Optimizing Power Consumption Of Mobile Devices For Video Streaming Over 4G LTE Networks

In recent several years,the mobile industry has been growing rapidly.Online video service as one of the most important big data applications has received much at-tention in mobile industry.In 2016 video dominates the mobile traffic and takes 50%of the total traffic(around 4.25 ExaBytes).According to GSMA,video will account for around 75%of mobile data traffic(around 34.5 ExaBytes)in 2022.In addition,mobile devices are arousing the increasing demand of video streaming,since users can experience video streaming services on a mobile broadband network with different mobile devices(e.g.,mobile tablets,smartphones).Video streaming,one of the most popular technologies for online video playback,has already been applied to 4G LTE networks.Previous work has been devoted to understanding the power consumption in general 4G LTE networks,while it is still unclear how the online video streaming makes impact on the power performance of mobile devices.Inspired by this,this dis-sertation investigates the relationship between the mobile device's power performance characteristics and the behaviours of video streaming in 4G LTE networks.This dis-sertation formulates the energy models together with an algorithm that can assist the analysis.Particularly,this dissertation designs a power analysis system,and conduct a comprehensive and also deep analysis on the power consumption of video streaming in 4G LTE networks,for the saving room,the number of RRC tails and the transmission pattern.In addition,to improve the power consumption of mobile devices for online video streaming,this dissertation studied the following problems:(1)How to control the buffer size efficiently;(2)How to merge the video segment fetching sessions wise-ly;and(3)How to reduce the power consumption when users' behavior modes are unavailable beforehand.This dissertation suggests the power saving methods.The central idea of the proposed methods is to exploit the continuous video watching time and the capacity of batteries to predict users' behavior modes and user demand,and then dynamically adjust corresponding parameters,so as to consume less energy.This dissertation made the following main contributions.1.This dissertation formulates the energy models,which characterize various s-cenarios.The models could be used to understand other features related to the energy consumption and video streaming.This dissertation also designs a pow-er analysis system that allows us to conduct a deep study on the power perfor-mance characteristics of video streaming in 4G LTE networks.Based on the energy models,this dissertation develops an algorithm that is used to assist us to roughly understand the energy saving room and some other useful information.2.This dissertation conducts a comprehensive analysis based on the proposed plat-form.The experiments reveal us a series of power-related findings—the saving room in the network part,the number of RRC tails and the transmission pattern could be promising for optimizing the power consumption of video streaming in 4G LTE networks.3.This dissertation proposes a user behavior self-adaptive method to save power consumption.This dissertation gives the rigorous theoretical analysis for the proposed method.The analysis verifies the feasibility and effectiveness of the presented method,from the theoretical perspective.This dissertation also con-ducts extensive experiments to demonstrate the effectiveness of the proposed method.Compared against other competitors,the performance of the proposed method is closest to that of the "ideal" prototype.4.This dissertation further develops a battery-aware method.According to dif-ferent capacities of the mobile batteries,the proposed method can aggressively adjust the video playing scheme to save the power consumption.Also,the theo-retical analysis and extensive experiments show the effectiveness of this method,comparing with other competitors.