Research On Energy Consumption Modeling And Analysis Based On Software Version Evolution

Power consumption estimation of software is the key supporting technology of the low power consumption software optimization study. However, the current software power consumption methods are divided into the measured energy Evaluation Method and method of based on energy consumption model, these two methods the former exist the difficult issues such as the scope of analysis, precision and accuracy are limited, the latter exist the difficult issues such as to establish the efficient and accurate software power consumption model. To solve these problems this paper will combine the method of measured with the method of energy modeling, by actual measured and analyzed the relevant data of software historical versions, from a point of software evolution view to study the association between the software power consumption and software high-level abstract characteristic parameters, to establish the corresponding energy assessment models, dynamically evaluate software power consumption in the same software ecosystem. In this paper, from a point of software evolution view to analyze the historical versions of the software, and find out the trends and patterns of power consumption evolution of high-level software characteristic parameters, and through these laws on power consumption to model software power consumption. This method has better real-time, low cost and so on. Through experimental validation the error estimation and measured values obtained by the method established model is6.74%, to prove that the energy consumption model with high accuracy.