Bandwidth allocation on video content distribution servers

Abstract

Video streaming applications have implementation issues when the transmission rate is much less than the video playback rate. Caching the beginning part of the video in large storage device before playing is a common solution. In this paper, we propose a bandwidth allocation scheme on video content distribution servers. This algorithm is aim to control the server sending rate with efficient bandwidth resource allocation. In order to reduce the impact of network rate jitter, we present a client cache scheme considering the average receiving rate. On servers, on-line clients are grouped into three queues based on different serving stages: playing, downloading or waiting to play. The sending rate is allocated to each client dynamically. Clients who are watching video have the highest priority. For the waiting clients, a weighted bandwidth allocation approach is presented to minimize users' waiting time. The simulation results show that the proposed bandwidth allocation scheme can achieve a high performance.