This talk will present our 3 year journey in creating 5G/B5G innovation testbeds in the Bristol region. I will discuss our technical approach for delivering end-to-end, multi-technology, multi-vendor private networks and our engagement with vertical sectors including digital creative, museums, public safety, manufacturing and logistics. I will reflect on our approach towards solutions co-creation and the impact generated both in terms of technical innovation but also knowledge transfer and digital transformation within these sectors. Last, I will project on future trends and challenges.

The need for cheaper and more precise localisation techniques has recently amplified. The initial approach has been to roll out high-level software running on smartphones and leveraging Bluetooth proximity sensing. However this approach lacks both precision in terms of ranging, and flexibility in terms of experimental framework to fully explore alternative schemes for contact event tracing. In this context, we thus provide open-access nodes in an open-access experimental platform for ranging and proximity tracking, letting researchers tinker freely with the full software stack on a swarm of multi-radio, low-power devices based on cheap microcontrollers. We provide a tutorial on how to use the platform and open source code building blocks to to program the devices, bare-metal. We then report on initial measurements we have performed using the platform. Perspectives with our platform include applicability studies and comparative evaluation for a large variety of localisation schemes combining the use of Ultra-Wide Band and Bluetooth Low-Energy for better precision and smaller energy budgets - and the use of complementary mechanism guaranteeing privacy protection, able to run directly on-board cheap IoT microcontrollers.

Current trends on 5G network programmability evidence the need for end-to-end flexibility from the node and edge all the way to the cloud. Such multi-domain scenarios require realistic testbeds where different task-offloading algorithms, scheduling functions, and service orchestration techniques can be deployed and tested. While many of these research components can be often explored locally in small and isolated testbeds, new 5G demands are requesting for inter-operable platforms with a wider and a more global scope. The goal for these global platforms is that they can cope with multi-tier hierarchical architectures that are capable to face intense computational processes and heavy network traffic loads, while preserving dependability and keeping a low latency on the task executions and data transmission.

In this paper we demonstrate a world-wide attempt to integrate different high-performance testing facilities, located in USA, Belgium, and The Netherlands, to enable experimentation on top of such large and complex architectures. In order to do this, we describe and deploy a multi-domain use case that can benefit from a global hierarchical infrastructure. Finally, we detail the performance characteristics of the deployment, discussing the experiences and technical challenges, and presenting the lessons learned we obtained when building and testing such experimental use case.

The need to find more efficient ways to share and use wireless spectrum has resulted in renewed interest in radio frequency (RF) propagation modeling. The open and programmable nature of the POWDER (Platform for Open Wireless Data-driven Experimental Research) mobile and wireless platform offers a unique environment in which to test and validate RF propagation modeling approaches. In this paper we present our work illustrating how POWDER based RF measurements can be performed, as a form of "ground truth", and compared with predicted RF signal strength based on a propagation model. We make use of the Shout RF measurement framework available on POWDER to perform a series of RF measurements. We compare these measurements with predicted power levels using the open source RF Signal Propagation, Loss, And Terrain (SPLAT!) analysis tool. We present our results and a brief terrain analysis to provide real-world context for it. Our work is "packaged" as a POWDER profile to allow others to repeat our analysis and to serve as a starting point for further RF measurement and propagation related research.

With the evolution of Internet protocols and communication networks, time-critical applications (e.g., smart grid, remote surgery) emerge with requirements on deterministic Quality of Service (QoS). Plenty of works have been done on the deterministic networking that provides transmission services with bounded delay and jitters. Most current works evaluate their designs and algorithms with software-based simulations. However, it is highly desired that innovative ideas can be evaluated on real systems or testbeds. The practical implementation and experiments in the real world shall yield insights that lead to further theoretical improvements and vice versa. Thus, this paper presents an implementation of the large-scale deterministic network to validate the availability of deterministic techniques. A large-scale deterministic network is constructed based on Deterministic IP (DIP)-enabled routers (i.e., a prototype for deterministic communication in the core network). Experiments are designed based on two network test facilities, including China Environment for Network Innovations (CENI) and Yangtze River Delta Comprehensive Test Environment (YZNET). We achieve the world's first delay and jitter control over a transmission distance of more than 2,000 kilometers. A maximum jitter beneath 30 µs is realized, which is independent of congestion levels and communication distances. Moreover, to demonstrate the potential of deterministic networking, a demo instance on PLC (Programmable Logic Controller) remote control is also presented. The experimental results imply that the deterministic networking is a promising technique for future networks to support time-critical applications on large-scale networks.

It has become a common requirement for testbeds to provide a service in charge of collecting and exposing metrics, thus assisting experimenters with the central task of data collection. This paper describes Kwollect, a service designed and developed in the context of the Grid'5000, to collect infrastructure metrics (including high-frequency wattmeters) and expose them to experimenters. Kwollect scales to high frequencies of metrics collection for hundreds of nodes. It can also be leveraged by the experimenter to collect custom metrics.

