Experimenting with real application-specific
QoS guarantees in a large-scale RINA demonstrator
March 2018 – Septembre 2018
Fed4FIRE+ Open Call 3
Over the last several years, large research attention has been given to clean-slate network architectures for the Future Internet, capable of efficiently and effectively solving the well-known limitations of the current TCP/IP-based Internet architecture, e.g., in terms of routing scalability, application-specific Quality of Service (QoS) delivery or built-in security. In this context, the Recursive InterNetwork Architecture (RINA) has emerged as a very promising architectural solution to address these challenges. Such is the case, that a substantial number of European research projects have been funded to date to bring RINA closer to its eventual 5G market adoption (FP7 IRATI, FP7 PRISTINE, GEANT3+ Open Call IRINA, H2020 ARCFIRE).
To keep paving the way to this ambitious goal, large-scale experimental validations as enabled by the Fed4FIRE+ Federation of test-beds become of paramount importance. The present proposal for a medium experiment, ERASER, targets a larges-cale experimental evaluation of the real QoS guarantees that RINA can deliver to heterogeneous applications. A RINA test-bed composed of 87 nodes will be considered, emulating a 5G
metro/regional network scenario spanning from the end-user terminal until the virtual machines where applications run in a datacentre. To illustrate the QoS capabilities of RINA, we have chosen high-definition video streaming as our test application, for which the end-user quality of experience will be validated under different load conditions by injecting synthetic traffic in the network reproducing real application traffic.
Our experiments, never performed before nor in the roadmap of the currently running project H2020 ARCFIRE, will also shed light on the most appropriate deployments of the QoS policies in the recursive stack of layers (e.g., only at the bottom in the metro/regional network segment, or only at the top close to applications and end-users, or at all network layers), collecting measurements of the obtained QoS metrics for each deployment case.
Combining optics and SDN in next generation data centre networks
January 2014 – December 2016
COSIGN proposes a new DC architecture empowered by advanced optical technologies and will demonstrate novel solutions capable of sustaining the growing resource and operational demands of next generation DC Networks. COSIGN aims to move away from today’s vendor specific, manually controlled, performance and scale limited DCs towards scalable DC solutions able to support future-proof dynamic, on demand, low-latency, energy efficient and ultra-high bandwidth DC solutions. COSIGN introduces disruptive transformations in the data plane, significant advances to the control plane and major innovations in the DC virtualization and service orchestration:
- In the DC Data Plane, COSIGN will deliver an entirely-optical solution enabling scalable top-of-rack switches, ultra-low latency and high volume DC interconnects with high spatial dimensioning.
- In the DC Control Plane, COSIGN will build upon and extend the Software Defined Networks (SDN) paradigm leveraging capabilities from high-performance optical technologies while developing technology agnostic protocols for software/user defined routing and control.
- For the DC Management and Orchestration, COSIGN will implement a coherent framework for optical network and IT infrastructure abstraction, virtualization and end-to-end service orchestration.
COSIGN brings together a unique combination of skills and expertise able to deliver, for the first time, a coordinated hardware and software architecture, which will guarantee the scale and performance required for future DCs. Results will be demonstrated in challenging industrial setting, leveraging a DC validation platform from Interoute – a leading European service provider.
Programmability in RINA for European Supremacy of Virtualized Networks
January 2014 – October 2016
The Internet as the global communications infrastructure has been successful in shaping the modern world by the way we access and exchange information. The Internet architecture originally designed in the 1960s has been supporting a variety of applications and offering a number of services till now but emerging applications demand better quality, programmability, resilience and protection. Any alterations to the Internet architecture have become restricted to simple incremental updates and plug-ins instead of radical changes by introducing new solutions.
