ProtoRINA is a user-space prototype of the Recursive InterNetwork Architecture (RINA) by Professors Abraham Matta and John Day of Boston University and graduate students Yuefeng Wang, Flavio Esposito (now at Exegy) and Nabeel Akhtar.  It is a new architecture that builds on the fundamental principle that networking is Inter-Process Communication (IPC) and only IPC. As a consequence, RINA subsumes existing Future Internet Architecture (FIA) proposals and inherently supports capabilities such as security, mobility, and manageability. RINA views the network as a collection of networks of processes, rather than a network of “boxes”. A “layer” in RINA constitutes a (virtual) transport network of processes, which can recursively provide communication service to a higher layer (Figure 1). RINA separates mechanisms and policies so all processes use the same mechanisms but may use different policies in different layers over different scopes. RINA has a complete naming/addressing architecture where processes, not interfaces, are named, and addresses are not static but relative to the layer in which the process resides. Each process employs only two policy-configurable protocols: a data transfer control protocol and an object-based management protocol.

Figure 1: Distributed IPC Facility (DIF) is RINA’s building block. A DIF is a special DAF (Distributed Application Facility) that provides communication service.

ProtoRINA provides a framework with common mechanisms so researchers do not have to implement these from scratch; rather they can focus on programming different policies (supported by user applications or network management applications).  ProtoRINA offers several features:

1. It is not restricted to the Internet Protocol (IP), so it enables experimentation with new control and management applications;
2. It supports research not only on user applications but also network management;
3. It can be used as a teaching tool by educators in networking and distributed systems classes; and
4. It can be used to run real experiments, both on local-area networks and on wide-area network testbeds such as GENI.

ProtoRINA (version 1.0) was released in October 2013 and has been used to demonstrate the RINA architecture and its advantages, and also to experiment with different policies.

In a GREE 2013 paper, RINA experiments over the GENI testbed demonstrate enrollment of an IPC process into a RINA layer, and the dynamic formation of a high-level layer on top of lower-level layers.  The latter illustrates how RINA naturally supports the provision of a virtual private cloud service (demonstrated in NSDI 2013). In a GREE 2014 paper, ProtoRINA on GENI is used to experiment with different routing policies configured over different layer topologies. In a CNERT 2014 paper, the programmability of ProtoRINA is demonstrated with the jitter-based routing of video traffic across multiple GENI aggregates.

Figure 2: Yufeng Wang describing the ProtoRINA implementation on GENI at CNERT 2014.

GENI has been instrumental in testing RINA as it supports repeatable wide-area and non-IP experiments. GENI capabilities and tools—such as stitching distant resources using GRE tunnels or VLANs, configuring the experiment using GUI tools, and visualizing real-time measurements using the GENI Desktop—provide the needed experimental support environment, which has markedly improved over the past two years. Future experiments will continue to test RINA’s inherent support for Network Function Virtualization and application-driven network management, demonstrate the programmability of additional recursive-networking policies, and run a long-lived RINA-managed slice on GENI to which users can opt-in to offer or use new services.

A tutorial on running ProtoRINA on GENI was delivered at GEC19; tutorial instructions and a video are available online.   The ProtoRINA code, along with its user / programmer’s manual and other documentation, are available on the project’s wiki page.

Join GENI and US Ignite, March 23-26 in Washington for an exciting look at the future of the Internet and the next-generation application it enables.

Under the leadership of the National Science Foundation, a growing community of researchers, educators, and industry partners have designed, developed, and deployed a working prototype of the next Internet. The Global Environment for Network Innovations (GENI) spans dozens of university campuses across the US, supporting thousands of users with access to deeply programmable cyberinfrastructure that transcends important limitations of the current Internet. The US Ignite initiative brings together corporations, cities, and their advanced application developers and learning communities to bring these new capabilities to bear in key areas of national interest, including education and workforce training, energy, transportation, health, and public safety.

