WebRTC at the hands of a telecom vendor.
The Telecom world has its own set of standards and needs. At times, they seem far remote from the way the Internet and WebRTC operates.
How do you bridge between the two? André Silva, Team Leader & WebRTC Product Manager at WIT Software tries to explain in this interview.
What is WIT Software all about?
WIT is a software development company specialized in advanced solutions for mobile telecommunications companies. The company has over 14 years of experience and a deep expertise in mobile communications and network technologies including IP Multimedia Subsystem (IMS), mobile voice (Mobile VoIP and Voice over LTE), messaging (SMS, MMS and IM), Rich Communication Suite (RCS) and Multimedia Telephony Services (MMTel). Located in Portugal, UK, Germany and California, the company has over 230 fulltime employees and a blue chip industry client base.
You’ve been working in the Telco space offering IMS and RCS products. What brought you towards WebRTC?
Back to 2008, WIT started the development of a Flash-to-SIP Gateway to support voice calls from web browsers to mobile phones. The first commercial deployment was done in 2011, enabling calls from a Facebook App to mobile subscribers connected to the Vodafone Portugal network. This first version included features like enhanced address-book, presence, IP messaging, IP voice calls and video calls.
When Google released the WebRTC project back in 2011, WIT started following the technology and as soon as it got stable we have implemented a new release of our Web Gateway with support for all the browsers in the market, including Chrome, Firefox and Opera that are WebRTC-compliant, but also Safari and IExplorer where we use the Flash-to-SIP capabilities.
How are your customers responding to the WebRTC capabilities you have?
Our customers are searching for ways to extend their mobile/fixed networks to web browsers and IP devices, either to extend voice calling with supplementary services and SMS, or to make more services available to off-net users. We are providing our WebRTC Gateway and our RCS capabilities to provide richer messaging and voice calling use-cases for the consumer and the enterprise market.
One of the facts that is much appreciated is the support for non-WebRTC browsers. The conversion of protocols (DTLS-SRTP and RTMP) to RTP is done by our Gateway and it is transparent for the network.
For codec transcoding, we support the standard JSR-309 to integrate with MRF’s in order to support extra codecs that are not natively available in WebRTC.
Recently we just announced a partnership with Radisys that is a leading provider of products and solutions, to address emerging media processing challenges for network operators and solution vendors.
What signaling have you decided to integrate on top of WebRTC?
We are using a proprietary JSON protocol over WebSockets. This is a lightweight protocol that exploits the best of asynchrony of WebSockets and provides the best security for Web Apps.
We have built a Javascript SDK that abstracts all the heterogeneity of the different browsers, and the technology that is used to establish calls. The Javascript SDK loads a Flash plugin when WebRTC is not available in the browser.
Backend. What technologies and architecture are you using there?
WIT WebRTC Gateway is a Java-based Application Server that can run in several containers. It can be scaled horizontally over several instances. The Gateway integrates with SIP Servlet Containers, for the integration with standard Media Servers, and with streaming servers, to make the media available over RTMP. Our Media engine copes with the WebRTC media and contains a STUN/TURN server to solve the NAT traversal issues.
Where do you see WebRTC going in 2-5 years?
I think WebRTC will become the standard for IP Communications that every VoIP application and server will support, either because they use the WebRTC native APIs, or because they will be improved to also support the extras brought by WebRTC specification.
In 2-5 years I expect to see web developers using the WebRTC JavaScript API to create new applications and just assume that WebRTC is there accessible in every browser, since Microsoft is moving forward to add WebRTC in the new browser.
On the negative side, I also expect browsers to continue having distinct implementations which will force developers to have specific code for each browser. Unfortunately, web development has always been like this.
If you had one piece of advice for those thinking of adopting WebRTC, what would it be?
WebRTC aims to enable VoIP without plugins. So you need to think about WebRTC alternatives for the cases where it is not available, because from our experience, the end user doesn’t really care what’s underneath the application, they just want it to work.
So, you should not filter the browsers or systems where your application will run and force the user to download a new browser.
Given the opportunity, what would you change in WebRTC?
Since H.264 is now one of the video codecs in the specification, a great step would be to add some audio codecs like AMR-WB and G.729 to avoid transcoding with some of the common codecs in existing services.
Also, I would give more focus to the advanced cases that depend on the renegotiation of the WebRTC sessions. We provide supplementary services like call hold, upgrade and downgrade and there are still some limitations in the APIs to allow us to have full control across browsers.
What’s next for WIT-Software?
We are creating WebRTC applications that will be launched later this year for the consumer market, and we are preparing a solution for the enterprise market that will leverage the best of WebRTC technology.
Our latest implementation adds support to voice calls between web browsers and VoLTE devices, and this is a major breakthrough for the convergence of Web Apps and new generation mobile networks.
For more information, please visit our product page at http://webrtc.gw
–
The interviews are intended to give different viewpoints than my own – you can read more WebRTC interviews.