Research And Implementation Of Energy-Efficiency Mechanism For Smartphones

With the rapid development of mobile Internet technology,smartphones are becoming more and more intelligent.And with the proposal of Internet plus,people use smartphones more frequently to some extent.Using smartphones,people can access the information on the Internet,obtain online services such as travel information and so on.The functions are usually achieved on the basic of real sensing and continuous communication,which makes the limitation of smartphones battery become the bottleneck problem.However,the slow development of battery technology can't meet the demand of users for using smartphones.Hence,the research on energy-efficient technology of smartphones becomes the key to solve the problem of smartphones energy consumption.The widely use of all kinds of applications leads to the users' continuous demand for accessed information,whereas the existing energy-efficient solutions are too dump and they only consider the handoff of sensors,sleeping mechanism,computation offloading.The above solutions must set the accurate time to handoff,sleep and offload and the setting of those time need some manual intervention.In the paper,we solve the energy-efficient problem in the communication of application with the server putting the data forecasting users to access first.That will be a new point in energy saving.In the paper,we take two factors into consideration.One is the similarity of data accessed by a group of similar users.The other is the predictability of user's behavior data under some specific condition.Based on the above factors,we propose a data forecasting-based strategy for energy efficiency on smartphones to reduce the energy consumption in the communication with the server.Firstly,we classify the known users into clusters using the mixed variable attributes-based improved K-means algorithm and establish the similar user group.Secondly,we forecast the expectation of user's continuous search using BG/NBD.Thirdly,aiming at valuable users,combined with collaborative filtering recommendation,we predict the data to be accessed by user based on the historical information of similar users.Lastly we utilize the data pre-storage mechanism to pre-store the above predicted data in smartphones in order to achieve the goal of energy saving in the way of reducing the communication times.The primary result illustrated that the strategy we propose can reduce the energy consumption by 13.3%,meet the demand of users and save expected energy in specific scenario.