On Wednesday, March 25, we will feature a special event, Beyond Today’s Internet: Experiencing a Smart Future. This event will showcase new applications running on deeply programmable cyberinfrastructure that transcends the current Internet and has the potential to completely transform how we live, work, learn and play.

A networking event with demonstrations and posters will be held the evening of Tuesday March 24 at George Washington University. Fifty teams will demonstrate next generation Internet technology and the exciting research and applications it enables.

At the conference you will learn what you can do with GENI by means of presentations and hands-on tutorials, and experience first-hand demos of next-generation applications run on this new network. Learn from engaging keynotes, tech presentations and workshops by leading researchers and visionaries. Hear from community leaders highlighting recent case studies, success stories and best practices to improve your group’s next-generation technology initiatives.

The conference is hosted by The George Washington University, Washington DC. Please register online via this link:
https://www.regonline.com/22ndgeniengineeringconferencegec22_1649371

For conference agenda, travel grant instructions and practical information:
http://groups.geni.net/geni/wiki/GEC22Agenda.

If you have questions about the event or would like to get your organization involved, please email Henry Yeh.

Understand. Innovate. Transform – Make an Impact on Global Networks.
In 2010, the NSF Directorate for Computer and Information Science and Engineering (CISE) established its Future Internet Architecture project to “stimulate innovative and creative research to explore, design, and evaluate trustworthy future Internet architectures”, and funded four principal projects in the summer of 2010.  At 16th GENI Engineering Conference (GEC 16), Darleen Fisher, Program Director at NSF for the FIA project, summarized the FIA vision, reviewed the four principal FIA projects and outlined how they could use GENI for validation. MobilityFirst is one of the principal FIA projects.

MobilityFirst Project

The MobilityFirst project assumes the future of the Internet is dominated by mobile devices, and that the Internet architecture at that time should fully support mobile devices without any special gateways or proxies, and provide service that is robust even as wireless links are disconnected and then re-connected. MobilityFirst is led by Dipankar “Ray” Raychaudhuri, director of WINLAB at Rutgers University, and is scheduled to run through September 2013.  As PI, Ray leads a large team of researchers, including those from eight other universities plus collaborators at several industrial research organizations.

See Figure 1 and the ACM AINTEC 2011 conference paper (by Ivan Seskar, et. al.) for a summary of the MobilityFirst architecture.  The core network is comprised of multiple domains, with connectionless communication driven by a hybrid combination of name- and address-based routing.  The routers are integrated with storage and computing resources.  A large, scalable Global Name Resolution Service (GNRS) binds the name of a network attached object to its current address.  Names are applied to devices, content and context, enabling context- and location-aware services. Self-certifying public key based identifiers for objects (i.e., their GUIDs) provide for an inherently trustworthy network.  There is edge-aware inter-domain routing, and storage-aware intra-domain routing.  End devices may home to multiple wireless domains, transferring packets wherever there is a connection.  When disconnected from the global network, devices may form ad-hoc wireless networks.

MobilityFirst Proof-of-Concept Prototype Using GENI Resources

An important task in the MobilityFirst project is to validate the architecture with prototypes.  First, key components were prototyped, including a network router, a host protocol stack and support services that embody the routing and name resolution services.

Second, a proof-of-concept prototype of the MobilityFirst architecture was demonstrated at GEC12

in late 2011 by Kiran Nagaraja of WINLAB .  The prototype included a Click-based router with name resolution (GNRS) and storage-aware routing functions (GSTAR), and the Linux/Android implementations of the host protocol stack.  An API provided a socket-like interface to all MobilityFirst services, including unicast, multicast, and anycast message delivery, content querying and retrieval, and multi-homing.

