The Research On Mobile Computation Offloading For Pulse Signal Processing

The use of smart phones and wearable pulse signal acquisition devices for convenient,fast and efficient pulse signal acquisition and processing is of great significance for human health monitoring.However,long-term,large-scale pulse signal acquisition and processing will inevitably consume too much s martphone computing resources and battery capacit y.There fore,how to carry out pulse signal acquisition and processing on smart phones becomes an urgent problem to be solved under the condition of limited computing power and limited battery capacity.This thes is studies the computat io n offloading technology for pulse signal processing.Based on the construction of the computation offloading software and hardware platform,through detailed experimental comparison and data analysis,it is proved that by offloading the pulse signal processing task from the smartphone to the cloud,It can not only reduce the power consumption of local smartphone pulse signal processing but also improve the real-time perfor mance of pulse signal processing,and conclude that the use of two remote cloud and Wi-Fi networks is the best way of computation offloading.On this basis,a computation offloading strategy based on the weight function is proposed,so that users can choose whether to process pulse signals in the local smartphone or the remote cloud according to their needs.The main work of this thesis is as follows:(1)A pulse signal acquisition system is designed.Firstly,the pulse signal of the thumb is collected by the green light reflection pulse sensor and the voltage signal of 0-3V is output through the hardware preprocessing circuit I/V conversion,filtering and amplification,and then the internal A/D converter of the single chip microcomputer is used.The voltage signal is converted into a digital signal,and finally the pulse signal is sent to the smartphone via Bluetooth to provide a data source for pulse signal processing.(2)A computation offloading platform for pulse signal processing is built.The platform mainly includes two parts: the local Android smartphone app and the web cloud application.This paper uses Eclipse,ADT plug-in and Android SDK to develop a pulse signal processing computation offloading app for local Android smartphones.The app has Bluetooth driver module,pulse signal receiving module to be processed,pulse signal storage module,pulse wave display module,The cloud server sett ing module,the pulse signal loading module,the local pulse signal processing module,the pulse signal to the cloud sending module,and the cloud processing post-processing pulse signal module a total of nine modules;the Web cloud pulse signal processing computation offloading program is developed by using MyEclipse and Tomcat,the program is mainly composed of Cloud pulse signal receiving module,cloud pulse signal processing module and sending processed pulse signal to local smartphone module a total of three modules.(3)Two different pulse signal processing computation offloading methods based on local cloud let and remote cloud are studied.Scientific and detailed experimenta l statistics and data analysis were perfor med on the pulse signal processing time and power consumption under different methods.The experimental results show that the computation offloading can reduce the power consumption of local smartphone pulse signal processing and improve the real-time processing.At the same time,the use of two remote cloud and Wi-Fi networks is the best way of computation offloading.On this basis,the computation offloading method was used to study the time reduction and power consumption reduction of pulse signal processing by computation offloading under different pulse data quantities.(4)Two pulse signal processing computat ion offloading decision methods are proposed.Firstly,a non-adaptive computation offloading decision-making method is proposed,which does not consider the user's attention to pulse signal processing time consumption and power consumption.The method can determine whether to proceed by comparing the time consumption and power consumption of pulse signal local processing and cloud processing.Then,a user adaptive computation offload ing decision method is proposed,which can flexibly select pulse signal processing on a local smartphone or a remote cloud server according to time consumption,power consumption and user's attention to these two consumptions.