MEC Resource Offloading for QoE-Aware HAS Video Streaming
Abd‐Elhamid M. Taha, Najah A. Abu Ali, Hao Ran Chi, Ayman Radwan · 2021
With the popularity of mobile edge computing (MEC), video streaming's peer-offloading strategy significantly affects the Quality of Experience (QoE) performance of video streaming for mobile users (MUs). Improving the service quality regarding the MUs' demand has become a vital challenge, reflected by QoE. This paper proposes a QoE-aware MEC-based peer-offloading method for HAS-based video streaming, called QOMECS. The proposed method considers dynamic MUs' demands and corresponding QoE requirements. We categorize the QoE KPIs into perceptual and systemic sectors. We formulate the corresponding transmission, computation, and offloading for MEC-based HAS into a QoE maximization problem. We propose a reverse-fuzzied particle swarm optimization (R-FPSO), to solve the highly nonlinear and perceptual-oriented optimization formulation. Unlike conventional fuzzy logic, R-FPSO reverses the fuzzification process by fuzzifying PSO's output (i.e. translated QoE KPIs into satisfactory levels) and further updates the particle values and velocities in the PSO process. Simulation results show that the proposed QOMECS dramatically improves the edge computing efficiency, with optimized QoE performance.