A drone detection, identification and location system and method may illuminate a target with one or multiple selected radio-frequency (RF) carrier frequencies. Both direct emissions received from the target and re-emissions generated by the target may be processed to determine whether the target is a drone. The re-emissions may be generated by circuitry of the target resulting from the illumination with the one or multiple RF carrier frequencies. The re-emissions may comprise cross-modulation products (CMPs) including forced non-linear emissions (FNLEs). The direct emissions and the re-emissions may be processed to generate an RF signature for the target. The target may be determined to be drone and the type of drone may be identified based on the RF signature.

A radar device includes a first radar and a second radar that are arranged at positions separated from each other, and of which detection ranges are at least partially overlapped; and a detection unit that detects at least one of a moving direction and a velocity vector of a reflection point existing in an overlapped portion of the detection ranges, based on a first detection result of the first radar and a second detection result of the second radar.

A target detection method includes a radar apparatus that performs at least one interference sensing on a plurality of first time domain resources and determines a second time-frequency resource used for target detection in a first time-frequency resource based on a result of the at least one interference sensing. The first time domain resources are a subset of time domain resources corresponding to a time-frequency resource of a first detection apparatus, and the first time-frequency resource is the time-frequency resource of the first detection apparatus.

A target detection method includes a radar apparatus that performs at least one interference sensing on a plurality of first time domain resources and determines a second time-frequency resource used for target detection in a first time-frequency resource based on a result of the at least one interference sensing. The first time domain resources are a subset of time domain resources corresponding to a time-frequency resource of a first detection apparatus, and the first time-frequency resource is the time-frequency resource of the first detection apparatus.

A method, apparatus and computer program product are provided to generate a model of one or more objects relative to a vehicle. In the context of a method, radar information is received in the form of in-phase quadrature (IQ) data and the IQ data is converted to one or more first range-doppler maps. The method further includes evaluating the one or more first range-doppler maps with a machine learning model to generate the model that captures the detection of the one or more objects relative to the vehicle. A corresponding apparatus and computer program product are also provided.

The present application provides techniques for reducing noise in sensor-based systems, such as radar systems. In particular, techniques referred to background supplemental cancellation (BaSC) and background supplemental loading (BaSL) are disclosed and facilitate improved detection of moving targets in certain types of radar systems, such as radar systems based on Reiterative minimum-mean square error (RMMSE) estimation formulations. The BaSC technique may utilize a hard cancellation, where clutter cancellation is performed prior to estimation, while the BaSL technique may utilize a “soft” cancellation technique whereby clutter cancellation is performed jointly with estimation. The clutter cancellation provided via the BaSC and BaSL techniques improves the accuracy of the radar system with respect to performing target detection.

Systems and methods for detection and reporting of small targets to an operational area. Exemplary embodiments are presented to detect targets such as avian species, UAS, UAV, and drones, and transmit unique small target identifier information via data link, such as ADS-B, to an operational area.

Disclosed are a method and a device for measuring a biometric signal by using a radar. The disclosed method measures a plurality of biometric signals by using a radar by: (a) receiving the plurality of biometric signals from the radar; (b) calculating distance information of the received plurality of biometric signals and classifying the same on the basis of a distance; (c) selecting a signal having a largest variance according to a time; (d) further selecting a number of signals among signals having a distance with the signal selected in the step (c) smaller than an arbitrary distance from the distance-based classified signals; (e) converting all signals selected from a time domain to a frequency domain; (f) calculating a reliability of each biometric signal from the converted distance-based signals; and (g) detecting a corresponding biometric signal by selecting the distance-based signal where the calculated reliability is highest.

An apparatus for detecting a target is disclosed. The apparatus of detecting a target includes: a frequency mixer configured to calculate a first beat frequency based on a transmitted signal and a received signal of first scanning and calculate a second beat frequency based on a transmitted signal and a received signal of second scanning performed with a predetermined time interval from the first scanning; a controller configured to extract a first moving component by comparing an up-chirp period frequency and a down-chirp period frequency of at least one of the first beat frequency or the second beat frequency; extract a second moving component by comparing up-chirp period frequencies or down-chirp period frequencies of the first beat frequency and the second beat frequency; and determine the moving target based on the first moving component and the second moving component.

In various examples, a deep neural network(s) (e.g., a convolutional neural network) may be trained to detect moving and stationary obstacles from RADAR data of a three dimensional (3D) space. In some embodiments, ground truth training data for the neural network(s) may be generated from LIDAR data. More specifically, a scene may be observed with RADAR and LIDAR sensors to collect RADAR data and LIDAR data for a particular time slice. The RADAR data may be used for input training data, and the LIDAR data associated with the same or closest time slice as the RADAR data may be annotated with ground truth labels identifying objects to be detected. The LIDAR labels may be propagated to the RADAR data, and LIDAR labels containing less than some threshold number of RADAR detections may be omitted. The (remaining) LIDAR labels may be used to generate ground truth data.