Research On Energy Consumption Optimization Of Mobile Browser Based On Webpage Characteristics

The development speed of the mobile Internet has greatly exceeded people's imagination,and it has profoundly changed the behaviors and habits of individuals,enterprises,and governments.The development of smart phones has played an important role in the entire mobile Internet tide.In recent years,with the rapid improvement of smartphone performance,the expansion of 4G coverage,the continuous maturation and gradual commercialization of 5G technology,a large wave of new industrial chains has emerged rapidly,and cultural entertainment such as short video,e-reading,and mobile games are gaining momentum.It quickly occupied the market,and the one-stop platform for mobile browsers brought together these new content,new forms,and new functions,reducing the number of APP installations,bringing users a smoother operating experience and immersive consumption scenariosHowever,with the faster and faster processor speed,larger and larger memory capacity,and higher and higher screen resolutions of smartphones,the development of battery technology has led to the endurance of smartphones due to the limitations of their physical processes.The time is getting shorter and shorter,which seriously affects the user experience and has become the bottleneck of the development of smart phones In response to the above problems,this article takes the Odroid development board as the experimental platform and the WebKit kernel-based Chrome browser as the research object,and explores the optimization of the browser's energy consumption without affecting the user experience,thereby increasing the battery life of the smartphone.First of all,this article analyzes the webpage characteristics of the top 1000 different industry types of websites on the comprehensive ranking on Aleax,and then surveys the use experience of 500 users on different types of websites through online questionnaires.It is found that 90%of users Film and television websites will take the loading completion time of pictures as the evaluation index of a good user experience;and 82%of users will use text loading completion time as an indicator of a good user experience for news and information websites.Through the above two surveys and analysis,it is concluded that users have different experience standards under different industry types of websites due to their different focus issues.Then experimentally explored the relationship between browser power consumption and CPU frequency,web page structure characteristics,and time to first screen.Combining theoretical knowledge of the WebKit kernel and analysis of experimental results,we found that under certain network conditions,CPU frequency exists.A threshold.When the CPU frequency exceeds this threshold,the rendering time of the webpage is almost no longer reduced.On the contrary,an excessively high CPU frequency will cause more energy loss.In this article,this threshold is defined as the optimal CPU frequency..Based on the above experimental results,this paper proposes a web-aware scheduling algorithm and establishes an energy consumption optimization model based on machine learning to intelligently schedule web pages using different CPU frequency configurations based on the current web page characteristics.For unknown webpages,first determine the website industry type based on the keywords of the meta tag,then extract the webpage characteristics,predict the optimal CPU frequency based on the energy consumption model of the corresponding industry type,and finally schedule by odroid-cpu-control Experimental results show that compared with the system's default CPU scheduling strategy,the proposed energy consumption optimization model can effectively reduce the energy consumption of the mobile terminal without affecting the user's experience,and achieve an ideal compromise between high performance and low energy consumption.