The paradigm shift where realtime communication will be handled on the web is soon becoming a reality. With only a few lines of Javascript code developers can easily turn a web browser into a voice or video terminal using the WebRTC API. While the IETF is actively working on standardizing the RTCWeb protocol to enable realtime communication across browsers and to avoid failures in session setups and enhance the session’s Quality of Experience (QoE), there is still a lack of a standardized congestion control mechanisms in the area. This paper proposes and evaluates a congestion avoidance mechanism and a rate adaptation model for WebRTC interactive video sessions in LTE networks where the model is based on realtime available bandwidth estimation and measurements. Our simulation results show that the proposed model provides a fair bandwidth allocation between TCP traffic flows and realtime media traffic flows. The results also show that excessive video frame delays and packet losses can be prevented with the proposed congestion avoidance mechanism. Consequently the proposed model improves performance and perceived QoE of WebRTC.
展开▼