QoS routing design for adaptive streaming in Software Defined Network

Abstract

Adaptive streaming over HTTP(DASH) is a popular method to deliver best possible quality media using available network resources. In this paper we propose QoS routing algorithm design for providing high quality and continuous stream of media streaming in SDN(Software Defined Network). To provide seamless DASH service over SDN is a challenge because traditional DASH service manner mismatches with SDN per-flow centralized management architecture. Therefore, it is necessary to have some interactions between communication layers for making routing decision. Firstly, we propose server driven bit rate estimation approach to computes video bit rate and inform the application QoS requirement to the control layer. The media server estimates the resources available and dynamically adapts the video bit rate in order to reduce additional software stacks on end-user devices. We also present a Kalman Filter based rate adaptation mechanism that predicts the next bit rate of the media in a seamless manner. Secondary, we propose a QoS routing design for adaptive stream, it allows SDN controller evaluates all passable paths based on whole network topology by taking the bit rate of the segments in to account. We perform an experiment and presented the server driven bandwidth estimation mechanism to compute the appropriate bit rate and rate switching in smooth manner. We also construct a SDN testbed for QoS routing algorithm for adaptive streaming, it shows additional reroute step to find path that satisfy the bitrate before downgrade the video quality.