Minimization technique for the amount of computation for land clutter simulation
Abstract
Introduction: Land clutter simulation is an important step in testing and evaluating the performance of radar stations. A surface simulation model is an area within the radar field of view divided into many small sites. Each of these sites is a point reflector. The echo signal from the surface area is a set of signals from the point reflectors this area contains, which requires significant computing resources. Purpose: To substantiate the techniques of minimizing the amount of computation when calculating reflected signals for the simulation of radar station operation. Results: The land clutter modeling requires the calculation of the convolution of the sequence describing the structure of the earth's surface and the sequence describing the probing signal. The calculation of the convolution is efficient according to the overlap-save method, where the entire sequence is divided into several sections. The study shows that when computing partial convolutions with the use of the Fourier transform, some harmonics have low energy and can be ignored without significantly affecting the accuracy of the calculations. The land clutter properties change depending on the distance of the particular area from the radar station. In an area which close to the radar station, the contribution of reflections from areas is highly dependent on their location. Reflections actually are formed only by nearby areas. We obtain the expressions for estimating the errors that arise due to the neglect of reflections from remote areas. At a considerable distance from the radar, the reflective areas take on a point character, which makes it possible to calculate the convolution without using fast algorithms. Practical relevance: The presented computational optimization techniques can be used to develop echo signal simulators.
