Effective Trajectory Data Storage for Tracking Applications

Geodata is an important component of most software. Coordinates of places and positions of moved objects are specific and different from common data formats. Geographic coordinates are presented in degrees of latitude (in the range of -90 to 90) and longitude (from -180 to 180). Precision of modern positioning sensors may reach up to one meter for lateral plane, which requires to use of up to 6 digits of floating point values of degrees. Thus, storing in memory one point of raw geographic coordinates of one-meter precision in a database requires 29 bits for storing longitude and 28 bits for placement latitude data. In this paper, we are focused on development of an effective algorithm for storing trajectories of finite periods of time, which is a common problem in tracking systems. We consider a compact position report (CPR) algorithm as an adaptive approach for storing geographic data sets of uncompleted trajectories. Effective complete trajectory data could be archived with a Relative Location Algorithm (RLA) which is based on transformation of trajectory data into a sequence of relative distances with initial reference point. Results of the study show the benefits of using RLA for supporting tracking of low-speed objects.