Clutter Cancellation Methods for Small Target Detection Using High-Resolution W-band Radar

Drones are currently being used for various applications. However, the detection of drones for defense or security purposes has become problematic because of the use of plastic materials and the small size of these drones. Any drone can be placed under surveillance to accurately determine its position by collecting high-resolution data using various detectors such as the radar system proposed in this paper. The W-band radar has a high carrier frequency, which makes it easy to design a wide bandwidth system, and the wideband FMCW signal is suitable for creating high resolution images from a distance. Unfortunately, the huge amounts of data gathered in this way also contain clutter (such as background data and noise) that is usually generated from unstable radar systems and complex environmental factors, and which frequently gives rise to distorted data. Accurate extraction of the position of the target from this big data requires the clutter to be suppressed and canceled, but conventional clutter cancellation methods are not suitable. Four clutter cancellation algorithms are assessed and compared: standard deviation, adaptive least mean squares (LMS), recursive least squares (RLS), and the proposed LMS. The proposed LMS has combined LMS with the standard deviation method. First, the big data pertaining to the target position is collected using the W-band radar system. Subsequently, the target position is calculated by applying these algorithms. The performance of the proposed algorithms is measured and compared to that of the other three algorithms by conducting outdoor experiments.