As the ability for more people to communicate in new ways expands, consumer needs are centered around fostering human connection through technology. Nitin’s talk will include the primary user needs which are driving the future of RTC, and how the latest innovations are shaping the future ahead including in the Metaverse.

Providing a natural real-time audio communication experience at the scale of billions of users across WhatsApp, Instagram and Messenger is challenging, particularly given the diversity of devices and network conditions we encounter. Echo (hearing yourself back) and audio quality under low bandwidth network conditions, are two well known problems that can seriously disrupt the call experience. In this talk, we will discuss the results of our long-term investments and innovation in these two areas as well as recent launches that revamp the core of Meta’s audio processing stack powering calling across our Family of Apps.

Providing a natural real-time audio communication experience at the scale of billions of users across WhatsApp, Instagram and Messenger is challenging, particularly given the diversity of devices and network conditions we encounter. Echo (hearing yourself back) and audio quality under low bandwidth network conditions, are two well known problems that can seriously disrupt the call experience. In this talk, we will discuss the results of our long-term investments and innovation in these two areas as well as recent launches that revamp the core of Meta’s audio processing stack powering calling across our Family of Apps.

Long distance calls can be challenging, but Meta's WhatsApp and Remote Presence orgs have some valuable insights to share. This presentation will explore how Meta's global network of edge clusters can be leveraged to build a stable and high-quality network infrastructure for serving over 500+ million long distance video and audio calls worldwide on daily bases. By utilizing Meta's extensive resources and expertise, we can improve the user experience and provide more reliable connections for long distance calls.

This presentation covers how Meta is improving the video quality for RTC apps by moving to next gen Video Codecs like AV1 to improve quality for low bandwidth, bringing HD video to high end networks and devices.

What does it take to develop and deploy AV1 in RTC applications? This presentation will discuss how RTC is different from other video coding applications and why that is challenging for developing an encoder that performs well. While other applications typically place coding performance above all other requirements, in RTC an encoder must meet demanding trade-offs with computational cost and be subject to many constraints. As encoding standards get more complex, meeting these constraints requires careful consideration of how to approach encoding.

5G and Edge Computing are now enabling new user experiences for immersive media such as 360-degree video, where content creation is in real-time and equally uses upstream and downstream links. Here we will present a 360 Immersive Media solution using the Intel Open WebRTC Toolkit (OWT) and Intel edge platform, which is composed of media ingestion from multiple cameras, remote control, and distribution over 5G networks. A performance evaluation showing the gain in terms of bandwidth and latency will be presented as well. In addition, a more generic framework is proposed for delivering spatially partitioned content using RTP.

Network resiliency algorithm is a key part to ensure seamless media in RTC. We will discuss Meta's recent improvements to Network Resiliency with Long Term Reference (LTR) and Reed Solomon Code.

WebRTC is going through turmoil these days. On one hand, contributions aren’t high to the codebase, but on the other hand, we see a lot of minor improvements and changes, hinting at the future of WebRTC. Who are the people contributing to the base libWebRTC codebase used in all modern browsers today, and what are the areas of interest they are covering? This is what we will review during this session, trying to glean out of these tidbits where we are headed to with WebRTC.

In this talk, we will go over observability for RTC applications which is critical to ensuring good operational monitoring and debugging to scale across billions of users. We'll cover the entire space in detail and also dive into the architecture of RAlligator, a real time data aggregation service built specifically for analyzing data from all Meta RTC applications. Finally, we'll go over all the different debugging tools and various different types of logs used by Rsys.

Sean will talk about lessons learned from his journey in Open Source. The joys of building something with huge impact. The frustrations when Open Source is done wrong. The success that can be found when doing it effectively. When done well Open Source has huge benefits for users, developers and companies.

The existing congestion control/bandwidth estimation algorithms have several hand-tuned parameters and actions depending on network conditions, leading to challenges in engineering, experimentation, maintenance, and operational monitoring. A machine learning based approach can offer a cleaner alternative with improved performance by addressing the problem holistically across layers (BWE, Network Resiliency, and Transport). This talk outlines our method for addressing BWE and CC issues in RTC using machine learning, discussing obstacles encountered, recent findings, and upcoming plans. Specifically, we explore network characterization and prediction problems, showcasing a few representative examples.