Data mining classification of cars based on GPS shadows in Forward Scatter Radar systems

Comparative Analysis of Two GPS Forward Scattering Systems For Cars Parameter Estimation

“…The Software-Defined GPS receiver contains the Acquisition block to identify satellites, and the Code & Carrier Tracking block to form the navigation message. Next, the obtained navigation message is integrated every hundreds milliseconds in order to form the radio shadow of the object [11]. To obtain the results described in this article, the programs for GPS signal processing (SoftGNSS) have been optimized and accelerated by using the Matlab capabilities to parallel data processing to maximize the load on modern multi-core computers.…”

“…The algorithm proposed in [10] is used for estimation of the received GPS shadow parameters (duration, depth, energy and power). The obtained GPS shadow parameters can then be used to classify moving objects (people, cars, buses, airplanes), for example, through the Data Mining approach [11].…”

“…After the preliminary analysis of data by semi-automatic processing of the recorded GPS signals, the operator selects the starting and ending points of the car shadows on the basis of visualization of the GPS signals from the cars. Then the estimates of the parameters (duration, depth, energy and power) of the resulting GPS shadows are automatically calculated based on the algorithm proposed in the article [11]. The thus obtained parameters of the GPS shadows of the two cars are represented in the form an Excel-table for the subsequent generalized statistical processing [11,12].…”

“…Then the estimates of the parameters (duration, depth, energy and power) of the resulting GPS shadows are automatically calculated based on the algorithm proposed in the article [11]. The thus obtained parameters of the GPS shadows of the two cars are represented in the form an Excel-table for the subsequent generalized statistical processing [11,12]. The preliminary rough assessment of the GPS shadow parameters of the car recorded by the two different GPS signal recording systems (GPS receiver 1 -is the GPS signal recording system developed by the Colorado University and GPS receiver -2 is the GPS signal recording system Stereo dual GNSS front-end) can be done by comparing their statistical parameters (Mean and STD).…”

“…A possible algorithm for target detection using GPS L5-based FSR system is described in [6], and the detection probability characteristics are analytically calculated in [7]. The next few articles [5][6][7][8][9][10][11][12] are devoted to experimental measurements made by using GPS L1-based FSR system and the Software-Defined GPS receiver, developed by the Aerospace Department in the University of Colorado [13], allowing to observe the geometric shadows (signal blocking) of ground objects of different sizes, mobile and stationary. The obtained recordings are not available for processing in the field conditions because the processing of radio shadows of objects requires the additional time in the laboratory, which further impedes research.…”