Adaptive Real-Time Video Streaming System for Best-Effort IP Networks

This paper presents an adaptive real-time video streaming (AVS) system that delivers real-time video stream from a network camera to multiple video receivers. The system provides an application layer rate-adaptation mechanism that adapts video quality and bit rate according to the network congestion conditions and the available bandwidth. The goal of the system is to improve the smoothness of the video playback at the receivers under varying network conditions. This is performed by concluding the highest possible video bit rate that can be used without facing bandwidth limitation problems. The system was tested and evaluated in private Ethernet local area network (LAN) in which packet loss and limitation on bandwidth was generated by utilizing a network impairment emulation tool. Based on the performed tests, the system behavior was predictable. The research is based on the constructive method of the related publications and technologies and the results are derived by the implemented AVS system.