The Research And Implementation Of Android Mobile Client Location Data Optimization Method

In recent years, with the popularity of mobile device and improvement of device’s performance, the demand of location-based services in indoor and outdoor environments is under a variety of increasingly urgent. Because satellite positioning signals of the existing satellite positioning system are effected by buildings or urban canyon area, so it can’t provide accurate position information in such environment. However, positioning based on WiFi RF signals can effectively compensate for this deficiency, because WiFi signals exist extensively in indoor and outdoor environment, especially in indoor. It doesn’t need additional hardware, and can provide accurate location information. WiFi-based positioning algorithm needs to collect fingerprint data in the offline training phase as training data, which will be a lot of labor power and money. Because of environment changing, equipment aging, radio frequency signal characteristics are no longer adapted to the new environment, system data should be maintained by specialized person, though the data is reliable, but it costs too much to do so.Currently, Android devices have been greatly improved in terms of performance which have accounted for a large percentage of the smartphone market. And also, Android has provided excellent platform for developers, which also provides powerful conditions for user feedback mechanism. Take advantage of real-time WiFi fingerprints that users collect to update the fingerprints in DB and users’correction information to improve the speed and accuracy of positioning. This paper takes user feedback into account and uses crowdsourcing model in the following aspects. Positioning parameters namely real-time fingerprint information that user collects are sent to the positioning server. After the position is calculated by using positioning algorithm, the sever analyses the fingerprints to get the APs’situation to decide whether the existing Aps are closed or new Aps are opened. In addition, if user’s current location is the kind that has relatively small fingerprints in DB, the server will send a message to the client to motivate users to participate in sampling. When users find the positioning results don’t match the real position, they can correct their position on the map, they just need to click on the phone screen to choose the correct location, and the position information, fingerprint information are sent to the server-side. Of course, these data is not reliable, there may be a malicious user or user’s careless operation resulting in inaccurate data, this paper will design several data detection algorithms to detect the data user submit, and compare these algorithms by a large number of experiments.Finally, according to the results of the research in this paper, introduce the technology related to systems development, analysis the requirements and design each function module in the system to build Android indoor positioning applications platform.