This demonstration utilized a wide variety of GENI resources as shown in Figure 2 to implement the prototype and support a robust content delivery service from one multi-homed  device (GUID_101) with Wi-Fi and WiMAX network access,  to another (GUID_201).  It consisted of seven routers deployed on programmable ProtoGENI hosts spread across the US, with two edge networks (located at BBN, Cambridge, MA and WINLAB, Rutgers, North Brunswick, NJ), each having a WiMAX base station and a WiFi access point for end user mobile device access

Figure 2. MobilityFirst Prototype Using GENI Resources to Provide Robust Content Delivery

The current MobilityFirst prototyping effort is aimed at showing that key MobilityFirst features, such as the GUID naming scheme and storage aware routing, can be supported using Software Defined Networks (SDNs) based on OpenFlow technology. A demo at GEC15 showed that such features could be implemented on an OpenFlow switch by using appropriate control  programs, and that client mobility can be handled seamlessly in an OpenFlow configuration. At GEC16 the demo showed MobilityFirst federated the GENI backbone with other MobilityFirst sites and with Japan.

This work in using GENI to prototype the MobilityFirst FIA is a step towards fulfilling GENI’s vision is to become the “world’s first laboratory environment for exploring future internets at scale, promoting innovations in network science, security, technologies, services and applications.”

Join GENI and US Ignite, March 23-26 in Washington for an exciting look at the future of the Internet and the next-generation application it enables.

Under the leadership of the National Science Foundation, a growing community of researchers, educators, and industry partners have designed, developed, and deployed a working prototype of the next Internet. The Global Environment for Network Innovations (GENI) spans dozens of university campuses across the US, supporting thousands of users with access to deeply programmable cyberinfrastructure that transcends important limitations of the current Internet. The US Ignite initiative brings together corporations, cities, and their advanced application developers and learning communities to bring these new capabilities to bear in key areas of national interest, including education and workforce training, energy, transportation, health, and public safety.

On Wednesday, March 25, we will feature a special event, Beyond Today’s Internet: Experiencing a Smart Future. This event will showcase new applications running on deeply programmable cyberinfrastructure that transcends the current Internet and has the potential to completely transform how we live, work, learn and play.

A networking event with demonstrations and posters will be held the evening of Tuesday March 24 at George Washington University. Fifty teams will demonstrate next generation Internet technology and the exciting research and applications it enables.

At the conference you will learn what you can do with GENI by means of presentations and hands-on tutorials, and experience first-hand demos of next-generation applications run on this new network. Learn from engaging keynotes, tech presentations and workshops by leading researchers and visionaries. Hear from community leaders highlighting recent case studies, success stories and best practices to improve your group’s next-generation technology initiatives.

The conference is hosted by The George Washington University, Washington DC. Please register online via this link:
https://www.regonline.com/22ndgeniengineeringconferencegec22_1649371

For conference agenda, travel grant instructions and practical information:
http://groups.geni.net/geni/wiki/GEC22Agenda.

If you have questions about the event or would like to get your organization involved, please email Henry Yeh.

The 21st GENI Engineering Conference (GEC), hosted by Jon-Paul Herron of the Indiana University Global Network Operations Center, was held from Monday October 20 through Thursday October 23 in Bloomington, IN.

 This GEC completed the transition to a new conference structure based on community feedback at GEC19.   Day 1 of the conference was primarily for newcomers to GENI; Days 2 and 3 formed the core of the conference with sessions of interest to experimenters, campus operators and developers; and Day 4 was primarily for GENI developers. This new format was well received, especially by developers who appreciated having time for discussions on Day 4 without being pulled away but other concurrent sessions.

 Over 80 attendees participated in the Day 1 newcomer tutorials. About half of them used the new jFed graphical tool while the other half used the new Jacks tool. A new session on this day had attendees pick one or more assignments of interest to them and work on them without the aid of step-by-step instructions. The objective was to prepare them to work independently on their own experiments.

Presentations and tutorial instructions for all the conference sessions are available on the GEC agenda page.

