A predictive video-on-demand bandwidth management using the Kalman filter over heterogeneous networks

In order to adapt the quality of an on-demand video stream over a time-varying bandwidth channel, a network-aware bandwidth estimation and rate control scheme are required. This paper proposes a predictive video-on-demand (VoD) bandwidth management and a feedback-based buffer control scheme for streaming fine granular scalability videos over wired/ WLAN/3G networks. The predictive VoD bandwidth management includes two parts: bandwidth estimation and rate adaptation. According to the measured information of packet round-trip-time, loss-rate, delay jitter and received bit-rate, an improved Kalman filter is proposed to predict an available bandwidth recursively, and to determine a proper transmission rate in consideration of buffer fullness of a decoder. The optimal parameters of the Kalman filter, e.g. a transition matrix and error covariances, can be initialized, converged and adapted to characteristics of the current network. In our experiments, distinct network traffic models are simulated in comparison with pathChirp and one Republic of China patent. The corresponding estimation results with respect to network information are also exhibited in the real networks.