Monitoring Human Heartbeat and Respiration

 

An experiment using X-band SDR-KIT 980AD with CW waveform for monitoring a stream of live human heartbeat and breathing data is shown in the Figure. From the continuous flow of data, the respiratory function and heartbeat function can be clearly monitored in real-time. After DC cancellation, data rate reduction, and I-Q imbalance correction, demodulation operation, phase unwrapping, differencing, and detrending processing, the spectrum of human heartbeat and respiration can be obtained. FMCW waveforms are also often used to provide additional target range information. 

 

Other frequency bands, such as C-band SDR-KIT 580 or K-band SDR-KIT 2400, can also be used for monitoring human vital sign.

 

Reference: The micro-Doppler Effect in Radar (Second Edition), Chapter 5, By Victor C. Chen, Artech House, 2019.

 

Traffic Monitoring

 

Ancortek SDR-KIT 580AD is one-transmitter and one-receiver radar kit in C- band with typical frequency limits 5.6-6.0 GHz. Its center frequency and the bandwidth of transmitted signals are selectable and adjustable. Its typical output power of the transmitter is 21 dBm. The maximum detectable range is about 120-m for 1-sm of RCS. It is designed to support human activity monitoring, vital-sign detection, gesture sensing, through-the-wall, and many other indoor and outdoor target detection and imaging.

 

In order to monitor traffic at beyond 0.5 km, an additional 1-watt power amplifier is needed. The Figure shows that the customized 1-Watt 580AD radar with horn antennas is monitoring a highway at 600m away from the radar. The highway is marked by the red dotted lines. From the range-velocity map we can see trajectories of vehicles on the highway within 15sec observation time duration.

 

Radar Tracking of Micro-UAV

 

The SDR-KIT 580AD2, centered at 5.8 GHz with dual receivers was used for testing the drone detection, tracking and classification. Its transmitting power is 19 dBm and the gain for both receiving antenna is 15dBi. For covering 70 degree azimuth angle, The spacing between two receiving channels is 0.8 wavelength.

 

With 300MHz bandwidth, the range resolution is 0.5m. If the sweep time T is 1ms and the number of sweeps for each frame is 64, the Doppler resolution is 15.625 Hz, or the velocity resolution is 0.4 m/s. The unambiguous velocity becomes ± 12.6 m/s. If we select 128 samples per sweep, the sampling rate becomes 128 kHz. Thus, the maximum range coverage is about 32m.

 

A drone can be detected in the range-Doppler map by finding local peaks in the range-Doppler domain. A beamforming method can be used to measure the angle-of-arrival information on the detected drones. A DJI Phantom 3 with radar cross section of 0.05sm (-13dBsm) was used in the experiment. The maximum speed of the Phantom 3 when flying horizontally is 15m/s, which is larger than the maximum unambigous velocity. For experimental purpose, we limit the speed of the drone within 3 m/s. To demonstrate tracking capability of the system, circular flying trajectory was recovered in a range and cross-range map as shown in the Figure. The drone started at around 10 degrees to the left and 4 m away from the radar (the green dot). At 11 meters, it made a turn and flew back next to its starting point (the red dot). ­­­­­­­­­­­­The range trajectory is also shown in the Figure.

 

Hand Gesture Recognition

 

An experiment of using SDR-KIT 2400 for hand gesture recognition is illustrated in the following Figure. It shows that only based on micro-Doppler features of hand gestures without using other features (such as range profiles, Doppler profiles, range-Doppler maps, or time-varying range-Doppler maps), it is possible to recognize different hand gestures. To recognize a hand gesture, the radar captured data must be processed and decomposed into a feature space. Then, the CNNs and deep learning may be applied to recognize the hand gestures.

 

The Figure also illustrates the process of the snapping fingers. It indicates that there is a higher negative Doppler peak and a lower positive Doppler peak located at the same time instant. The higher negative Doppler peak corresponds to flipping the middle finger onto the palm and the lower positive Doppler peak corresponds to flipping the thumb forward the radar. Then the three fingers are reset back to their initial positions and there is another low positive Doppler peak appearing.

 

Based on radar micro-Doppler signatures of a set of pre-selected hand gestures, simple monostatic CW Doppler radar can perform very well for the hand gesture recognition. An example of a micro-Doppler signature-based hand gesture recognition system was demonstrated for controlling a CD player. For simplicity, the demonstration only used four gestures to control a CD player: snap fingers for “Play”, swipe away for “the Next”, swipe towards for “the Previous”, and flip fingers for “Stop”.

