Implementation and Performance Analysis of a Video Streaming System using WEBRTC and Equal Cost Load Balancing

Abstract

Real-time Internet based video communication applications are becoming widely adopted by internet users. However, the exponentially rising numbers of internet users is proving to be a real challenge to guarantee a good Quality of Service (QoS) for such applications. This paper investigates how video streaming quality is impacted by the increasing number of users using a video streaming application developed with Web Real Time Communications (WebRTC). The application considers both live video streaming from of an event captured by the initiator’s webcam and live streaming of a video file from the initiator’s device. Tests were performed using a Wi-Fi connection with different types of video sources. Moreover, equal cost load balancing was used to assess the impact on the video quality two servers are used at the initiator’s site. Videos were recorded for five minutes on both the initiator and client sides using the application’s record function and the Peak Signal to Noise Ratio (PSNR) was used to assess the quality of the videos. Results showed that with equal cost load balancing using two servers a maximum gain of up to 5.39 dB in PSNR could be obtained with live streaming from a webcam to ten clients.