Android-based Low-power Optimization For Embedded System

Today the information technology is developing rapidly, with the fusion and rapid development of integrated circuit technology and mobile internet technology, the multicore processors mobile phone based on the android embedded system has become more and more popular, and it becomes the indispensable electronic product in people’s daily life. However, due to the limit of cell phone size, weight, heat and other factors, the standby time of Android smartphone often can not satisfy people’s need. Therefore, the power optimization of multicore processors smartphone based on android embedded system, improvement of standby time is becoming a hot topic issue.This thesis optimizes the power consumption of smartphone which is based on android embedded system, by optimizing power consumption of the multicore processors which are at the bottom of the android embedded system.Firstly this thesis introduces the development of integrated circuit and the power consumption of CMOS integrated circuit. Since the multicore processors based on mobile embedded system are constituted by CMOS transistor circuit, this thesis then describes the theory of static power consumption and dynamic power consumption of the CMOS transistor circuit. Three key technologies of optimizing the power consumption of multicore processors are described:dynamic voltage frequency scaling technology (Dynamic Voltage and Frequency Scaling, DVFS), dynamic core scaling technology (Dynamic Core Scaling, DCS) and dynamic power management technology (Dynamic Power Management, DPM). This thesis uses them to reduce the power consumption of the android smartphone system. DVFS technology reduces and optimizes dynamic power mainly by reducing the supply voltage and operating frequency. DCS technology is mainly based on the need to adjust the working core’s number of multicore processors. After determining the number of working core, the design optimizes the power consumption of cores by DPM technology. DPM technology optimizes the static power consumption of system mainly by shutting down the cores or placing them into a sleep state. DCS technology and DPM technology are often used together.Secondly, in this thesis, the power consumption of multicore processors is analyzed theoretically. Through analyzing the power consumption factors of multicore processors, a simple model of three-dimensional space discrete points is established. It simplifies the complex issues to deal with, and gets the theoretical basis of low-power optimization of multicore processor through a simple theoretical derivation. Thereby facilitating power optimization and guiding power optimization direction.Finally, this thesis presents an optimized and fine grained DVFS algorithm, a simple flow chart of low power consumption for mobile terminal based on android embedded system is proposed. Android programs are written to respectively realize dynamic voltage frequency scaling (Dynamic Voltage and Frequency Scaling, DVFS) technology, dynamic core scaling (Dynamic Core Scaling, DCS) technology and dynamic power management (Dynamic Power Management, DPM) technology。Through experiments, this thesis uses a combination of three power optimization technologies to achieve the optimization based on android embedded system power consumption. Through experiments and measurements, after comparing the joint application of three kinds of technologies to optimize the power consumption with no power consumption optimization, system’s power consumption which was optimized can reduce 8.59 percent, and it improves the cruising ability of smartphone.