Research On Power Consumption Optimization Of Streaming Media For Smart Terminals

In recent years,with the development of wireless communication technology,people can access different types of network services on mobile devices.At present,the proportion of video in global mobile dataflow is about 65%,and users watch videos on the mobile terminals more frequently than before.Such video applications are data and compute intensive.They require long network transmission time and large resource overhead,which including network resources,CPU resources,and battery resources.This paper focuses on the energy consumption of streaming media applications in smart terminals,and researches on the wireless data receiving and processor configuration.Finally,the energy optimization strategy and model are proposed in this dissertation.The main research contents of this article include: 1.Introduction of the research status of energy consumption optimization for mobile devices,and summaries the low-power optimization methods for CPU and wireless data transmission for streaming media applications.In depth study of the composition structure,operation flow and protocol model of the streaming media system.2.Aiming at the problem of wireless data reception for streaming media applications,the relationship between queue stability and energy consumption in video buffer is analyzed.With the help of the improved Lyapunov framework,a low-power data receiving strategy(LDR)which based on Lyapunov queue optimization is proposed.This strategy combines the Lyapunov migration theorem with the change of the video buffer queue and the data receiving rate of wireless network interface under different network conditions,and the change of the queue length of the buffer to select the data receiving mode.Therefore,the low-power data reception of mobile devices in many available networks can be realized,achieving the purpose of enhancing the user experience.3.To solve the problem of CPU configuration for streaming media applications,a CPU power optimization model(VMDP)which based on Markov Decision Process(MDP)is proposed on heterogeneous platform.This model combines the states,actions,transition probabilities and rewards of the Markov decision process with the playback process of the streaming video in the terminal,so as to reasonably allocate Cortex-A15 and Cortex-A7 processors in the Odroid XU3 in the principle of awards maximization to achieve CPU power optimization under heterogeneous multi-core architecture,extending the service time of the terminal.4.The experiments demonstrates that the effectiveness of the above two optimization strategies and models.The experimental results show that the performance of LDR strategy is improved by 23.8% and 75.9% respectively,when compared with other two common strategies.Compared with the heterogeneous multi-processing strategy HMP in the heterogeneous multi-core architecture,VMDP,the total percentage of processors in sleep state is increased by 12.8%.And it can reduce the energy consumption of streaming media video terminals by up to 21%.