RINA, the Recursive InterNetwork Architecture, is an emerging clean-slate programmable networking approach, centring on Inter-Process Communication (IPC) paradigm, which will support high scalability, multi-homing, built-in security, seamless access to real-time information and operation in dynamic environments. The heart of this networking structure is naturally formed and organised by blocks of containers called Distributed Information Facilities (DIFs) where each block has programmable functions to be attributed to as they required. A DIF is seen as an organizing structure, grouping together application processes that provide IPC services and are configured under the same policies. Virtualization is a fundamental attribute of the architecture itself. Based on the above fundamental aspect, PRISTINE intends to:
- Design, develop and implement the innovative internals of this clean-slate architecture that include the programmable functions for: security of content and application processes, supporting QoS and congestion control in aggregated levels, providing protection and resilience, facilitating more efficient topological routing, and multi-layer management for handling configuration, performance and security.
- Demonstrate the applicability and benefits of this approach and its built-in functions in use-cases driven by the end-users, service providers and equipment vendors in the consortium. This will ensure that the applications and tools we develop will be deployable by providers, and have a greater potential for future exploitation.
The open portal of Measurement Tools and Datasets for experimental research.
Users can download any public tools and dataset. Registered users (registration is free of charge) can upload their own tools and dataset. This initiative is supported by the FP7 EULER project.
Its general functionalities are:
- Easy access to the tools
- RSS feed
- A search engine
- A wiki/FAQ guide to the basics of the portal
- Two forums for getting help or discussing about the portal
Each tool uploaded obtains these functionalities:
- Overview with brief information about its scope
- Contact information of its developers
- Wiki page to describe its details
- Recent activities (e.g. an update, new module, new documentations, etc.)
- Possibility to publish news
- Add user and technical documentations
- Repository of files with subversion support
Sustainable Network Infrastructure Enabling the Future Digital Society
January 2015 – June 2018
The ICT eco-system has been rapidly and dramatically changing in the last several years. Emerging cloud services, mobile and social network technologies are creating new communication patterns, requiring architectural changes to the underlying networks in order to enable scalable growth in traffic volume, while supporting a high level of dynamic connectivity where applications and application components are frequently provisioned, released, and moved around. SUNSET project focuses on overcoming the existing bottlenecks of current architectural solutions to provide a successful support of the future Digital Society, paving the way to new-coming cloud ICT services as well as enhancing a sustainable use of these services by mature economic sectors, improving this way their competitiveness through cloud ICT technologies. More specifically, SUNSET proposes a novel architecture including all network segments (access, metro, core) and data centre network, empowered by advanced optical technologies and Software Defined Networking (SDN), capable of sustaining the growing resource and operational demands of next generation networks.
SUNSET aims to achieve several objectives as:
- Researching and developing modulation techniques and signal processing techniques for a 10x improvement in transmission data rate in metro-access networks using low cost commercial devices.
- Developing SDN-enabled nodes implementing elastic wavelength and space multiplexing as well as to implement SDN controllers able to manage legacy optical equipment.
- Designing and developing SDN-based orchestration frameworks, SDN-enabled monitoring capabilities, resource optimization algorithms, and investigating future SDN technologies.
- In the long term, SUNSET also investigates further reducing power consumption and enhancing performance at larger scales higher factors of 100 to 1000 in medium-long terms, by hybrid photonic and wireless technologies towards Data-Centres-in-a-Box, supported by new nanostructured materials.
Definitely, the SUNSET project brings together a combination of expertise and resources to deliver novel scalable and future-proof network overall infrastructure solutions.
Communications and Broadband Architecture (CBA) research group
January 2014 – December 2016
2014 SGR 1427
The CBA research group is recognised both as a research group of the UPC and as a consolidated research group of the Catalan Government Generalitat de Catalunya (2005SGR-00481 for the period 2005-2008, 2009SGR-1140 for the period 2009-2013, and 2014SGR-1427 for the period 2014-2016).
