RTCWeb Evolution

• Applicants: Professor Dr.-Ing. Erwin P. Rathgeb; Professor Dr. Michael Tüxen
• Subject Area: Security and Dependability, Operating-, Communication- and Distributed Systems
• Term: from 2014 to 2019
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 259236644

The use of the Internet for integrated audio and video communication is increasing substantially. At the moment, there exist multiple proprietary solutions, e.g. Skype, Facetime or Google Hangouts. A significant limitation of these solutions is that they require the installation of proprietary software and that they are not interoperable. Therefore, the IETF and the W3C are working together in order to develop a standardized extension of web browsers enabling a direct real-time communication between browsers for media (audio/video) and non-media (general data) application data. The communication protocols are defined by the IETF working group RTCWeb, whereas the programming interface (Javascript API) is developed by the W3C working group WebRTC. The protocol stack currently being specified by the IETF under the name RTCWeb has to provide a secure communication over NAT boundaries (Network Address Translation) for media as well as non-media streams. To achieve this, already existing protocols are combined in a new way. The transport protocol SCTP (Stream Control Transmission Protocol) with its congestion control feature is one of the key building blocks of the RTCWeb architecture and is in the focus of this project. The novel combination of existing protocols in the RTCWeb context will be evaluated, improved and extended systematically during the project. Of particular scientific interest is the question how to coordinate and control the various media and non-media streams in the RTCWeb context taking into account their heterogeneous requirements and properties in order to provide the expected behavior for the user. RTCWeb combines the various streams within one application. This provides the opportunity to coordinate the involved mechanisms - in particular congestion controls - in an active and target-oriented way. Options to exploit this shall be developed and evaluated. Initial investigations of the project partners have already shown that the standard congestion control of SCTP is not able to protect the media streams from concurrent non-media streams. Therefore, alternative congestion controls shall be developed and evaluated in the project. In addition to the consideration of the SCTP congestion control alone also the currently proposed methods with respect to congestion control for media streams as well as the active queue management mechanisms for access routers shall be taken into account. These very diverse mechanisms are currently developed separately and their interaction is a crucial aspect which has not yet been evaluated systematically. The goal of the project is to achieve a significant performance improvement of the overall system by means of targeted evaluations in combination with the development of concepts and mechanisms for coordinated congestion control.

• DFG Programme: Research Grants