Mobile networks are embracing disaggregation, reflected by the industry trend towards Open RAN. Private 5G networks are viewed as particularly suitable contenders as early adopters of Open RAN, owing to their setting, high degree of control, and opportunity for innovation they present. Motivated by this, we have recently deployed Campus5G, the first of its kind campus-wide, O-RAN-compliant private 5G testbed across the central campus of the University of Edinburgh. We present in detail our process developing the testbed, from planning, to architecting, to deployment, and measuring the testbed performance. We then discuss the lessons learned from building the testbed, and highlight some research opportunities that emerged from our deployment experience.

Throughput and latency critical applications could often benefit of performing computations close to the client. To enable this, distributed computing paradigms such as edge computing have recently emerged. However, with the advent of programmable data planes, computations cannot only be performed by servers but they can be offloaded to network switches. Languages like P4 enable to flexibly reprogram the entire packet processing pipeline. Though these devices promise high throughput and ultra-low response times, implementing application-layer tasks in the data plane programming language P4 is still challenging for an application developer who is not familiar with networking domain. In this paper, we first identify and examine obstacles and pain points one can experience when offloading server-based computations to the network. Then we present P4rrot, a code generator (in form of a library) which allows to overcome these limitations by providing a user-friendly API to describe computations to be offloaded. After discussing the design choices behind P4rrot, we introduce our proof-of-concept implementation for two P4 targets: Netronome SmartNIC and BMv2. To demonstrate the applicability of P4rrot, we investigate case studies in the context of publish-subscribe sensor data processing and real-time data streaming, supporting, in particular, MQTT-SN and MoldUDP packets.

This April 2022 issue contains five technical papers and two editorial notes. The first technical paper, Data-Plane Security Applications in Adversarial Settings , by Liang Wang and colleagues, investigates security issues that may arise when creating and running data-plane applications for programmable switches. This work moves security analysis and design forward in this particular area. This paper also calls for a more thorough rethinking of security for data-plane applications for programmable switches. The second technical paper, One Bad Apple Can Spoil Your IPv6 Privacy , by Said Jawad Saidi and colleagues, leverages IPv6 passive measurements to pinpoint that a non-negligible portion of devices encodes their MAC address in their IPv6 address. This threatens users' privacy, allowing content providers and CDNs to consistently track users and their devices across multiple sessions and locations. Overall, the paper is an excellent contribution toward privacy-by-design solutions and a nicely executed measurements study that clarifies the problem and provides solid suggestions to mitigate the problem. The third technical paper, Hyper-Specific Prefixes: Gotta Enjoy the Little Things in Interdomain Routing , by Khwaja Zubair Sediqi and colleagues, investigates the presence of high-specific prefixes (HSP) on the BGP Internet routing during the last decade. These prefixes are more-specific than /24 (/48) for IPv4 (IPv6) and are commonly filtered by Autonomous Systems operators. Overall this paper offers a nice contribution to the understanding of the BGP universe, with a clear message and a nice quantification of the phenomenon. The authors clearly present and motivate the work, offering also to not experts a nice view of the routing complexity of the nowadays internet. The fourth technical paper, Programming Socket-Independent Network Functions with Nethuns , by Nicola Bonelli and colleagues, proposes a new solution to transparently develop packet-processing programs on top of different network I/O frameworks. The authors design and develop an open-source library, nethuns, serving as a unified programming abstraction for network functions that natively supports multi-core programming. Not only is this work very relevant to our community, but also the code is released open-source through a BSD license, which can be used to foster more research in the area, towards unifying programming mechanisms of end-host networking. The fifth technical paper, Measuring DNS over TCP in the Era of Increasing DNS Response Sizes: A View from the Edge , by Mike Kosek and colleagues, studies one of the foundations of today's Internet: the Domain Name Service (DNS). The original RFC document of DNS instructs to send queries either over UDP (DoUDP) or TCP (DoTCP). This paper presents a measurement study on DoTCP focusing on two perspectives: failure rates and response times. Finally, we have two editorial notes. A Case for an Open Customizable Cloud Network , by Dean H. Lorenz and his colleagues, argues for the desirability of the new ecosystem of managed network solutions to connect to the Cloud, outlines the main requirements and sketches possible solutions. Recommendations for Designing Hybrid Conferences , by Vaibhav Bajpai and colleagues, presents guidelines and considerations-spanning technology, organization and social factors-for organizing successful hybrid conferences. I hope that you will enjoy reading this new issue and welcome comments and suggestions on CCR Online (https://ccronline.sigcomm.org) or by email at ccr-editor at sigcomm.org.

