Efficient delivery of multimedia content over multipath networks

Abstract

The objective of the proposed research is to develop a unique approach to efficiently deliver high-quality multimedia data by leveraging concepts from various multipath networking models in conjunction with ABR protocols that supplement the reliance on client-side network bandwidth estimation and/or buffer-state. Prior research into the development and evaluation of the performance of multipath-enabled multimedia streaming techniques are presented. Machine learning based multipath-enabled techniques are proposed in the development of selective transport protocols for streaming applications and validating their effectiveness.