Highlights of the general sessions:

• Tutorials by four NSF Future Internet Architecture (FIA) projects: A ChoiceNet tutorial led by Tilman Wolf (U. of Massachusetts) and James Griffieon (U. of Kentucky), MobilityFirst tutorial led by Ivan Seskar and Francesco Bronzino (Rutgers U.), an NDN tutorial led by Alex Afanasyev (UCLA) and Steve DiBenedetto (Colorado State U.), and an XIA tutorial led by Dan Barrett, Srinivasan Seshan, Matt Mukerjee and Yuchen Wu of CMU. All four of these tutorial were well attended.
• A series of tutorials on scripting and scaling up experiments. Divya Bhatt and Mike Zink did a tutorial on LabWiki, a tool for scripting experiments; Thierry Rakotoarivelo and Max Ott did a tutorial on OEDL used for scripting and orchestrating experiments; Nick Bastin did a tutorial on geni-lib, a Python library useful for parsing advertisement and manifest RSpecs and programmatically generating request RSpecs; and Sarah Edwards and Xuan Liu did a tutorial on systematically scaling up experiments.
• Tutorials on wireless experimentation using the Belgian iMinds w-iLab.t testbed with mobile robots by Pieter Becue, Brecht Vermeulen and Thijs Walcarius; using GENI WiMAX resources by Fraida Fund, Ivan Seskar and Abhimanyu Gosain; and installing 3^rd party services on home WiFi routers by Suman Banerjee.
• An experimenter/developer roundtable session where a panel of experimenters from the Future Internet Architecture projects and a couple of other advanced GENI experimenters described their experiences using GENI. A separate panel of GENI developers responded to the experimenter feedback. There was a good discussion about improving the experimenter experience of building WAN topologies and the developers identified improvements they will implement and deploy.
• A GENI Operations session that covered a combination of topics on in-process and future planned operations and integration work.   There was a lively discussion about the dearth of VLANs at some campus and regional endpoints and what can be done about it.

The demonstrations and poster session included 39 demonstrations and posters. As with the last GEC, attendees voted for the best demonstration. The winning demonstrations were:

• Best Demo: Vehicular Sensing and Control by Hongwei Zhang, Yuehua Wang, Jing Hua, Chuan Li, Hai Jin, Yu Chen, Pengfei Ren, Ling Wang, Anthony Hold, Patrick Gossman and Xiaohui Liu of Wayne State University; Jayanthi Rao of Ford Motor Company; George Riley of Georgia Institute of Technology and Weidong Xiang of the University of Michigan-Dearborn.
• First Runner-up: GENI Cinema by Ryan Izard, Kuang-Ching Wang, Joseph Porter, Benton Kribbs and Qing Wang of Clemson University and Aditya Prakash and Parmesh Ramanathan of the University of Wisconsin-Madison.
• Second Runner-up: GENI Over the Air by Shengli Fu and Yixin Gu of the University of North Texas.

The GEC20 plenary session featured five demonstrations by the New York Polytechnic University, Rutgers University, Clemson University, Wayne State University and the University of Wisconsin that used GENI wireless resources and GENI services such as Instrumentation and Measurement services and archival services. These demonstrations, which showcased different application domains, used GENI resources, tools and services. The projects, slices and resources for each demo were reserved using the GENI Portal. The layer 2 connectivity for multi-site demos was reserved over existing AL2S multi-point VLAN(s) that extended to the wireless edge of the network.

Figure 1. Connectivity diagram for Vehicular Sensing and Control.

The GENI-Enabled Vehicular Sensing and Control demonstration by Hongwei Zhang, Jing Hua, Yuehua Wang, Jin Hai and Chaun Li of Wayne State University showed the real-time collection of a vehicle’s internal state and surrounding data while it was driven around the Wayne State campus (Figure 1). As seen in the diagram, data was sent from the vehicle to a GENI Rack for processing over multiple wireless technologies: the GENI WiMAX network when in range of the campus WiMAX base station and the Sprint LTE network when out of range. A private network in GENI was used to carry the data from the GENI WiMAX network to the GENI Rack. The team used the GENI LabWiki GENI tool for monitoring system performance.