Computer science testbeds often require extensive experiment automation to be used efficiently. Jupyter notebooks can contain the scientific reasoning, experiment orchestration, and experiment results in an easy to read, easy to execute format. However some level of support is required from the testbeds to facilitate notebook use on their platforms. Additionally this format can be used for more than driving experiments, and different uses have different requirements in terms of support.

In this paper we present an analysis of different notebook uses to be expected on computer science testbeds, as found through the feedback of users of the Grid'5000 testbed and a survey of other similar testbeds. Then we present the technical implementation of Jupyter on the Grid'5000 to support those uses.

The Chameleon project developed a unique experimental testbed by adapting a mainstream cloud implementation to the needs of systems research community and thereby demonstrated that clouds can be configured to serve as a platform for this type research. More recently, the CloudBank project embarked on a mission of providing a conduit to commercial clouds for the systems research community that eliminates much of the complexity and some of the cost of using them for research. This creates an opportunity to explore running systems experiments in a combined setting, spanning both research and commercial clouds. In this paper, we present an extension to Chameleon for constructing controlled experiments across its resources and commercial clouds accessible via CloudBank, present a case study of an experiment running across such combined resources, and discuss the impact of using a combined research platform.

Future wireless networks must be able to support Quality of Service (QoS) requirements of emerging 5G and other next-generation applications. REACT is a distributed resource allocation protocol can be used to negotiate airtime among nodes in a wireless network. In this paper, REACT is extended to support QoS airtime. Experimentation on the w-iLab.t wireless testbed in an ad hoc setting shows that these extensions allow REACT to converge on allocations where any node requesting the higher class of airtime receives its allocation, while nodes requesting the lower class are allocated remaining airtime fairly.

Real world testbeds, like the POWDER platform (Platform for Open Wireless Data-driven Experimental Research), enable a broad range of mobile and wireless research. Given the flexibility of this platform, a key concern for platform users is selecting a set of wireless resources that will satisfy the requirements of their experiments. In this paper we present the design and implementation of WiMatch a wireless resource matchmaking system. We illustrate the utility of our approach by evaluating it in the POWDER platform.

Software-defined Radio Access Networks (SD-RANs) enable unparalleled flexibility and the opportunity to customize and/or optimize network operations. In particular, network function virtualization suggests that network functions are no longer tightly coupled to any specific hardware instance or location. In this paper we present our work on BoTM, basestation-on-the-move, where we exploit network function virtualization in an SD-RAN environment to dynamically "move" a mobile base station from one location to another to realize a general RAN management primitive. Specifically, we show how an SD-RAN environment enable orchestration across both the virtual infrastructure layer (i.e., the base station) as well as the mobile network functional layer (i.e., the protocol interactions in a mobile network) enable a base station and its associated endpoints to be moved to enable network management functionality. We present the design and implementation of this network management primitive using Open Air Interface and the FlexRAN framework, with experimental results and efficiency metrics.

This talk does not have an abstract.

EdgeNet is a software-only distributed systems testbed in the family of PlanetLab. It is realized as a multi-tenant global Kubernetes cluster running on volunteered virtual machines. In this paper, we describe our modifications to Kubernetes which enable trusted execution with accountability on untrusted, volunteered hardware, user control of the placement of Kubernetes pods, and control of a global cluster from a single central head node.

Experimentation in a large-scale testbed environment requires a large amount of traffic to ensure realistic scenario execution and better experiment results. Emulated traffic from real users (i.e., clients) can be generated from a large number of human activities (e.g., web browsing, email, or video streaming). Deploying and producing a single activity from an individual user is simple, but emulating and automating it from multiple users with a wide-range of activities is very challenging. We designed a containerized human agent (i.e., bot) to generate a single human activity. Thus, a large number of bots can be deployed and controlled by a single orchestration system. Due to the complexity and wide-range usage of container orchestration systems, we need to develop a simpler system that leverages widely-used open-source container orchestrator. So, researchers and scientists can easily use it to define and execute activity requirements with a few command lines or single specification files. This demo shows the implementation of several unique bots and their deployment in a production testbed environment. Several use cases of activity are also presented to verify the functionality of our system to support large-scale experimentation.

The Tor anonymity network is an essential protection for those living in oppressive regimes, who may be persecuted for their speech or other activities on the Internet. However, its usability is severely affected by high latency. Furthermore, because of the difficulty associated with conducting research evaluations on the Tor network, there has been insufficient research aimed at this problem. To address this, we develop a CloudLab profile that instantiates a private Tor testing network for easy, reproducible experimental evaluations on Tor. We demonstrate how to use this CloudLab profile to measure the page load time of selected web pages over a Tor network under different network conditions representative of typical Tor usage.

