The protection of critical infrastructure objects such as airports, sea ports or border crossings is one of the essential security problems. In order to perform this task an electromagnetic emission passive monitoring method may be used. This is realized with distributed passive wideband radio frequency (RF) sensors utilizing the time difference of arrival (TDoA) method of electromagnetic emission source location. The complexity of the algorithms implemented in such systems requires high computational power of the signal and data processing systems. This implies the need for computational complexity reduction of implemented algorithms. In this article a computationally efficient algorithm for cross-correlation of received signals in time difference of arrival method is presented.
Today’s solutions of reconnaissance and electronic warfare systems require high accuracy in determining the location of the recognized object. It also involves the need to track the trajectory of the object's movement with particular attention to maneuvers and direction changes. The article presents a tracking simulation study of a land object carried out in MATLAB environment for generated various types of trajectories. For each case, the values of root mean square and relative errors are determined. These errors, as well as the evaluation of graphical imaging, allow to assess the usefulness of implemented tracking algorithms using nonlinear Kalman’s filtration.<p> </p> In the next stage, measurements were taken in real terms using a system built from four stationary ultra-wideband (UWB) modules and one mobile built-in land vehicle. The measurement data contained both the raw distance data and the locations obtained from the system tracking application. At this stage, preliminary verification of the designed tracking algorithms was carried out using raw distance data between the modules and the object to create own tracking application. The results of the above operations for different trajectories are presented in the form of appropriate graphical imagery. They confirm the possibility to track the trajectory of a land object using distance measurement from ultra-wideband modules based on the own implemented algorithms. The determination of location is made with high accuracy while improving parameters in relation to factory algorithms at the time of maneuver.
Each modern navigation or localization system designed for ground, water or air objects should provide information on the estimated parameters continuously and as accurately as possible. The implementation of such a process requires the application to operate on non-linear transformations. The defined expectations necessitate the use of nonlinear filtering elements with particular emphasis on the extended Kalman filter. The article presents the simulation research elements of this filter type in the aspect of the possibility of its practical implementation. In the initial phase of the study the conclusion was based on nonlinear one-dimensional model. The possibility of improving the precision of the output through the use of unscented Kalman filters was also analyzed.