Figure 2. In-network video transcoding in MobilityFirst

Francesco Bronzino, Ivan Seskar, Kiran Nagarajan of Rutgers University demonstrated the latest advancements in the MobilityFirst (MF) Future Internet Architecture project. They showed their network protocols support at-scale mobility, trustworthiness and improve data transport to mobile devices. Their topology; depicted at Figure 2, spanned seven GENI sites: U. of Utah, U. of Wisconsin, UIUC, NYSERNET, GPO-BBN, NYU/NYU-Poly and Rutgers-WINLAB. Each site has one or more core network routers running the whole MobilityFirst network stack and are interconnected over the GENI backbone. The core interconnection is set up using a multi-point VLAN provided by Internet2′s advanced layer-2 services (AL2S). At three sites (Rutgers, NYU and Wisconsin) there are also edge networks connected to the MobilityFirst backbone. Each edge network had clients connecting over the local WiMAX network. The edge network also hosted a compute cloudlet running a video caching and transcoder service. This setup was used to run the in-network rate-adaptation service for a DASH video streaming application, with the DASH-enabled content server at the Wisconsin site, and clients at the Rutgers and NYU WiMAX networks. The access link was monitored using a routing layer service at the access router and the rate-adaptation was dynamically invoked if the access bandwidth dropped below the server encoded rate. For the demonstrations, the drops in access bandwidth were emulated by adjusting bandwidth reports, to simulate link quality variation from client mobility. GENI OMF/OML instrumentation and measurement tools were used during the demo to monitor the network.

Figure 3. Setup for the GENI takes Flight demonstration.

Suman Banerjee of the University of Wisconsin and Anthony Kapela of 5Nines Network used two kinds of aerial vehicles—drones and small aircraft— in their GENI takes Flight demonstration to illustrate capabilities of the GENI’s wireless infrastructure. In each case, the goal was to stream live, interactive video and audio from the aerial vehicle to the GEC audience at UC Davis, The video and audio streams were captured using a camera (mounted on the drone and the aircraft) and was streamed over WiMAX base stations, via an InstaGENI VM installed in the GENI rack in Madison, WI. Since the capacity of the WiMAX link can be variable, they installed a media transcoding service in the aerial vehicle. The transcoding service was implemented as part of the compute layer of the MobilityFirst network stack and installed in a Raspberry PI connected directly to the camera system.

The aircraft flew around downtown Madison and streamed two separate views: (1) A videos stream of downtown Madison and (2) an interactive Skype session between personnel in the aircraft and the audience at UC Davis. The entire session, including the audio and video flows worked well over the aerial link connecting the aircrafts and the WiMAX base stations in the Madison area, successfully demonstrating the transcoding service and the value of the GENI WiMAX infrastructure.

The demonstration by Thanasis Korakis and Fraida Fund of New York Polytechnic School of Engineering was titled Dynamic sensor value estimation for minimizing message exchange in wireless sensor networks. They demonstrated a protocol that uses temporal and spatial relationships between nodes in a wireless sensor network to minimize message-passing in the network. In an offline phase, they collected temperature, light, and humidity measurements from all of the sensors in the network over the course of several days, and applied machine learning techniques to estimate correlations in time and space. In the live demo, they used this information to selectively determine which sensors to request values from, and when, to minimze the total number of messages while still having more than 90% monitoring accuracy across the sensor network. The demo used the GENI WiMAX testbed at NYU, compute resources at the Rutger’s GENI rack that were interconnected through the GENI backbone. ,Standard OMF/OML tools were used for instrumentation, archiving of the measurements, and experiment orchestration.

