Design And Implementation Of Mobile Sensing System Based On Hadoop

With the rapid development of mobile Internet, mobile Internet products need tohave a better user experience, services need to be more humane, user data need to bestored reliably and securely, and large amounts of data need to be handled quickly. Toaddress these issues, mobile sensing system based on Hadoop is designed in this paper.The system uses various sensors on the smart phone to perceive a variety of userinformation, and get a better grasp of user needs. The system uses Hadoop as thedistributed storage and computing platform, not only to improve the capability of largeamount of data processing, but also to ensure the reliability and scalability.In this paper, the features and needs of mobile Internet products based on userperception are analyzed. The most important user perception–user location is designedas the basic function of the system. Based on these issues, mobile sensing system basedon Hadoop is overall designed, then the system architecture and network architectureare described. After the overall design of the system, the distributed storage andcomputing layer, the control layer and the service layer of the server are designed andimplemented in detail respectively. After the design and implementation of the server,the latest algorithm called SensLoc which is implemented on smart phones to locateuser is introduced and implemented. Various functions of the client on Android phonesare designed and implemented in detail. Finally, function testing and performancetesting of the system are implemented to obtain test results which are analyzed carefully.