LAUSR

ABSTRACT With so many different devices with varying video processing capabilities in use today, an adoptive distribution of live video streaming that is capable of meeting the end user expectations and that utilizes the end user's hardware to the maximum has become a major priority for live video streaming service providers. Meeting this challenge requires upgrading of the traditional video scheduling algorithm so that the provision of best video quality does not compromise the integrity of the video stream. This paper analyses modelling of the video chunk scheduling for layered live video streaming in mesh pull-based CDN-P2P network and introduces a new algorithm that improves modelling performance. The proposed chunk scheduling algorithm is based on a taboo search and optimizes the video throughput and the video delivery ratio. It uses chunk playback deadline together with chunk layer dependency to discredit chunks whose deadlines have expired and those chunks that do not have the corresponding lower layer values. Finally, the simulation results show that the proposed algorithm performs significantly better than the traditional chunk scheduling approach.