Data Centers/High Performance Computing: Technologies, optical networks and systems

Sub-microsecond optical data center networks

Network architectures, topology

Hardware (FPGA/ASIC) based scheduling-control for circuit switched & packet switched networks

  • J. L. Benjamin, T. Gerard, D. Lavery, P. Bayvel and G. Zervas, "PULSE: Optical Circuit Switched Data Center Architecture Operating at Nanosecond Timescales," in Journal of Lightwave Technology, vol. 38, no. 18, pp. 4906-4921, 15 Sept.15, 2020, doi: 10.1109/JLT.2020.2997664.
  • P. Andreades and G. Zervas, "Parallel Modular Scheduler Design for Clos Switches in Optical Data Center Networks," in Journal of Lightwave Technology, vol. 38, no. 13, pp. 3506-3518, 1 July1, 2020, doi: 10.1109/JLT.2019.2963160.

Disaggregated and heterogeneous data centers

Memory disaggregation; Remote memory transaction protocols; Optical network and data center architecture; Compute and network resource allocation; FPGA/MPSoC systems; Remote memory transaction protocols


  • Georgios Zervas, Hui Yuan, Arsalan Saljoghei, Qianqiao Chen, and Vaibhawa Mishra, "Optically Disaggregated Data Centers With Minimal Remote Memory Latency: Technologies, Architectures, and Resource Allocation [Invited]," J. Opt. Commun. Netw. 10, A270-A285 (2018)
  • V. Mishra, J. L. Benjamin and G. Zervas, "Demonstrating Optically Interconnected Remote Serial and Parallel Memory in Disaggregated Data Centers," 2020 Optical Fiber Communications Conference and Exhibition (OFC), San Diego, CA, USA, 2020, pp. 1-3.
  • M. Bielski et al., "dReDBox: Materializing a full-stack rack-scale system prototype of a next-generation disaggregated datacenter" 2018 Design, Automation & Test in Europe Conference & Exhibition (DATE), Dresden, 2018, pp. 1093-1098, doi: 10.23919/DATE.2018.8342174.

Data center SDM networking

Multi-dimensional (space-wavelength) resource allocation

  • Hui Yuan, Marija Furdek, Ajmal Muhammad, Arsalan Saljoghei, Lena Wosinska, and Georgios Zervas, "Space-Division Multiplexing in Data Center Networks: On Multi-Core Fiber Solutions and Crosstalk-Suppressed Resource Allocation," J. Opt. Commun. Netw. 10, 272-288 (2018)
  • H. Yuan, A. Saljoghei, A. Peters and G. Zervas, "Comparison of SDM-WDM based Data Center Networks with equal/unequal core pitch Multi-Core Fibers," 2018 Optical Fiber Communications Conference and Exposition (OFC), San Diego, CA, 2018, pp. 1-3.
  • Y. Liu, H. Yuan, A. Peters and G. Zervas, "Comparison of SDM and WDM on Direct and Indirect Optical Data Center Networks," ECOC 2016; 42nd European Conference on Optical Communication, Dusseldorf, Germany, 2016, pp. 1-3.

Optical sub-nanosecond switching and AI/ML optimisation

Testbeds and experiments on ultrafast space and wavelength switching/tuning.


  • Related papers:C. W. F. Parsonson, Z. Shabka, W. K. Chlupka, B. Goh and G. Zervas, "Optimal Control of SOAs with Artificial Intelligence for Sub-Nanosecond Optical Switching," in Journal of Lightwave Technology, doi: 10.1109/JLT.2020.3004645.
    • C. Parsonson, Z. Shabka, K. Chlupka, B. Goh, and G. Zervas, “https://github.com/cwfparsonson/soa driving,” May 2020. (AI tool and code for SOA optimization)
  • Thomas Gerard, Christopher Parsonson, Zacharaya Shabka, Polina Bayvel, Domaniç Lavery, Georgios Zervas, "SWIFT: Scalable Ultra-Wideband Sub-Nanosecond Wavelength Switching for Data Centre Networks", arXiv preprint arXiv:2003.05489

Space division multiplexing: From technologies to networks

SDM multi-core fiber switching

Multi-core fibre switch design and experimentation


A. Saljoghei et al., "MCF-SMF Hybrid Low-Latency Circuit-Switched Optical Network for Disaggregated Data Centers," in Journal of Lightwave Technology, vol. 37, no. 16, pp. 4017-4029, 15 Aug.15, 2019, doi: 10.1109/JLT.2019.2920354.
C. Deakin, M. Enrico, N. Parsons and G. Zervas, "Design and Analysis of Beam Steering Multicore Fiber Optical Switches," in Journal of Lightwave Technology, vol. 37, no. 9, pp. 1954-1963, 1 May1, 2019, doi: 10.1109/JLT.2019.2896318.

