Anti-drone radars are mainly designed to accurately monitor low-altitude airspace below 1,000 meters Above Ground Level (AGL). By integrating dedicated signal processing modules and high-gain antennas, they can efficiently capture clutter signals generated by ground objects, aerial targets, and various environmental interferences (see Figure 2), providing high-quality basic data support for subsequent target identification, trajectory tracking, and countermeasure decision-making. According to the general airspace classification standards in the aviation field, airspace below 1,000 meters is clearly defined as low altitude, among which the range below 100 meters is ultra-low altitude. Affected by factors such as terrain occlusion and building reflections, this area has more complex environmental clutter. Meanwhile, it matches the endurance and operational needs of small drones, thus becoming a primary activity scenario for consumer aerial photography drones, industrial inspection drones, and even some maliciously used drones. Taking pulse Doppler radar, the most widely used and technologically mature radar in the current anti-drone field, as an example, the typical Low, Slow, and Small (LSS) characteristics of drones will significantly restrict the detection accuracy, continuous stability, and anti-interference capability of radar systems from multiple dimensions, including signal strength, movement trajectory, Radar Cross Section (RCS), and flight attitude stability (as shown in Figure 3). This is also a core technical challenge that needs to be prioritized in the design, research and development, and performance optimization of anti-drone radars.
Firstly, the core characteristic of drones—"low-altitude flight"—places strict requirements on the multi-scenario adaptability and target discrimination capability of anti-drone radars. They need to accurately identify various moving targets on the ground, in low altitude, and ultra-low altitude in different complex terrains and environments such as urban buildings, mountainous hills, and open areas, covering pedestrians, ground motor vehicles, migrating bird flocks, as well as drones of different sizes and flight modes (e.g., multi-rotor, fixed-wing, and vertical take-off and landing). To reduce the interference of ground clutter (such as building wall reflections, terrain undulation interference, and ground vegetation scattering) on detection results, some anti-drone radars adopt an optimization strategy of dynamically adjusting the pitch angle. By real-time changing the irradiation direction, coverage angle, and energy distribution of radar beams, they actively avoid areas with concentrated ground clutter and improve the signal-to-noise ratio of target signals. However, this passive avoidance method has obvious technical limitations and is prone to a high "false negative rate" in drone detection. Since the conventional operational airspace of most consumer and industrial small drones is concentrated below 100 meters (ultra-low altitude), radar beams can hardly achieve non-dead-angle coverage of this area after adjusting the pitch angle. Especially in complex terrains such as high-density urban buildings and mountain gullies, occlusion blind spots are further expanded, and the risk of false negatives increases significantly. Therefore, an efficient and reliable anti-drone radar system must be equipped with mature Automatic Target Recognition (ATR) capability. Through deep learning algorithms, it extracts, classifies, and verifies captured signals, accurately distinguishing drone targets from clutter, birds, and other interference sources, fundamentally reducing the risks of false negatives and false positives, and ensuring the reliability of detection results.
Secondly, the inherent characteristic of drones—"small size"—results in an extremely low Radar Cross Section (RCS). The RCS value of most small drones, especially consumer multi-rotor drones, is only 0.01-0.1 square meters, much lower than that of traditional aircraft such as fighter jets and helicopters. The radar signals reflected by them are weak and easily masked by environmental clutter and electromagnetic interference, posing great challenges to signal capture. This characteristic places extremely high requirements on the detection sensitivity of radar detectors, which need to have strong capabilities in weak signal extraction, amplification, and filtering. While effectively filtering electromagnetic interference and environmental clutter, they must also cover a wide detection range to achieve the dual performance goals of "long-distance detection and short-distance precise positioning". The realization of this core performance goal must be based on high detection and recognition credibility, requiring the construction of a "hardware + algorithm" collaborative system through multi-dimensional technical optimization. At the hardware level, upgrade core components such as high-sensitivity antennas and low-noise receivers to improve signal reception and conversion efficiency. At the algorithm level, introduce advanced technologies such as adaptive filtering, pulse compression, and Constant False Alarm Rate (CFAR) detection to enhance the recognition capability of weak target signals. This ensures the accurate capture, feature recognition, and stable locking of weak target signals, avoiding the impact of signal misjudgment and missed judgment on the disposal efficiency and accuracy of subsequent countermeasure links, and meeting the needs of practical application scenarios.
Finally, the characteristic of drones—"slow flight speed"—also poses considerable challenges to the stable tracking function of radar systems. The flight speed of most small drones ranges from 10 to 50 kilometers per hour, and some drones operating in low-altitude hovering have a speed close to zero. In this low-speed flight state, their motion characteristics are barely distinguishable from those of interfering targets such as floating clutter, slow-flying birds, and falling objects. Traditional tracking algorithms can hardly achieve effective discrimination through speed differences, which not only fails to continuously and stably lock drone targets but also may mislead the judgment of auxiliary sensors such as optical and infrared sensors, leading to data deviations and decision-making errors in multi-sensor fusion systems. Such deviations will further transmit to the countermeasure units in Counter-Unmanned Aircraft System (C-UAS) solutions, such as directional jamming equipment, physical interception devices, and laser countermeasure systems, resulting in delayed countermeasure actions and insufficient accuracy, failing to intercept target drones in a timely and effective manner, and even possibly causing disturbance to surrounding innocent targets. To address this issue, radar systems need to have high scan update rates and fast target recognition capabilities. By increasing the beam scanning frequency, optimizing dynamic tracking algorithms and target trajectory prediction models, they can real-time update target motion parameters (speed, trajectory, attitude, flight trend), quickly distinguish low-speed drones from various interfering targets, and provide real-time, accurate, and continuous target data support for subsequent countermeasure units. This ensures the accuracy and timeliness of tracking and countermeasure links, fully meeting the rapid disposal needs of practical scenarios such as security, military, and event protection.
