Invited Papers

Abstract: As 5G Radio Access Networks (RAN) are moving closer to commercial exploitation, research efforts are now strongly focusing on Optical/Wireless convergence. Several techniques have been proposed in literature for transporting mobile traffic through optical links to centralized baseband controllers, exploiting either Analog or Digitized  radio signals over the fiber channel. Even though the latter, currently appears as the preferred option from the 5G industry cooperation, the poor scalability of D-RoF bandwidth requirements makes the A-RoF concept a highly appealing fronthaul approach for future Centralized/Cloud Radio Architectures (C-RAN). In this paper, we explore the role of analog Rof solutions for future 5G architectures, highlighting the impact of  digital-signal-processing to proposed network implementations. Moreover, an Analog RoF-based Mobile Fronthaul architecture compatible to 5G requirements is presented and experimentaly evaluated using a centralized DSP-enabled approach.

Abstract: Ethernet has evolved from a protocol for local area Network transport to a more advanced carrier class metro Ethernet transport protocol. Ethernet is evolving continuously as new features are brought into Ethernet. Recently, industrial, automotive and 5G mobile front-haul network applications have been addressed. Several new mechanisms are proposed and standardized, e.g. enabling deterministic latency. This paper reviews and discusses differences between Ethernet and ITU-T G.709 – Optical Transport Network OTN, and analyses if Ethernet may replace OTN in the optical transport network.

Abstract: Emerging mobile and cloud applications drive ever-increasing capacity demands, particularly for short-reach optical communications, where low-cost and low-power solutions are highly required. Spatial division multiplexing (SDM) techniques provide a promising way to scale up the lane count per fiber, while reducing the number of fiber connections and patch cords, and hence simplifying cabling complexity. This talk will address challenges on both system and network levels, and report our recent development on SDM techniques for optical data centers.

Abstract: The Open and Disaggregated Transport Network is an operator-led initiative to build data center interconnects using disaggregated optical equipment, common and open standards, and open source software. We discuss the project objectives, roadmap, and design.

Abstract: Information and communication technologies (ICT) are being deployed at scales at which continued growth becomes problematic across multiple dimensions. Historical architectures for technologies including servers, mobile access networks, and optical systems are being re-examined. Sub-system disaggregation has received much attention as a means of improving system efficiency and scalability. We examine disaggregation as an architectural transformation mechanism generally and consider its potential for realizing hyperscale attributes in optical networks.

Abstract: We present elasticity and agility in softwarized optical network construction, service continuation, and service update. Existing services are kept transient quality against sudden traffic changes and failures. We show proper optical power management in network protection. Programmability is a necessary solution for lower CAPEX and agile network setup. We show beyond 100 Gbps programmability trial in optical edge.

Abstract: Future 5G networks will bring important challenges to network operators, not only in terms of increased volume of traffic load due to the proliferation of mobile broadband communications but also to support appealing features such as network virtualization and network slicing. In this context and aiming at reducing the network investment to be realized for supporting a myriad of heterogeneous services, the convergence among the nowadays separated infrastructures (fixed and mobile) is considered essential. In this work, we address the above objective describing a set of experimentally validated solutions based on SDN to handle fixed and mobile convergence combining and leveraging the benefits of both transport technologies packet and optical.

Abstract: A key challenge in multi-vendor heterogeneous virtual optical networks is providing transparent access to network resources and virtual functions in a manner that enables users to combine them appropriately into meaningful end-to-end services. In this talk, we will present a solution that consists of two components: an Open Marketplace where vendors and users of network resources and functions meet to establish economic relationships; and a suite of orchestration algorithms for creating end-to-end services across various time scales.

Abstract:  In this talk, vertical cavity surface emitting laser (VCSEL) technology is presented as potential prominent performer to address key challenges and novel functionalities of future optical metro networks. The adoption of VCSEL-based modules is particularly attractive for the implementation of programmable (SDN-enabled) transceiver architectures, targeting a radical reduction of cost, power consumption and footprint. Different flavours of these architectures will be presented to enable agile, scalable and high capacity optical metro networks. Furthermore, advanced functionalities and programmability aspects will be analysed to explore the potential of adopting solutions based on this technology.

Abstract: The introduction of optical switching in datacenter interconnection networks is key for solving several of the shortcomings of state-of-the-art electronic switched solutions. Limitations in the port count and reconfiguration speed of optical switches require novel network designs offering network scalability and dynamicity. We review the NEPHELE data center interconnection network which relies on hybrid electro-optical top-of rack (TOR) switches to interconnect servers over multi-wavelength optical rings. The bandwidth of the rings is shared and an efficient utilization of the infrastructure calls for co-ordination in the time, space and wavelength domains. For inter-datacenter communication, a hierarchical orchestration approach is followed to jointly allocate capacity in the respective intra-datacenter networks and the elastic optical wide area network that connects them.

Abstract: Failure detection is of paramount importance in optical networks as a result of the huge amount of traffic that optical connections support, so recovery techniques can be applied. Additionally, the cause of failure needs to be localized so failed resources can be excluded from the computation of the restoration path. In the case of soft failures, their prompt detection, identification, and localization makes that recovery can be applied before excessive errors in optical connections translate into errors on the supported services or even become disrupted. Monitoring and Data Analytics (MDA) become of paramount importance for in the case of soft failures. In this paper, we review a MDA architecture that reduces remarkably detection and identification times, while facilitating failure localization. In addition, we rely on Optical Spectrum Analyzers (OSA) deployed in the optical nodes as monitoring devices acquiring the optical spectrum of outgoing links. Analyzing the optical spectrum of optical connections, specific soft failures that affect the shape of the spectrum can be detected. Three different machine learning –based approaches, designed to be integrated in the aforementioned MDA architecture, will be studied to analyze the optical spectrum of a given optical connection acquired in a node and to determine whether a filter failure is affecting it, and in such case, what is the type of filter failure and its magnitude. Exhaustive results are presented that allow to evaluate the three proposed approaches.

Abstract: In this work, we present a complete rethink of the decision-making process in allocating bandwidth with machine intelligence. We highlight the use of an artificial neural network (ANN) at the central office to learn the uplink latency performance using multiple network features over a heterogeneous fiber-wireless network. In turn, the trained ANN enables the central office to facilitate flexible bandwidth allocations under diverse network scenarios in meeting low-latency communication demands.

Short Bio: Dr Shuangyi Yan is a lecturer in High Performance Networks group at the department of Electrical & Electronic Engineering, University of Bristol. He received the B.E degree in information engineering from Tianjin University, Tianjin, China in 2004. In 2009, he got the PhD degree in optical engineering from Xi’an Institute of Optics and Precision Mechanics, CAS, Xi’an, China. From 2011 to 2013, he worked on the spectra-efficient long-haul transmission system and low-cost short-range transmission system in Photonics Research Centre, Dept. EIE of the Hong Kong Polytechnic University, Hong Kong. In July 2013, he joined the High Performance Networks Group at University of Bristol as a senior research associate. His research interests include multi-dimensional programmable optical networks, multi-layer network analytics for network optimization, and next generation data center networks. He is currently working on different national and European projects (e.g., INSIGHT, TOUCAN, MetroHaul). He is the author or co-author of over 50 publications, of which consist 6 patents and several post deadline papers in optical communication related top-level conferences.