WebRTC standards improve interactive audio/visual communications via the world-wide web
WebRTC (Web Real-time Communication) enables real-time interactive video conferencing for the first time as a native component of web browsers. Dr Perkins’ research developed circuit breakers and congestion control feedback to ensure WebRTC is safe to deploy and has good performance. It has been incorporated into the international technical standards for WebRTC, helping browsers and other video conferencing systems recognise and react to network congestion, adapting video quality to match network capacity. Global deployment of WebRTC-based video conferencing has transformed education, telemedicine, business, and family communications — a transformation which has accelerated during the COVID-19 pandemic. Dr Perkins’ work helps ensure such use does not overwhelm the network.
Context and societal impact
Developed by the Internet Engineering Task Force (IETF) and the World Wide Web Consortium (W3C), WebRTC is a free, open source browser extension used in modern web browsers, adding new network protocols and programming interfaces to support secure, high-quality, peer-to-peer, real-time interactive voice, video, and data. This allows new classes of interactive audio/visual web applications, including high-quality video conferencing, to be developed and deployed within any web browser.
From the first cross-browser video call in February 2013, WebRTC has been adopted by all major web browsers, desktop and mobile, supporting applications including Facebook Messenger, Google Meet and Hangouts, Discord, Snapchat, Cisco WebEx, Microsoft Teams, and Skype.
This innovation comes with a risk, however. High-quality interactive video needs significant network resources, and rapid deployment of such web applications has potential to congest the network due to the increased traffic demand, degrading the user experience and disrupting other Internet applications.
Research led by UofG's Dr Colin Perkins focussed on removing that risk through the development of circuit breaker algorithms that monitor progress of an ongoing video call and its impact on other traffic, halting the call if it is causing congestion that will disrupt other users of the network. These algorithms monitor reception quality feedback provided by the WebRTC Real-time Transport Protocol (RTP) and determine if the video is causing persistent and excessive congestion or network disruption based on heuristics derived from knowledge of user perception of video quality, video coding, network dynamics; and behaviour of network transport protocols (e.g., RTP, UDP, TCP/IP).
Using the information obtained, the algorithms determine if it is necessary to reduce the sending rate or terminate the flow, breaking the circuit. Dr Perkins led the incorporation of this research into the WebRTC standards developed by the IETF. The IETF is the leading international standards organisation that develops the technical standards that define the internet (IETF developed the network protocol aspects of WebRTC; the World-Wide Web Consortium (W3C) integrated with the web programming model). Dr Perkins has been involved in IETF since the mid-1990s, co-authoring >30 RFCs, chairing the Audio/Video Transport (1998–2008), Multiparty Multimedia Session Control Working Group (2000–2007), and Real-time Media Congestion Control Working Groups (2016–). He chairs the Internet Research Task Force (https://irtf.org/), the research arm of IETF, and is a member of the Internet Architecture Board (https://iab.org/).
The circuit breaker is referenced as “MUST implement” in the WebRTC media transport specification (DOI:10.17487/RFC8834). Independently, the Real-time Streaming Protocol standard (DOI:10.17487/RFC7826), a core component of the 3GPP Multimedia Broadcast/Multicast Service, cited a draft of the circuit breaker as “MUST implement”.
WebRTC also uses detailed congestion control feedback to adapt video quality to match available network capacity, rather than simply terminating problematic flows. Devising scalable feedback algorithms that can operate across the range of scenarios where WebRTC is used (from point-to-point phone calls to high-quality multiparty interactive video conferences) is difficult due to timing constraints inherent in the underlying RTP protocol, the limited adaptation rate of video codecs, and limits on the feedback overhead that can be tolerated. Our research showed how to adopt Explicit Congestion Notification from the network for use with the UDP transport underpinning WebRTC, and in collaboration with researchers from Ericsson and the University of Oxford, Dr Perkins worked to incorporate this into the WebRTC standards.
Dr Perkins co-chairs the Real-time Media Congestion Avoidance Techniques working group in the IETF (https://datatracker.ietf.org/wg/rmcat/about/), developing congestion control algorithms for future versions of WebRTC, and was a key member of the design team that developed congestion control feedback for WebRTC to allow interoperability of congestion control algorithms from different vendors, and performed simulations to demonstrate its effectiveness and low-overhead. Dr Perkins’ work on Explicit Congestion Notification is seeing interest for future versions of WebRTC and was also published as an IETF standard. He was also editor of the IETF WebRTC media transport standard that incorporates these results and extensions, and has been a long-term editor of the Session Description Protocol signalling standard. These are two of the mandatory-to-implement core specifications that define WebRTC.
Finally, Dr Perkins' team collaborated with engineers from Ericsson, Huawei, and Vidyo to develop extensions to the RTP standards that improve its scalability and reduce overheads, allowing WebRTC to work effectively on low-capacity links, including dial-up connections in developing regions, and on poor quality 3G/4G wireless links. These have enhanced the reach of WebRTC, allowing it to compete with traditional telephony in developing regions, and providing a more robust user experience in wireless environments.
WebRTC is integral to a new ecosystem of applications and services such as Skype for Business, Zoom, and Google Duo, facilitating communication in areas such as education, telemedicine and business. A Google news search for WebRTC finds 71,400 articles relating to the technology, its applications, and companies building on the platform. For example, Attend Anywhere has led in the transformation of healthcare delivery across the UK and Australia, initially using Vidyo, and then WebRTC from 2014 to provide high-quality secure video consultations across the healthcare sector.
The global restriction in travel in response to the coronavirus pandemic combined with the cross-platform browser capability of WebRTC has facilitated real-time audio-visual communication for business, healthcare and education across the Internet. In the first 4 months of the COVID-19 pandemic, real-time voice and video has grown more than 200% in traffic and daily conferencing minutes together with 20-fold increases in users of conferencing platforms. In April 2020, Google announced that over 2 million new users were connecting on Google Meet, collectively spending more than 3,800 years of secure meetings each single day. This has continued to be supported as people work and learn from home, with 100,000 schools across 20 countries connecting using Zoom. The UofG work on circuit breakers, congestion control feedback, and scalability for WebRTC have helped the network adapt to these changes, allowing businesses to connect with customers, patients with their doctors, students with their teachers, and friends and families across the globe to stay in touch during the global pandemic. The resultant global market is anticipated to reach USD50 billion by 2026.