The Design And Implementation Of Crowd Sensing Oriented Distributed Proactive Service Framework

Recently, with the development of hardware technology, various kind of sensors and devices have been appearing, make it possible for computers to understand human’s situation. Scholars start to research on crowd sensing applications based on data from sensors, however these research is always limited within the laboratory. The main reason is crowd sensing relies not only on sensing, but also on the infrastructure it runs on. Devices and services’ compatibility, as well as the pressure of large scale accessing, must be taken into consideration to realize crowd sensing, the study in this area is still immature.In this paper we study the problems encountered by the crowd sensing oriented service framework in distributed environment, analyze the possibility of realizing compatibility for heterogeneous sensors through message based Web Service. Study the way to grant data source stability in unpredictable environment with the help of device manager. Analyze the need for devices to support large scale accessing in distributed environment, and proposed a solution of using context data server as transfer stop. Provide the algorithm for implementing high availability for server clusters referring to the mainstream solution.Aiming at its application in real life, this paper proposed a distributed crowd sensing framework based on context data servers called DASF. Hardware device layer, context data layer and smart service layer are introduced, as well as their cooperative progresses.Based on some open source software, this paper implements a DASF framework, through several experiments, DASF’s stability and performance are testified, proving its availability for distributed crowd sensing application. In the end, a simple entrance guard service is introduced to prove DASF’s compatibility and proactive serving capacity.