The SciWiNet demonstration by Jim martin and Adam Hodges of Clemson University illustrated how GENI/WiMAX experiments can be extended with coverage from a commercial cellular network. At Clemson, they had a GENI WiMAX handset that had access to SciWiNet (using Sprint’s 3G network). A student at Clemson moved out of WiMAX coverage and into coverage of Sprint’s network.

The device periodically sent location, signal strength and ping RTT data to a server (a GENI VM) on the OpenGENI rack in Clemson. They showed live visualizations: location/access network, ping RTT time series and audience members could point their browser at a publically available website to see a live map. The visualization showed the device seamlessly moving location and changing access networks.

“Essentially, all models are wrong, but some are useful.” – George Box, 1987

Over 250 students at the NYU Polytechnic School of Engineering, the University of Thessaly, the University of Wisconsin – Madison, and Arizona State University know that real networks don’t always obey the models they learn about in the classroom. These students have all run lab experiments on GENI wireless resources as part of classes on computer networks, wireless communications, or mobile computing. GENI wireless resources are deployed at over 14 campuses across the United States. These resources enable researchers to conduct experiments in real-world conditions.

F. Fund

T. Korakis

GENI researchers Thanasis Korakis and Fraida Fund at the NYU Polytechnic School of Engineering have developed a set of prepared experiments on a range of topics, including: wireless link adaptation, IP mobility, TCP congestion control, adaptive video streaming, wireless signal propagation, software defined radio, and QoS of wireless networks. Each experiment is designed to show a side of computer networks that wouldn’t usually be covered in textbooks and classroom lectures, and to highlight the differences between the simplified models taught to students and the behavior of real networks.

Figure 1: This isn’t the TCP you learned about in school! In the TCP lab exercise, students measure the congestion window dynamics of four different congestion control algorithms, highlighting how much TCP congestion control has changed in the last 40 years. Shown above: results of experiment with TCP CUBIC, which is the current default TCP in the Linux kernel.

In their feedback, students consistently refer to three favorite parts of the lab experience: (1) Collecting their own data on a real network, (2) Plotting data using built-in tools, and (3) Running an experiment to answer a question that is relevant to their own lives. Comments include:

• “I liked that we can find ourselves the answer to a question after experimentation.” (Link Adaptation Lab)
• “This lab was very interesting because of how I encounter these problems in real life.” (Adaptive Video Streaming Lab)
• “Getting our own dedicated cellular network was pretty cool.” (Cellular Measurement Lab)
• “I liked that I see the real time measurements.” (Link Adaptation Lab)
• “I enjoyed how straight-forward and example-centered this lab was.” (Adaptive Video Streaming Lab)
• “I liked that we could take our measurements, not have them be given to us.” (Wireless Signal Propagation Lab)

The GENI WiMAX lab materials are offered as a fully hosted service – the personalized lab website, course materials, and testbed infrastructure are all hosted at the NYU Polytechnic School of Engineering. The course materials include assessment materials which are automatically graded by computer – multiple-choice quizzes on the content and automated grading of experiment results – as well as open-ended questions which can be graded by an instructor directly on the lab website (Figure 2).

Figure 2: As an instructor, you’ll have access to the course management system for your site, so you can modify any content, add announcements, and decide which lab exercises your students will run.

If you are an educator interested in using these materials, please contact us here or email witestlab@poly.edu.

Begun in the fall of 2010, MobilityFirst is one of five Future Internet Architecture proposals funded by NSF’s Future Internet Architecture project. Led by WINLAB at Rutgers University, the MobilityFirst project is a multi-university collaborative research effort that proposes mobility and trustworthiness as central design goals for the future Internet. It addresses mobility by cleanly separating names or identifiers of network objects (includes all prominent principals – hosts, services, content, context, etc.) from their addresses, and proposes a globally scalable naming service to resolve from names to addresses in real time to support seamless mobility.  Shortcomings of today’s end-to-end transports in handling wireless challenges (e.g., variable link quality and disconnections) are addressed by the use of a hop-by-hop segmented data transport and the use of routing protocols that leverage both in-network storage and information of traffic conditions at the network edge. Scalable support for delivery services such as anycast, multicast, and multihoming are emphasized to enable efficient future realizations of content- and context-centric applications that use multi-endpoint communication patterns extensively.

