The appearance of small UAVs on the commercial market poses a real threat to both civilian safety and to military operations. In open terrain a radar can detect and track even small UAVs at long distances. In an urban environment with limited line-of-sight and strong static and non-static background, this capability can be severely reduced. The radar cross section of these UAVs are normally small compared to the background. However, the rotors of the UAVs produce a characteristic micro-Doppler signature that can be exploited for detection and classification. In this paper, we investigate in an experimental set-up whether it is possible in the radar non-line-of-sight to retrieve the micro-Doppler signature of the UAV rotors. This is done by exploring up to three multipath bounces in the measured signal. The measurements were made with a semi-monostatic single receiver-transmitter radar system operating at X-band in a pulsed single frequency mode. The radar response of the UAV, with plastic and metallic rotors, was measured at several positions inside a 4 m wide corridor with metallic walls. In this paper, data from one line-of-sight and two non-line-ofsight positions are presented. Results show that we are able to detect the micro-Doppler of the rotors and to retrieve the number of revolutions per minute, for both rotor types. Free space Finite-Difference Time-Domain calculations have also been performed on a CAD-model of the UAV rotor to determine the optimal choice of polarization and the short-time Fourier transform filter length.
The possibility of detecting moving people around the corner using multipath propagation of radar waves has previous been demonstrated in experimental set-ups. Here, we present measurements from a realistic scene using a radar system operating at X-band in a stepped-frequency mode. The moving objects include person(s), bicycle, and car(s). A semi-monostatic single receiver-transmitter radar system was used as a data collector. All detections were made by using Doppler filter processing.
We can identify target returns after one and two wall reflections. Two persons moving in the scene can be separated, when allowed by the range resolution.
Fully polarimetric through-the-wall radar measurements with high spatial resolution have been attained by using the
ISAR (Inverse Synthetic Aperture Radar) technique. Polarimetric methods may reduce the effects of the wall interaction
and increase the contrast between humans and the background.
The main scene in the measurements was a human sitting in a small wooden cabin. The cabin was placed on a turntable
and rotated, to obtain ISAR imaging. By switching the transmitter and receiver antennas between horizontal and vertical
polarizations, four polarization combinations were obtained. Phase coherence was maintained through a whole
measurement series. This enabled co-processing of the whole collected data set with coherent methods. A statistical
description of the measured data was used, with the polarimetric coherency matrix applied to the received signals.
ISAR images produced for the TTW scenes show that the human can be discerned from the background. The contrast
between the human and the background was found to be greater with vertical polarization at transmit and receive, with
less contrast using cross-polarization or horizontal co-polarization, due to the horizontal wall grain orientation.
A classification scheme based on the eigenparameters of the coherency matrix (entropy, anisotropy and alpha angle) and
the backscatter power has been tested to discriminate between different target objects in the cabin. The method shows
some promise, but a reliable classification has not yet been attained.
An approach toward real-time for radar through-the-wall (TTW) sensing is presented. The aim is to detect and classify
human motions behind walls. In the future a system could support an operator with information on the activity in a
closed room, e.g., in a hostage situation.
To meet this objective, radar TTW measurements have been performed on moving person(s) inside a closed room and an
MTI-based signal processing algorithm, using coherent subtraction between different frequency sweeps, has been
The radar was equipped with ridge horn antennas which were directed toward an outer door. For each measurement the
radar frequency was swept between 5 and 10 GHz 1893 times, with a sweep sampling rate of ~94 Hz.
A crucial algorithm parameter is the time distance between two subtracting sweeps, or the sweep difference. Fast and
slow motions are captured by using separate sweep differences.
Applying the algorithm to the TTW data, we find that human motions behind a door can be detected with the background
well suppressed. Even a person who is standing still without breathing is fairly easy to detect. The results are promising
with low false alarm rates and fast signal processing rates, enabling real-time operation capability.
Detection of moving objects around corners, with no direct line-of-sight to the objects, is demonstrated in experiments using
a coherent test-range radar. A setting was built up on the test-range ground consisting of two perpendicular wall sections
forming a corner, with an opposite wall, intended to mimic a street scenario on a reduced scale. Two different wall materials
were used, viz. light concrete and metallic walls. The latter choice served as reference, with elimination of transmission
through the walls, e.g. facilitating comparison with theoretical calculations. Standard radar reflectors were used as one kind
of target objects, in horizontal, circular movement, produced by a turntable. A human formed a second target, both walking
and at standstill with micro-Doppler movements of body parts. The radar signal was produced by frequency stepping of a
gated CW (Continuous Wave) waveform over a bandwidth of 2 or 4 GHz, between 8.5 and 12.5 GHz. Standard Doppler
signal processing has been applied, consisting of a double FFT. The first of these produced "range profiles", on which the
second FFT was applied for specific range gates, which resulted in Doppler frequency spectra, used for the detection. The
reference reflectors as well as the human could be detected in this scenario. The target detections were achieved both in the
wave component having undergone specular reflection in the opposite wall (strongest) as well as the diffracted component
around the corner. Time-frequency analysis using Short Time Fourier Transform technique brought out micro-Doppler
components in the signature of a walking human.
These experiments have been complemented with theoretical field calculations and separate reflection measurements of
common building materials.
This paper describes the research efforts made at the Swedish Defence Research Agency (FOI) concerning through-the-wall imaging radar, as well as fundamental characterization of various wall materials. These activities are a part of two FOI-projects concerning security sensors in the aspects of Military Operations in Urban Terrain (MOUT) and Homeland Defence.
Through-the-wall high resolution imaging of a human between 28-40 GHz has been performed at FOI. The UWB radar that was used is normally a member of the instrumentation of the FOI outdoor RCS test range Lilla Gåra. The armed test person was standing behind different kinds of walls. The radar images were generated by stepping the turntable in azimuth and elevation. The angular resolution in the near-field was improved by refocusing the parabolic antennas, which in combination with the large bandwidth (12 GHz) gave extremely high resolution radar images. A 3D visualization of the person even exposed the handgun tucked into one hip pocket. A qualitative comparison between the experimental results and simulation results (physical optics-based method) will also be presented.
The second part of this paper describes results from activities at FOI concerning material characterization in the 2-110 GHz region. The transmission of building, packing and clothing materials has been experimentally determined. The wide-band measurements in free space were carried out with a scalar network analyzer. In this paper results from these characterizations will be presented. Furthermore, an experimental investigation will be reported of how the transmission properties for some moisted materials change as a function of water content and frequency. We will also show experimental results of how the transmission properties of a pine panel are affected when the surface is coated with a thin surface layer of water.
This paper describes the experimental research efforts performed at the Swedish Defence Research Agency (FOI) concerning fundamental characterization of different wall and clothing materials as well as through-the-wall imaging.
Results from on-going activities at FOI concerning material characterization in the millimeter wave range are presented. Wide-band measurements of five building materials have been carried out in two different ways. In the frequency range 0.04-40 GHz a vector network analyzer was used and the samples were positioned in waveguides. In the 2-120 GHz region a scalar network analyzer was used and transmission measurements of the materials were performed in free space. Transmission measurements in free space of two clothing materials were also performed.
Results from measurements of a human target standing behind an inner wall are presented.
In certain radar imaging applications one encounters the problem of reconstructing a reflectivity function from information about its averages over circles with center on a straight line. A robust inversion method is a filtered backprojection method, similar to the one used in medical tomography. We will present a fast algorithm for this backprojection operator. Numerical examples are given.