基于Android系统的能耗计算研究

发布时间：2018-05-25 20:27

本文选题：智能手机 + Android　；参考：《西北大学》2014年硕士论文

【摘要】：在当今社会,随着嵌入式技术的迅速发展,人们对移动智能终端的要求越来越高,许多功能强大、界面丰富的应用程序出现在智能手机上。应用程序的繁荣带来的问题却是智能手机的待机时间越来越短,无法满足用户的日常需要。Android系统的智能手机作为目前用户使用量最大的智能手机,该问题就显得尤为突出。由于电池的发展受到其物理工艺的影响,单从提高电池的续航能力根本无法满足移动智能终端设备对电量的需求。所以,必须考虑软件对移动智能终端设备功耗的影响。然而,现在市面上流行的大多数应用程序都没有进行相应的低功耗处理,主要原因在于没有一款工具能在系统能耗比较低的条件下在线测量智能手机中应用程序的能耗,并且准确的分析出智能手机中哪部分硬件组件在消耗电能。基于上述原因,本文从智能手机的硬件组件出发,研究了基于硬件组件能耗模型,并在此基础上提出基于组件运行时间的能耗模型来估算应用程序的能耗。完成的主要内容如下： 1.对智能移动终端的发展现状和国内外的研究现状进行了分析与总结。尤其是针对Android系统的智能手机,研究分析了Android系统的移动终端在能耗方面所存在的问题,阐述了能耗估算对降低系统能耗的重要性。 2.从智能手机的硬件体系结构出发,深入研究了智能手机中能耗的产生与分布,重点分析了智能手机中能耗影响比较大的组件。 3.针对Android系统的智能手机中能耗影响较大的组件,本文以真机为例,研究分析了组件的工作机制及其在不同工作状态下所产生的功耗,并对每种硬件组件建立了组件功耗模型。 4.针对硬件组件能耗模型在估算能耗时,存在估算时间长和自身能耗比较大的缺点,本文提出了基于组件运行时间的能耗模型。该模型是在硬件组件能耗模型的基础上建立的,文中将组件分为动态组件和静态组件两类,动态组件的能耗获取依赖于组件的平均功耗,而静态组件的能耗则由硬件组件能耗模型获得的。同时,利用组件运行时间的能耗模型为实验真机的组件建立了能耗模型。 5.针对两种不同的能耗估算模型,本文在Android4.1操作系统的智能手机上,分别对4款常用的应用程序进行能耗测量。通过对实验结果分析,我们发现在相同的条件下,基于组件运行时间的能耗模型在测量应用程序能耗时性能要明显优于基于硬件组件能耗模型。
[Abstract]:In today's society, with the rapid development of embedded technology, people have higher and higher requirements for mobile smart terminals, many powerful, interface rich applications appear on smart phones. The problem brought about by the prosperity of applications is that the standby time of smartphones is getting shorter and shorter, which can not meet the daily needs of users. The Android system is the most widely used smartphone, and this problem is especially prominent. Because the development of battery is affected by its physical technology, it is impossible to meet the demand of mobile intelligent terminal equipment by improving the battery's ability of battery life. Therefore, the influence of software on the power consumption of mobile intelligent terminal equipment must be considered. However, most of the popular applications in the market do not have the appropriate low power processing, mainly because there is no tool to measure the energy consumption of smartphone applications online when the power consumption of the system is relatively low. And accurately analyze which parts of the smartphone hardware components consume electricity. Based on the above reasons, this paper studies the energy consumption model based on hardware components of smart phone, and puts forward an energy consumption model based on component running time to estimate the energy consumption of application program. The main elements completed are as follows: 1. The development status of intelligent mobile terminal and domestic and international research status are analyzed and summarized. Especially for the smart phone of Android system, the problems in energy consumption of mobile terminal in Android system are studied and analyzed, and the importance of energy estimation in reducing system energy consumption is expounded. 2. Based on the hardware architecture of smart phone, the generation and distribution of energy consumption in smart phone are deeply studied, and the components which have great influence on energy consumption in smart phone are analyzed. 3. Aiming at the components whose energy consumption is more important in the smart phone of Android system, this paper takes the real machine as an example, studies and analyzes the working mechanism of the component and the power consumption under different working conditions, and establishes the component power consumption model for each hardware component. 4. In order to overcome the disadvantages of long estimation time and large energy consumption in estimating the energy consumption of hardware components, this paper proposes an energy consumption model based on component running time. This model is based on the hardware component energy consumption model. In this paper, the components are divided into dynamic components and static components. The energy consumption of dynamic components depends on the average power consumption of components. The energy consumption of static components is obtained by hardware component energy consumption model. At the same time, the energy consumption model of the real computer is established by using the energy consumption model of the component running time. 5. Aiming at two different energy consumption estimation models, this paper measures the energy consumption of four commonly used applications on the smart phone of Android4.1 operating system. Through the analysis of the experimental results, we find that under the same conditions, the energy consumption model based on the component running time is superior to the hardware component model in measuring the time-consuming performance of the application.
【学位授予单位】：西北大学
【学位级别】：硕士
【学位授予年份】：2014
【分类号】：TN929.53;TP316

【参考文献】