Design and Simulative Performance Evaluation of a QoE Fair Adaptive Streaming Mechanism

Type:
Master Thesis Computer Science
    Status:
    completed
    Tutor:
    Examiner:
    Volker Gruhn
    Examiner:

    Abstract

    In this thesis, we design a Quality of Experience (QoE) fair adaptive streaming algorithm and a simulation to evaluate its performance. We propose a design with an external coordinator that assesses QoE for clients behind a bottleneck and tries to maximize the QoE while reducing stalling and low video quality playback. The simulation was written with simpy, a discrete event framework in Python. We simulated multiple clients with configured and random arrival times and assessed the performance of the design adaptation strategy in comparison to other throughput- respectively buffer-based adaptation strategies. We found that the adaptation strategy did not outperform the buffer-based strategy on fairness and quality level playback