cs 15(2): e4

Research Article

A New QoE-Driven Video Cache Allocation Scheme for Mobile Cloud Server

Download1016 downloads
  • @ARTICLE{10.4108/eai.19-8-2015.2260126,
        author={Xiaojiang Zhou and Mengyao Sun and Yumei Wang and Xiaofei Wu},
        title={A New QoE-Driven Video Cache Allocation Scheme for Mobile Cloud Server},
        journal={EAI Endorsed Transactions on Cloud Systems},
        volume={1},
        number={2},
        publisher={EAI},
        journal_a={CS},
        year={2015},
        month={9},
        keywords={quality of experience; mobile cloud computing; video cache allocation},
        doi={10.4108/eai.19-8-2015.2260126}
    }
    
  • Xiaojiang Zhou
    Mengyao Sun
    Yumei Wang
    Xiaofei Wu
    Year: 2015
    A New QoE-Driven Video Cache Allocation Scheme for Mobile Cloud Server
    CS
    EAI
    DOI: 10.4108/eai.19-8-2015.2260126
Xiaojiang Zhou1, Mengyao Sun1, Yumei Wang1,*, Xiaofei Wu1
  • 1: Beijing University of Posts and Telecommunications
*Contact email: ymwang@bupt.edu.cn

Abstract

With the advent of mobile cloud computing, video cache technologies at local cellular networks have attracted extensive attention. Nevertheless, existing video cache allocation schemes mostly made decisions only according to the video coding requirements, without considering users’ individual perception for the video service. In this paper, we propose a new video cache allocation scheme with the consideration of quality of experience (QoE) of users under limited storage space. We make use of the linear regression algorithm to map the relationship between the requested video rate, the replied video rate, the channel condition and the QoE value, which then helps to obtain the different video rates to be stored in the server. Meanwhile, we define the parameter to represent the popularity of a video clip. We optimize the cache space allocation for each video clip based on these parameters in the mobile cloud server of local cellular networks. The experiments demonstrate that the proposed scheme has a better performance in terms of the overall QoE of users with the constraint of the total cache size.