Low latency and high throughput dynamic network infrastructures for high performance datacentre interconnects
November 2012 – October 2015
The main objective of the LIGHTNESS project is the design, implementation and experimental evaluation of a high-performance network infrastructure for data centres, where innovative photonic switching and transmission solutions are deployed. Harnessing the power of optics will enable data centres to effectively cope with the unprecedented demand growth to be faced in the near future, which will be driven by the increasing popularity of computing and storage server-side applications in the society. Indeed, the deployment of optical transmission systems leveraging Dense Wavelength Division Multiplexing (DWDM) allows the transmission of more than a hundred of wavelength channels operating at 10, 40, 100 Gb/s and beyond. This effectively results in “unlimited” bandwidth capacities of multiple Terabit/s per fibre link, which can be efficiently utilized through next-generation all-optical switching paradigms like Optical Circuit Switching (OCS) or Optical Packet Switching (OPS). In this context, LIGHTNESS will join efforts towards the demonstration of a high-performance all-optical hybrid data plane for data centre networks, combining both OCS and OPS equipment to implement transport services tailored to the specific applications’ throughput and latency requirements. To this goal, an OPS node suitable for intra- data centre connectivity services will be developed and prototyped during the project, together with an enhanced Top of the Rack (TOR) switch seamlessly connecting servers in each rack to the hybrid OCS/OPS inter-cluster network. As an additional achievement of LIGHTNESS, the OCS/OPS inter-cluster network will be empowered with a network control plane able to dynamically provision flexible connectivity services in the hybrid OCS/OPS data centre network. Such a control plane will also be developed and prototyped for integration in the final LIGHTNESS demo throughout the project.
Experimental UpdateLess Evolutive Routing
October 2010 – June 2014
The main objective of the EULER exploratory research project is to investigate new routing paradigms so as to design, develop, and validate experimentally a distributed and dynamic routing scheme suitable for the future Internet and its evolution. The resulting routing scheme(s) is/are intended to address the fundamental limits of current stretch-1 shortest-path routing in terms of routing table scalability but also topology and policy dynamics (perform efficiently under dynamic network conditions). Therefore, this project will investigate trade-offs between routing table size (to enhance scalability), routing scheme stretch (to ensure routing quality) and communication cost (to efficiently and timely react to various failures). The project will develop appropriate tools to evaluate the performance of the proposed routing schemes on large-scale topologies (order of 10k nodes). Prototype of the routing protocols as well as their functional validation and performance benchmarking on the iLAB experimental facility and/or virtual experimental facilities such as PlanetLab/OneLab will allow validating under realistic conditions the overall behaviour of the proposed routing schemes.
Design and optimization of multi-layer green optical networks
January 2011 – December 2014
The DOMINO project aims at designing novel architecture, algorithms and protocols solutions fulfilling the energy efficiency and awareness requirements of future multi-layer green optical networks. DOMINO leverages the capacities of ultra high dynamic multi-layer optical networks to decrease the ICTs carbon footprint, and relies on five innovative concepts: a thorough analysis of the energetic issues in networks, including the benefits of using novel sub-wavelength switching devices and extending the energy-oriented model to multi-domain scenario; novel network planning strategies accounting for energy, cost and performance metrics; specialised dynamic routing algorithms where multiple constraints such as energy consumption, resource utilisation, and signal quality are optimised; energy-oriented operations and protocols in the network control plane to support the designed strategies and algorithms; dedicated techno-economic studies to evaluate the overall impact the novel concepts have on current network infrastructure and provide a possible migration roadmap.
Energy efficiency in large scale distributed systems
23 January 2009 – 4 May 2013
ICT Action IC0804
This COST Action will propose realistic energy-efficient alternate solutions to share IT distributed resources. While much effort is nowadays put into hardware specific solutions to lower energy consumptions, the need for a complementary approach is necessary at the distributed system level, i.e. middleware, network and applications. The Action will characterize the energy consumption and energy efficiencies of distributed applications. Then based on the current hardware adaptation possibilities and innovative algorithms it will propose adaptive and alternative approaches taking into account the energy saving dimension of the problem. The Action will characterize the trade-off between energy savings and functional and non-functional parameters, including the economic dimension.