QoS-sensitive transport of real-time MPEG video using adaptive redundancy control

  • Authors:

  • K. Park;W. Wang

  • Affiliations:

  • Network Systems Lab, Department of Computer Sciences, Purdue University, West Lafayette, IN 47907, USA;Network Systems Lab, Department of Computer Sciences, Purdue University, West Lafayette, IN 47907, USA

  • Venue:
  • Computer Communications
  • Year:
  • 2001

Quantified Score

Hi-index 0.24

Visualization

Abstract

This paper presents an adaptive end-to-end protocol for quality of service (QoS)-sensitive transport of real-time MPEG video using packet-level forward error correction in dynamic networks. The objective is to facilitate a user-specified QoS end-to-end - i.e. without special network support - for real-time MPEG video whose timing constraints rule out the use of retransmission-based congestion control and QoS provisioning schemes. The degree of redundancy - overcode in forward error correction - injected into the network is adjusted as a function of network state, decreasing when the network is well-behaved to minimize unnecessary network resource consumption, and increasing when it is not to compensate for adverse network effects so as to maintain an invariant level of end-to-end QoS. We describe an adaptive packet-level forward error correction protocol called AFEC and analyze its properties with respect to optimality and stability. The optimal control problem is nontrivial due to the fact that increased redundancy, beyond a certain point, can backfire resulting in self-induced congestion which impedes the timely recovery of information - MPEG video frames - at the receiver. We experimentally evaluate the efficacy of our end-to-end QoS control by implementing and customizing AFEC to the transport of real-time MPEG video. We realize the system as a transport/application layer protocol running on UNIX workstations, and measure its performance over controlled network environments. We show that AFEC is able to effectively hide potentially adverse network effects such as packet drops and delays stemming from traffic burstiness and nonstationary structural changes using adaptive redundancy control, exporting a constant QoS service commensurate with user-specified QoS.