance and network management. Software-defined networks (SDNs) decouple the data plane and control plane and enable network programmability in optical networks with a centralized network c...

In the context of the fifth generation of mobile technology (5G), multiple technologies will converge into a unified end-to-end system. For this purpose, software defined networking (SDN) is proposed, as the control paradigm will integrate all network segments and heterogeneous optical and wireless network technologies together with massive storage...

In this work, we experimentally demonstrate highly flexible and intelligent inter-domain coordinated actions based on adaptive software-defined networking (SDN) orchestration. An advanced multi-domain multi-technology testbed is implemented, which consists of a 400-Gb/s variable-capacity optical packet switching domain and a Tb/s-class flexi-grid w...

We will demonstrate the composition and operation of a virtual infrastructure (VI) for NFV content delivery. The demonstrated VI will be controlled through SDN controller. Furthermore, an infrastructure replanning mechanism will be also demonstrated

We propose a scheme to introduce real-time optical layer monitoring into optical network virtualization enabled by software defined networking. The optical layer factors that impact the virtualization are characterised and investigated using this scheme experimentally.

We experimentally demonstrate a monitoring scheme utilizing both intermediate node and receiver monitoring for network re-planning. Either modulation format switching or light-path re-routing is adopted for restoration. The recovered signal performs better compared with static planning.

For future multi-dimensional optical networks, vast network resources provided by space division multiplexing and wavelength division multiplexing technologies, require new network architectures to scale up current network functions. The huge switch-granularity range requires a more dynamic way to deploy network resources. In this paper, we propose...

This paper describes the problematic filter narrowing effect in the context of next-generation elastic optical networks. First, three possible scenarios are introduced: the transition from an actual fixed-grid to a flexi-grid network, the generic full flexi-grid network, and a proposal for a filterless optical network. Next, we investigate differen...

This paper reports an FPGA-based switch and interface card (SIC) and its application scenario in an all-optical, programmable disaggregated data center network (DCN). Our novel SIC is designed and implemented to replace traditional optical network interface cards, plugged into the server directly, supporting optical packet switching (OPS)/optical c...

Theoperation of multidomain and multivendorEONs can be achieved by interoperable sliceable bandwidth variable transponders (S-BVTs), a GMPLS/BGP-LS-based control plane, and a planning tool. The control plane is extended to include the control of S-BVTs and elastic cross connects, which combine a large port-count fiber-switch (optical backplane) and...

An SDN based monitoring framework is proposed and demonstrated to aggregate multi-layer monitoring information and enable end-to-end network analytics for converged packet and optical networks. New network functions are demonstrated with the proposed framework.

A common Transport API is proposed for joint cloud/network orchestration, allowing interworking of heterogeneous control planes to provide provisioning/recovery of QoS-aware E2E services. We present its first demonstration in a multi-partner testbed including data-plane monitoring. OCIS codes: (060.4256) Networks, network optimization; (060.4253) N...

We introduce multi-technology transport layer monitoring to facilitate coordinated virtualization of optical and packet networks supported by optical virtualize-able transceivers. The scheme is experimentally demonstrated to show a holistic optimization across both layers.

Cognitive SDN orchestration over 400-Gbps OPS and Tbps-class Flexi-WDM networks is demonstrated, where the SDN-controllable OFDM transponders and the extended transport API enable the congestion-aware provisioning of end-to-end real-time services.

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 and a planning tool. This paper reports the first full demonstration and validation this end-to-end architecture.

We present an all-optical flexible disaggregated flat DCN architecture utilizing Hollow-Core Bandgap Fibre, reconfigurable 4×16/8×12 Spectrum Selective Switches and FPGA-based switch and interface intra/inter-blade cards enabling chip-level access while utilizing cut-through low-latency tuneable WDM/WDM-TDM.

A feasible optical virtualize-able transceiver is proposed to support virtual optical networks by creating multiple co-existing virtual transceivers controlled by SDN. This solution is experimentally demonstrated, showing on-demand selection of subcarriers, channel spacing and modulation formats.

We experimentally demonstrate multiple advanced functionalities of a cost-effective high-capacity sliceable-BVT using multicarrier technology. It is programmable, adaptive and reconfigurable by an SDN controller for efficient resource usage, enabling unique granularity, flexibility and grid adaptation, even in conventional fixed-grid networks.

An adaptive optical data center network architecture is, for the first time, proposed and simulated. Data center traffic loads are efficiently accommodated by FTP solution, supported by synthetic optical packet/circuit switching scheme and dynamic function-topology management.