Multi-core fibers

AI-based weakly coupled MCF design, MCF characterization and crosstalk modelling


  • H. Yuan et al., "Experimental Investigation of Static and Dynamic Crosstalk in Trench-Assisted Multi-Core Fiber," 2019 Optical Fiber Communications Conference and Exhibition (OFC), San Diego, CA, USA, 2019, pp. 1-3.
  • Alessandro Ottino, Hui Yuan, Yunnuo Xu, Eric Sillekens, Georgios Zervas, "Random Walk for modelling Multi Core Fiber cross-talk and step distribution characterisation", arXiv preprint arXiv:2008.13511, 2020
  • Hui Yuan, Alessandro Ottino, Yunnuo Xu, Arsalan Saljoghei, Tetsuya Hayashi, Tetsuya Nakanishi, Eric Sillekens, Lidia Galdino, Polina Bayvel, Zhixin Liu, Georgios Zervas, "Experimental Analysis on Variations and Accuracy of Crosstalk in Trench-Assisted Multi-core Fibers", arXiv preprint arXiv:2008.08034, 2020

Multi-core fiber systems and networks

AI based MCF transceiver optimization

MCF based networks

  • A. Ottino et al., "Genetic Algorithm Optimization of Multi Core Fibre Transmission Links based on Silicon Photonic Transceivers," 2019 Optical Fiber Communications Conference and Exhibition (OFC), San Diego, CA, USA, 2019, pp. 1-3.
  • Hui Yuan, Marija Furdek, Ajmal Muhammad, Arsalan Saljoghei, Lena Wosinska, and Georgios Zervas, "Space-Division Multiplexing in Data Center Networks: On Multi-Core Fiber Solutions and Crosstalk-Suppressed Resource Allocation," J. Opt. Commun. Netw. 10, 272-288 (2018)


Space division multiplexed networks

Routing, Spectrum, Core, and Modulation allocation allocation

Open/White-Box/Disaggregated versus Black-Box/Integrated Optical Node/Networking


  • Ajmal Muhammad, Georgios Zervas, and Robert Forchheimer, "Resource Allocation for Space-Division Multiplexing: Optical White Box Versus Optical Black Box Networking," J. Lightwave Technol. 33, 4928-4941 (2015)

Reconfigurable computing: hardware design (FPGA/MPSoCs) for networking and computing

Hardware-based network scheduling


  • J. L. Benjamin, T. Gerard, D. Lavery, P. Bayvel and G. Zervas, "PULSE: Optical Circuit Switched Data Center Architecture Operating at Nanosecond Timescales," in Journal of Lightwave Technology, vol. 38, no. 18, pp. 4906-4921, 15 Sept.15, 2020, doi: 10.1109/JLT.2020.2997664.
  • P. Andreades and G. Zervas, "Parallel Modular Scheduler Design for Clos Switches in Optical Data Center Networks," in Journal of Lightwave Technology, vol. 38, no. 13, pp. 3506-3518, 1 July1, 2020, doi: 10.1109/JLT.2019.2963160.

Runtime/hitless reconfigurable computing

Hardware programmable network function service chain


  • Qianqiao Chen, Vaibhawa Mishra, Jose Nunez-Yanez, Georgios Zervas, "Reconfigurable Network Stream Processing on Virtualized FPGA Resources", International Journal of Reconfigurable Computing, 2018, https://doi.org/10.1155/2018/8785903
  • Q. Chen, V. Mishra, J. Nunez-Yanez and G. Zervas, "Synchronizing reconfiguration of coherent functions on disaggregated FPGA resources," 2017 International Conference on ReConFigurable Computing and FPGAs (ReConFig), Cancun, 2017, pp. 1-6, doi: 10.1109/RECONFIG.2017.8279822.
  • Q. Chen, V. Mishra, N. Parsons and G. Zervas, "Hardware programmable network function service chain on optical rack-scale data centers," 2017 Optical Fiber Communications Conference and Exhibition (OFC), Los Angeles, CA, 2017, pp. 1-3.
  • PQ. Chen, V. Mishra and G. Zervas, "Reconfigurable computing for network function virtualization: A protocol independent switch," 2016 International Conference on ReConFigurable Computing and FPGAs (ReConFig), Cancun, 2016, pp. 1-6, doi: 10.1109/ReConFig.2016.7857183.
  • V. Mishra, Q. Chen and G. Zervas, "REoN: A protocol for reliable software-defined FPGA partial reconfiguration over network," 2016 International Conference on ReConFigurable Computing and FPGAs (ReConFig), Cancun, 2016, pp. 1-7, doi: 10.1109/ReConFig.2016.7857184.

Tools

... more information upon request

Nanosecond speed optical circuit switched data center network simulation

Matlab based simulator (verified with SystemVerilog) that explores optical circuit switching based to hardware scheduler for time-slot and wavelength allocation.

Cycle accurate emulator for optical packet switch

SystemVerilog based emulator for investigating data (buffering, switching) architectures such as Clos and Crossbar and critically hardware based packet scheduling and control plane functionality.

Artificial intelligence for multi core fibre design

A tool has been developed to design homogeneous and heterogeneous weakly coupled multi-core fibers.

Trafpy: Generator of network flow and computational job demands

Traffic generator to stress test networks and computing systems (resource/job scheduling algorithms, topologies, protocols, switches). Trafpy generates either flow distributions or a computation graph, job distributions, whose dependencies (edges) may form flows. Both flow-centric and job-centric network demand generation and management are supported by TrafPy.

AI based optimization of semiconductor optical amplifiers

A tool and source code is available that used genetic algorithm, particle swarm optimization, and ant colony optimization to generate multi-level drive signals to calibrate/optimize the on-off (switching) function (settling time) of any semiconductor optical amplifier.

https://github.com/cwfparsonson/soa_driving

FPGA/RFSoC for real-time signal and data processing