Research And Design Of Data Services Technology For The Internet Of Things

Data service is one of the most basic and important functions of the IoT (Internet of Things),via which, people can get access to the data of state of human or objects whenever and wherever itis. Then corresponding decisions can be made according to these data. However, various kinds ofsensing devices result in problems in data service of the IoT, such as great amount of dataacquisition, types undetermined, variety in ways to appear, difficulties to get disposed andanalyzed. Aiming at overcoming these problems, an overall structure of data service and a plan oftransition for the integrity of data are proposed for the IoT. Besides, a multi-threaded schedulingstrategy for the technology of data service is designed for the IoT. And also, a new algorithm ofdata processing and a hierarchical storage structure based on the fast sort algorithm are putforward. All of these make the data collected by all kinds of sensing devices of the IoT getdisposed correspondingly and stored according to its style. What’s more, the performance of thedata service of the IoT is promoted efficiently. The research of this paper will lay the foundationfor the further work of the data service technology of the IoT. The main contribution and creativityof this paper are as below:1. An overall architecture of data service that is oriented to the Iot is presented. The applicantneeds of the overall architecture of the data services for the IoT is analyzed. By using thisarchitecture, the characteristics of the entire system of the IoT like real-time, stability andmaintainability can be enhanced.2. A multi-threaded scheduling strategy for the technology of data service that is oriented tothe Iot is designed. With the combination of the state machine mechanism, the real-timeperformance of the system is improved by using the hybrid scheduling strategy. A basic idea thatthe multi-threaded scheduling mixes with the state machine mechanism to form the schedulingstrategy is proposed. Various states among the master and slave station model is described basedon the decision model of Markov. The transition probability for all kinds of state is analyzed. Aswith the mixed scheduling strategy, according to the multi-threaded scheduling, the primary threads to execute is selected. And then, the selected threads execute the specific scheduling ofstate machine choosing. As a consequence, the scheduling time of choosing threads later isreduced. The scheduling efficiency is improved as well.3. A new kind of data processing algorithm and a hierarchical storage structure based on thefast sort are put forward. And the data storage system based on the improved and time-sensitivealgorithm of fast sort is realized. Storing data in storage devices comes true. With the form ofsecondary level directory, data is stored according to its classification. In addition, the files thatstore data are named after the dynamic time, so that the data can be queried easily. The system isnot only functional completely, but also fast to get access, convenient and simple and low cost.Finally, gived the situation that the technology of data service is applied to the subwayvehicle system, the scheduling time is analyzed. In the view of the applicantion effects, thereal-time performance of the hybrid scheduling strategy improved more than that of otherscheduling strategy. In terms of data storage, data can be very intuitive to be viewed and analyzed.