Net-Cetric Logo 2021
   
 

 

[Th1-1] Introduction

Bijan JabbariIsocore

back to program ^

 

[Th1-2] IETF Network Slicing (tentative)

Adrian FarrelOld Dog Consulting

Network slicing has emerged as an important concept in the operation of networks that deliver a wide variety of complex services over disparate and shared network resources. With its roots firmly in the 3GPP specification of 5G networks, and with the intention of enabling flexible and performance-sensitive applications, network slicing is being considered as a key operational practice in networks that use IETF technologies.
The IETF has been working to specify a common architecture and framework that is applicable to any network that uses an IETF technology (whether it be IP, Segment Routing, MPLS, or a GMPLS-managed lower layer network). This places the IETF Network Slice as a "transport slice" in the context of the 3GPP's end-to-end slicing architecture, but also allows operators to provide network slice services to customers for their own use.
This talk will look at the history of IETF Network Slicing, present the architecture, and explain the common terms. As the IETF's work progresses (including at this week's interim meeting of the TEAS working group), the presenter will include the latest updates on the northbound "slicing as a service" YANG interface, and introduction to the three different solution models on the table.

back to program ^

 

[Th1-3] SURFnet8 - Evolving from device automation to service orchestration

Peter Boers, SURF

During this presentation we will share our views and experiences regarding automating and orchestrating networks. What are the pitfalls? What are the challenges? But also, what are the unforeseen benefits? And how has our service delivery improved now we have a fully automated network? If you are at the start of automating networks we hope you will find our presentation enlightening.

back to program ^

 

[Th1-4] OpenShift/K8s CNF base disaggregation from CPU centric network computing across core and edge

Hidetsugu Sugiyama, Red Hat Japan

The traditional computer architectures are no longer able to keep pace with ever-growing demand. The future computing infrastructure needs to be dynamic, accelerated and secure in nature, just like the containerized software today is composable and on-demand.
Cloud platform is expanding to the edge of the network while supporting data and AI workloads requires new approaches to hardware. Network workloads such as Open vRAN also require new approaches to hardware in order to replace traditional vendor specific hardware. System architectures are transitioning away from the general purpose computing model. Adding more CPU is not sufficient and offloading with traditional hardware acceleration approaches cannot meet the demand. Some communities such as IOWN Global Forum already started to address disaggregated network and compute functions with heterogeneous and composable computing in Data centric infrastracture.
This session elaborates system architecture innovation in open source and how OpenShift and Linux with heterogeneous composable computing are enabling these exciting network functions on disaggregated infrastructure with cloud autoscaling manner. We will discuss vision, use cases, architectural approach, and future with valid examples.

back to program ^

 

[Th2-1] FABRIC: A Network Construction Kit for the Future Internet

Ilya Baldin, RENCI

The presentation describes FABRIC testbed that enable novel experiments investigating Future Internet architectures, protocols and applications.

back to program ^

 

[Th2-2] Technical Challenges of Distributed Virtual Resources Addressed in BRIDGES

Jerry Sobieski, BRIDGES Project (NSF and GMU)

Virtualization as a broad based cyber-infrastructure architecture promises to deliver great new capabilities to future CI. However, it can also introduce interesting technical issues that must be resolved along the way. This talk will present two examples of such challenges and the BRIDGES approach to resolving them: a) Migration and grooming of globally distributed real-time virtual services, and b) a trusted loop detection and mitigation algorithm in an secure [opaque] multi-domain transitive resource reservation system.

back to program ^

 

[Th2-3] R&D Testbed toward Beyond 5G Network

Hiroaki Harai and Fumihide Kojima , NICT

NICT provides a testbed for R&D purpose. This is suitable to experiments for networking and service applications. The presenters will give a talk about current NICT's testbed overview, and present experiment examples. Then, testbed overview toward beyond 5G is presented, where high-reliability and high-elasticity are featured in mobile, optical and NFV environment.

back to program ^

 

[F1-1] Physical Unclonable Functions use in IPv6 deployments

Ciprian Popoviciu, East Carolina University

Abstract Coming Soon.

back to program ^

 

[F1-2] Towards SINET6: Next-Generation Japanese Academic Backbone-Network

Takashi KurimotoNII

This presentation gives a brief report on a new 400-Gbps academic backbone network, called SINET6, which will be started full-scale operations in April 2022. SINET6 has more than 70 backbone routers and forms a fully-meshed topology by using layer-2 multiplex (L2MUX) networks in order to provide researchers in every Japanese prefecture with 400-Gbps access, minimized-latency. SINET6 also gives a secure mobile network accesses for data processing of IoT applications with inter-cloud environment. Multi-layer network design, mobile access and cloud-oriented functions are reported.

back to program ^

 

[F1-3] Towards a new Network-Integrated Paradigm for Data Intensive Sciences

Harvey Newman, Caltech

The science workflows of the largest data intensive science programs, such as the Large Hadron Collider (LHC), the International Gravitational Wave Observatory Network (IGWN), IceCube, and the Vera Rubin Observatory, as well as programs in earth observation, genomics, the multiscale Brain Initiative and many others, demand reliable and predictable data flows across multi-domain networks, coordinated with resource-intensive computation and storage. All of these programs face the challenge of collaboratively processing data at sites around the world, on an exponentially rising scale that is expected to reach several exabytes per year, thereby outstripping the affordable network capacity on transoceanic and campus links, as well as some continental and regional backbone segments within the next few years.
In response to this challenge, leading research and education networks, the data management teams of the LHC experiments and some other science programs, and advanced R&D projects and platforms including the Global Network Advancement Group (GNA-G) and its working groups, and the Global Research Platform and its regional counterparts, have come together to develop a new dynamic and adaptive paradigm of software driven programmable networks that will meet the needs of the science programs in a matter compatible with the network needs of the at-large academic and research community. Achieving these goals within a resource constrained environment will require a new systemwide level of end-to-end awareness, an in-depth approach to monitoring, decision support tools enabling traffic engineering, workflow scheduling coordinating site and network resources, and system tools and optimization methods building on recent developments in machine learning.
In this talk I will briefly review the challenges, requirements, and paths to the new paradigm, referring to a range of key developments, emerging and persistent testbeds and R&D projects along with their progress and plans, from the perspective of the GNA-G Data Intensive Sciences Working Group.

back to program ^

 
 

© 2001-2021 ISOCORE CORP. ALL RIGHTS RESERVED | Sitewide Privacy Statement | Contact the Webmaster | Contact Info