Autonomous Systems (ASes) exchange reachability information between each other using BGP---the de-facto standard inter-AS routing protocol. While IPv4 (IPv6) routes more specific than /24 (/48) are commonly filtered (and hence not propagated), route collectors still observe many of them. In this work, we take a closer look at those "hyper-specific" prefixes (HSPs). In particular, we analyze their prevalence, use cases, and whether operators use them intentionally or accidentally. While their total number increases over time, most HSPs can only be seen by route collector peers. Nonetheless, some HSPs can be seen constantly throughout an entire year and propagate widely. We find that most HSPs represent (internal) routes to peering infrastructure or are related to address block relocations or blackholing. While hundreds of operators intentionally add HSPs to well-known routing databases, we observe that many HSPs are possibly accidentally leaked routes.

Virtual reality systems today cannot yet stream immersive, retina-quality virtual reality video over a network. One of the greatest challenges to this goal is the sheer data rates required to transmit retina-quality video frames at high resolutions and frame rates. Recent work has leveraged the decay of visual acuity in human perception in novel gaze-contingent video compression techniques. In this paper, we show that reducing the motion-to-photon latency of a system itself is a key method for improving the compression ratio of gaze-contingent compression. Our key finding is that a client and streaming server system with sub-15ms latency can achieve 5x better compression than traditional techniques while also using simpler software algorithms than previous work.

Presentation of the July 2021 issue of CCR.

This interview is part of a series on Great Educators in Computer Networking, where we interview some of the most impactful and skilled educators in our field. Here, we interviewed Australian Bruce Davie, the self-described computer scientist/engineer/runner/cyclist, who agreed to talk to us about his thoughts on computer networking education, his role in it, his thoughts about the big ideas in our field, and how the pandemic is changing our work. Bruce has over 30 years of industry experience and is well known for a broad spectrum of educational initiatives such as co-authoring several textbooks, as well as his contributions to many networking standards and technologies, including IP quality of service, network virtualization, software defined networking, and more

On 16-17 December 2020, CAIDA hosted the 11th interdisciplinary Workshop on Internet Economics (WIE) in a virtual Zoom conference. This year our goal was to gather feedback from researchers on their experiences using CAIDA’s data for economics or policy research. We invited all researchers who reported use of CAIDA data in these disciplines. We discussed their successes and challenges of using the data, and how CAIDA could help these fields via Internet measurement and data curation. To avoid Zoom fatigue, we had a conversation-focused rather than presentation-focused workshop. Research topics we discussed included: Internet data for macroeconomics; connectivity and its effect on economic interdependence; effects of the EU’s new GDPR on internet interconnection; measuring corporate cyber risk; measuring work-from-home trends; measuring the economic value of open source software; and more generally how to best support evidence-based policymaking.

Since the beginning of the COVID-19 pandemic, governments introduced several social restrictions. As of 18 March 2020, more than 250 million people were in lockdown in Europe. This drastically increased the number of online activities. Due to this unprecedented situation, some concerns arose about the suitability of the Internet network to sustain the increased usage. Italy was severely hit by the first wave of the pandemic and various regions underwent a lockdown before the main country-wide one. The Italian network operators started sharing information about improvements carried out on the network and new measures adopted to support the increase in Internet usage. In this report, by means of a questionnaire, we collect information and provide a quantitative overview of the actions undertaken by network operators in Italy. The attitude of Italian operators was synergic and proactive in supporting the changed market conditions caused by the public health emergency.

Many countries are deploying Covid-19 contact tracing apps that use Bluetooth Low Energy (LE) to detect proximity within 2m for 15 minutes. However, Bluetooth LE is an unproven technology for this application, raising concerns about the efficacy of these apps. Indeed, measurements indicate that the Bluetooth LE received signal strength can be strongly affected by factors including (i) the model of handset used, (ii) the relative orientation of handsets, (iii) absorption by human bodies, bags etc. and (iv) radio wave reflection from walls, floors, furniture. The impact on received signal strength is comparable with that caused by moving 2m, and so has the potential to seriously affect the reliability of proximity detection. These effects are due the physics of radio propagation and suggest that the development of accurate methods for proximity detection based on Bluetooth LE received signal strength is likely to be challenging. We call for action in three areas. Firstly, measurements are needed that allow the added value of deployed apps within the overall contact tracing system to be evaluated, e.g. data on how many of the people notified by the app would not have been found by manual contact tracing and what fraction of people notified by an app actually test positive for Covid-19. Secondly, the 2m/15 minute proximity limit is only a rough guideline. The real requirement is to use handset sensing to evaluate infection risk and this requires a campaign to collect measurements of both handset sensor data and infection outcomes. Thirdly, a concerted effort is needed to collect controlled Bluetooth LE measurements in a wide range of real-world environments, the data reported here being only a first step in that direction.