In this paper some capabilities of field programmable gate arrays (FPGA) as a suitable platform for radar applications based on an exemplary implementation of matched filter are discussed. Relations between the resource usage, precision of calculations and design performance are presented. Results of a multi-aspect analysis can help in elaboration of specific implementation conclusions both at the stage of formulating processing algorithms and their implementation. A wider view of the possibilities and limitations of programmable logic in combination with the features of other available platforms leads to the effective use of modern heterogeneous systems, including FPGA, GPU and CPU, which in turn allows to take advantage of the possibilities and compensation for the limitations of those technologies.
Classic methods of detection of objects assume a certain threshold, that separates the echo signal of the object from noise. In case of objects with high linear velocity their echoes may be spread in time and can be under the threshold. The reduction of the threshold should improve a probability of detection, but on the other hand it will increase the probability of false alarm. This side effect can be reduced using special methods usually referred to as Track-before-Detect. Their principle of operation base on the analysis of object’s state estimation in several subsequent scans of sensors’ observation space. Subsequently the estimates of the object position are combined in accordance to assumed strategy. If the parameter of association cost achieves the threshold, it means that an object is detected. There are many TBD strategies described in the literature, but two of the most popular are the Multiple Hypothesis Tracking and the Dynamic Programming. The paper presents their principles of operation and a comparison of their effectiveness.
This article presents CUDA architecture as an effective tool for the digital beam forming in radar system. The article contains the results of a series of tests, which verify the fulfillment of stability and data processing time requirements. In the article, the authors presents the results of implementations that illustrate different approaches to this issue and some methods to increase the efficiency of implemented algorithms. In addition, the presented results represent implementations on devices adapted for military applications.
There are known only few ballistic object tracking algorithms. To develop such algorithms and to its further testing, it is necessary to implement possibly simple and reliable objects’ dynamics model. The article presents the dynamics’ model of a tactical ballistic missile (TBM) including the three stages of flight: the boost stage and two passive stages – the ascending one and the descending one. Additionally, the procedure of transformation from the local coordinate system to the polar-radar oriented and the global is presented. The prepared theoretical data may be used to determine the tracking algorithm parameters and to its further verification.
In this paper we present the concept of multiple sensors data acquisition from onboard of an Unmanned Air
Vehicule (UAV). Because of flight instabilities caused by atmosferic movements (winds, thermals etc..) it is necessary to
apply active stabilization in order to obtain reliable readings from observation sensors. The most stabilization-demanding
sensor is Synthetic Aperture Radar (SAR) and in this paper two methods of stabilization are presented: hybrid (electromechanical)
Space-Time Adaptive Processing (STAP) is a well known technique used for dealing with clutter in order to detect
moving targets. This technique was derived under assumption, that clutter has Gaussian characteristics.
Unfortunately when dealing with sea clutter, Gaussian assumption is no longer valid . This causes increased
number of false alarms. In this paper we present improved detector to deal with non-Gaussian clutter. Detector was
derived from Generalized Likelihood Ratio Test (GLRT), assuming Spherically Invariant Random Process (SIRP) as
a model for the clutter. Resulting detector was named Two Dirac-Deltas (TDD) detector and it has additional
parameter (Δ) in comparison to classical STAP. Based on simulations, it is shown that it is crucial to choose Δ