We showcase the capabilities of V2Verifier, a new open-source software-defined radio (SDR) testbed for vehicle-to-vehicle (V2V) communications security, to expose the strengths and vulnerabilities of current V2V security systems based on the IEEE 1609.2 standard. V2Verifier supports both major V2V technologies and facilitates a broad range of experimentation with upper- and lower-layer attacks using a combination of SDRs and commercial V2V on-board units (OBUs). We demonstrate two separate attacks (jamming and replay) against Dedicated Short Range Communication (DSRC) and Cellular Vehicle-to-Everything (C-V2X) technologies, experimentally quantifying the threat posed by these types of attacks. We also use V2Verifier's open-source implementation to show how the 1609.2 standard can effectively mitigate certain types of attacks (e.g., message replay), facilitating further research into the security of V2V.

Software-defined Radio Access Networks (SD-RANs) enable unparalleled flexibility and the opportunity to customize and/or optimize network operations. In particular, network function virtualization suggests that network functions are no longer tightly coupled to any specific hardware instance or location. In this paper we present our work on BoTM, basestation-on-the-move, where we exploit network function virtualization in an SD-RAN environment to dynamically "move" a mobile base station from one location to another to realize a general RAN management primitive. Specifically, we show how an SD-RAN environment enable orchestration across both the virtual infrastructure layer (i.e., the base station) as well as the mobile network functional layer (i.e., the protocol interactions in a mobile network) enable a base station and its associated endpoints to be moved to enable network management functionality. We present the design and implementation of this network management primitive using Open Air Interface and the FlexRAN framework, with experimental results and efficiency metrics.

Precise time synchronization amongst network devices is a basic requirement for time critical applications. Despite that time synchronization is a well-established functionality in the wired network domain, its wireless counterpart is very basic or even non-existing. This is because commercial off-the-shelf (COTS) wireless chipsets still focus on basic network connectivity for consumer applications, hence no dedicated software/hardware features are required for time-bounded services in professional wireless environments. This work leverages on openwifi - an opensource Wi-Fi chip design based on Software-Defined Radio (SDR), by adding the support of hardware timestamping in openwifi to support precise time protocol (PTP) application. In addition, openwifi Access Point (AP) is connected to nodes in wired network through a TSN capable switch, and synchronization offset between devices in wired and wireless network is measured. Next, the measurement is done with the same setup in wired network but replacing openwifi by COTS Wi-Fi devices. We observe the synchronization offset with COTS is 10 4 fold larger than the offset achieved with openwifi. The experiment setup is in w-iLab.t, an open testbed infrastructure freely accessible for the academic wireless research community.

It has become a common requirement for testbeds to provide a service in charge of collecting and exposing metrics, thus assisting experimenters with the central task of data collection. This paper describes Kwollect, a service designed and developed in the context of the Grid'5000, to collect infrastructure metrics (including high-frequency wattmeters) and expose them to experimenters. Kwollect scales to high frequencies of metrics collection for hundreds of nodes. It can also be leveraged by the experimenter to collect custom metrics.

As voice- and video-based services that require low latency become increasingly essential, a tremendous research effort has been devoted to improving the latency performance of networks. However, there remains a "latency divide" between those whose Internet access is subject to high latency and those with low-latency access. Furthermore, it is unclear whether the aforementioned research efforts narrow this divide (by improving the latency performance of poor-quality connections the most), maintain the divide (by improving poor- and high-quality connections equally), or even widen the divide (by mostly reducing latency of high-quality network connections). In this work, we design a reproducible experiment for evaluating the performance of Copa, a delay-based congestion control protocol, in different categories of residential broadband connections. We use empirical data from U.S. home network connections to evaluate whether Copa leads to more equitable network access. We confirm that Copa has potential to narrow the ``digital divide'' by improving latency performance for users with a low-quality residential broadband connection, although care must be taken to set appropriate parameters for best performance.

The need to find more efficient ways to share and use wireless spectrum has resulted in renewed interest in radio frequency (RF) propagation modeling. The open and programmable nature of the POWDER (Platform for Open Wireless Data-driven Experimental Research) mobile and wireless platform offers a unique environment in which to test and validate RF propagation modeling approaches. In this paper we present our work illustrating how POWDER based RF measurements can be performed, as a form of "ground truth", and compared with predicted RF signal strength based on a propagation model. We make use of the Shout RF measurement framework available on POWDER to perform a series of RF measurements. We compare these measurements with predicted power levels using the open source RF Signal Propagation, Loss, And Terrain (SPLAT!) analysis tool. We present our results and a brief terrain analysis to provide real-world context for it. Our work is "packaged" as a POWDER profile to allow others to repeat our analysis and to serve as a starting point for further RF measurement and propagation related research.

Real world testbeds, like the POWDER platform (Platform for Open Wireless Data-driven Experimental Research), enable a broad range of mobile and wireless research. Given the flexibility of this platform, a key concern for platform users is selecting a set of wireless resources that will satisfy the requirements of their experiments. In this paper we present the design and implementation of WiMatch a wireless resource matchmaking system. We illustrate the utility of our approach by evaluating it in the POWDER platform.