Figure 1: GEC-18 deployment of MobilityFirst (MF) prototypes at five rack sites on the GENI wide-area testbed, along with wireless edge deployments at NYU Poly and Rutgers campuses (shown expanded).

The current emphasis within the project is towards at-scale validation of protocols and architecture through both realistic testbed evaluation and real-world field trials. Prototypes of the key components of the architecture including a Click-based software router, a distributed global name resolution service (GNRS), storage- and edge-aware routing protocols (GSTAR and EIR, respectively), and a name-based service API and host stack have been implemented on Linux and Android platforms. These prototypes are readily deployable on the wide-area GENI testbed utilizing layer-2 network connectivity and deep programmability of host and network elements, to enable realistic evaluations of the MobilityFirst features. To date, several experiments and demonstrations have been conducted on the GENI testbed, including the demo at the recently concluded GENI Engineering Conference (GEC-18), where a novel context-based messaging application was built and demonstrated over a long-running deployment of the MobilityFirst prototype network on GENI.

Contextual Messaging Using Name-Based Networking Concepts in MobilityFirst

A contextual messaging application — Drop It — was developed using name-based networking abstractions provided by MobilityFirst, which allows users to drop messages at a location, and to pick up messages left by others at a location. MobilityFirst allows locations (contexts, in general) to be assigned unique names (a GUID – globally unique ID) which help identify them for network operations such as send,recv or get (retrieves content by name). Locations in physical space may be defined (or fenced) by a set of GPS coordinates, for example, and a persistent GUID can be assigned to them by a well-known service.  Next, by maintaining meaningful address mappings for a location GUID in the GNRS, endpoints can send and receive messages to/from this context. For instance, a mapping of location GUID to the set of all phones that dropped messages at that particular location can enable a pure peer-to-peer realization of the contextual messaging service, where the ‘pick-up’ can implemented as an efficient multicast request to each of the phones by using MobilityFirst’s get API. It is also possible to realize alternate approaches to pure p2p, where in-network message caches could enable a more robust operation when phones go offline.  Further information about the MobilityFirst services and the API can be obtained from the project website and the MobiArch’2013 paper: Network Service Abstractions for a Mobility-Centric Future Internet Architecture.

The GEC-18 demonstrations utilized both GENI wide-area deployments over Internet2’s Advanced Layer 2 Service (AL2S), as well as mobile phones handled by volunteers on the demo floor. A total of ten Click-based prototype MF routers (each with a GNRS server instance), five of which functioned as MF edge routers providing WiFi access were used to construct the core-edge topology shown in Figure 1. Five MobilityFirst software routers ran within virtual machines at the five GENI rack sites (U. Utah, U. Wisconsin, BBN, Rutgers, and NYU) and formed the core network. The WiFi/WiMAX wireless edge access was provided by a combination of custom deployed MF edge routers (as WiFi APs) and the GENI campus WiMAX deployments at Rutgers University and NYU Polytechnic. Ten Android phones (some with dual WiFi/WiMAX interfaces), each running the MobilityFirst protocol stack and the “Drop It” application (shown in Figure 2) were carried around by volunteers (except two which were static at Rutgers) who performed message drop and pick-up operations at the several preset locations on the demo floor. Each location was marked with a QR-code tag that encoded the location’s GUID, like the one shown in Figure 2. Note that QR codes were used to identify locations primarily due to the difficulties of using GPS indoors. However, QR codes and RFID tags are also the most effective way when defining fine-grain contexts such as a demo booth, which can be quite cumbersome to define using GPS-coordinates.

