A Localization Correction Algorithm of Location-Based Services Based on Point Clustering

Cell phone location-based services (LBS) is the basis of many smartphone applications, and location accuracy is the most core parameter of location-based services. Currently, LBS can be realized in two ways: GPS positioning and network base station positioning. Among them, the accuracy of network base station positioning is relatively low, and the phenomenon of positioning point drift will occur. Most of the positioning point drift occurs in the stationary state of the positioning object by the network positioning of the random positioning deviation, to the application of LBS-based services have brought about an impact. To address this problem, this paper first uses k-means algorithm to cluster the positioning data points, the positioning points are divided into different sets, to distinguish which are the positioning data when the positioning object is moving, which are the points generated by the drift deviation when it is stationary, and then use the expectation to obtain relatively accurate positioning information, to achieve a certain degree of correction of the positioning drift deviation. In this paper, based on android platform, we call the basic Gaode positioning API to obtain the positioning service, and store the positioning data in the cell phone database through greenDao component as the collected positioning points, in addition to the data sharing function and so on. This paper further implements the k-means algorithm for the collected positioning data, verifies the above ideas, and achieves certain results.