The 18th GENI Engineering Conference (GEC), hosted by Prof. Thanasis Korakis of the Polytechnic Institute of New York University, was held from Sunday October 27 through Tuesday October 29 in Brooklyn, NY.

GENI WiMAX resources played a prominent role at GEC18.  Two plenary demonstrations were based on WiMAX:

• MobilityFirst Future Internet Architecture.  Ivan Seskar and Kiran Nagaraja of Rutgers university demonstrated the MobilityFirst system using Android phones.  The phones were connected to both WiMax and Wi-Fi infrastructures and exchanged traffic using non-IP, Mobility First routing.
• WiRover.  Suman Banerjee of the University of Wisconsin led a demonstration of the WiRover system, designed to significantly enhance Internet connectivity to moving vehicles.  Using a local WiRover box that used WiMAX, WiFi and 3G connectivity, Suman did a live video call from NYU Poly to a moving vehicle at downtown Wisconsin.  The call was dynamically and automatically routed over the best available wireless link.

GENI WiMAX resources were also used by a tutorial on experimentation using GENI WiMAX resources.  This tutorial was led by Fraida Fund of the Polytechnic Institute of NYU, Ivan Seskar of Rutgers University and Abhimanyu Gosain of the GENI Project Office.

The WiRover Demonstration: Video call from Madison WI to the GEC Plenary.

GEC18 featured a number of tutorials ranging from introductory tutorials for newcomers to advanced OpenFlow tutorials.  A three-part Getting Started with GENI tutorial that debuted at the last GEC was offered again, as were the popular OpenFlow and WiMAX tutorials.  New at this GEC were tutorials on: (1) ToMaTo, a topology management tool from the German-Lab future Internet testbed project, and (2) Designing and running experiments using the LabWiki tool and the OMF Experiment Description Language.

New GENI Solicitation 4 projects that will be running shakedown experiments, developing GENI-based curriculum, building new experimenter tools or taking on GENI Rack operations and monitoring presented their project plans and explored opportunities for collaboration.  Many of these projects presented posters at the GEC Posters and Demonstrations session.

International efforts at federating GENI with other research testbeds were discussed at the Federation Tools Support session.  Aki Nakao (University of Tokyo) discussed the VNode architecture and the use of the Slice Exchange Point (SEP) to support joining slices from different federation; Brecht Vermeulen (Ghent University, Belgium) reviewed the jFed tool suite developed at iMinds to test federation API’s, stitching computation services and aggregate managers; Serge Fdida and Loic Baron (University Pierre & Marie Curie) discussed experiences in developing MySlice as a portal to the OneLab federation; and Vasilis Maglaris (National Technical University of Athens) reviewed his experiences with the recently completed NOVI project , describing lessons learned in areas of monitoring architectures, semantic resource description, virtual resource brokering and federated virtualization technologies.

The GEC18 Posters and Demonstration session featured over 30 posters and demonstrations of experiments related to GENI and federated international testbeds.  Among those presenting posters were high school students who ran GENI WiMAX based experiments at a camp run by Fraida Fund of the Polytechnic Institute of New York University.

Details of the GEC sessions including presentations and tutorial material from these sessions are available at http://groups.geni.net/geni/wiki/GEC18Agenda.  Video recordings of select sessions are available athttp://www.youtube.com/user/GENIConference.

Workshop on GENI in Education.  Jeanne Albrecht (Williams College), Jai Aikat and Kevin Jeffay (U. of North Carolina, Chapel Hill) organized this workshop in conjunction with GEC18.  Over 25 faculty that teach graduate and undergraduate networking and distributed systems classes participated.  The workshop included a keynote address by James Kurose, instructors sharing their experiences using GENI in their classrooms and breakout group discussions that addressed specific questions raised by participants who had never used GENI in their classes but were considering doing so.  Feedback from the participants was excellent with many expressing an interest in a follow-on workshop.