 

Reference: The micro-Doppler Effect in Radar (Second Edition), Chapter 6, By Victor C. Chen, Artech House, 2019.

 

ISAR Image of a Vehicle

 

A range-Doppler image of a moving target can be generated by a Doppler radar. The Doppler resolution is achieved by the antenna aperture. ISAR (inverse synthetic aperture radar) utilizes the relative rotation between the radar and the target to synthesize a larger antenna aperture and, thus, a higher resolution range-Doppler image.

 

The C-band SDR-KIT 580, X-band 980 or K-band 2400 can be used to collect radar data for generating ISAR images of moving car as shown in the Figure. To form an ISAR range-Doppler image of a moving target, the first step is to carry out the translational motion compensation (TMC).  It estimates target’s translational motion parameters and removes the extra-phase term. In this case, the rotational motion compensation (RMC) must be carry out to correct for the rotation. Then, by taking 2-D Fourier transform, the range-Doppler image of the target can be reconstructed as seen in the right-site of the Figure.

 

Radar Level Gauge

 

Radar level gauge is an effective tool for non-contacting measurement of the level of liquid or solid in a large-scale tank. To test radar level gauge for measuring the exact distance from radar to surface level in a tank, we use SDR-KIT 980B and 2500B with FMCW signal waveform for short-distance measurement.  The beat frequency, i.e., the frequency difference between the transmitted and the received FMCW signals, is directly proportional to the distance to be measured. By accurately measuring the beat frequency using digital signal processing, the distance can be measured precisely and accurately.

 

level gauge 1

 

The experimental system diagram using SDR-RF 2500B module and SDR-PM 402 processor module, and output/display as shown below. A reflective metal plate was mounted on an adjustable precise positioning table. The distance between the radar and the metal plate is 0.828 m. After precisely adjusting the position of the metal plate by 1 mm stepping, the range reading showing in the GUI did reflect the 1 mm steps.  The test verified that with a 250 MHz signal bandwidth, the 0.6 m range resolution/precision can be improved to more than 1,000 times and a precision of <1.0 mm can be achieved.

 

level gauge 2

level gauge 3

Environmental Monitoring

 

In this trial, the SDR-KIT 580B operating in FMCW mode was used to monitor the environment, such as surrounding vehicles, pedestrians, and even animals. The antenna was at approximately 4-meter height above the ground monitoring a gas station where vehicles or persons are in and out frequently.

 

Tracking walking person

 

In this trial, the SDR-KIT 580B was used to track walking persons.  FMCW waveforms allow to measure distances of moving or static objects. A person was walking back and forth inside a room.  The waterfall display in Figure (b) shows walking track of the person. The strong vertical stripes are caused by the returns from the wall.  If the option of background subtraction in the SDR GUI is selected, the wall will be subtracted as shown in the figure below.

 

Two People Walking

 

In this trial, the SDR-KIT 580B was used to track walking person. Two persons were walking in opposite directions. An option of the DC subtraction function in the SDR GUI was enabled.  The Range-velocity map shows two hotspots walking towards opposite directions.

 

TwoPersonWalking

Tracking of Vehicle

 

In this trial, the SDR-KIT 620B was used to track the vehicle. The antenna was mounted at approximately 4 meter height above ground. FMCW modes was used to generate a range-velocity map.  Figure (a) is a snapshot of a moving car towards the line-of-sight and is registered at a range of around 19 meter with a velocity of -4 m/s. Figure (b) is the range-velocity map of the vehicle.

 

TrackingVehicle

 

 

Vehicle Speed

 

In this trial, the SDR-KIT 620B working in CW mode was used to track the speed of the vehicle. A car was running towards the antenna at a speed of about 30 mile per hour (13.4m/s) as shown in the figure (a).  A snapshot of the measured velocity plot from the SDR-GUI is shown in the figure (b). The detected speed was 13.4 m/s, which very well matches to the reading from the speedometer of the car.

 

TrackingSpeed

Image of a Vehicle

 

In this trial, the SDR kit 620B was used for imaging a walking person crossing a parking lot. The experiment was setup as shown in figure (a), where the SDR-KIT is located at 4-meter height above the ground. Figure (b) is range-Doppler image of the walking person showing in its range and cross-range. The torso and swinging foots can be seen.

 

TrackingSpeed

 

 

Image of Person Walking

 

In this trial, the SDR-KIT 620B working in CW mode was used to track the speed of the vehicle. A car was running towards the antenna at a speed of about 30 mile per hour (13.4m/s) as shown in the figure (a).  A snapshot of the measured velocity plot from the SDR-GUI is shown in the figure (b). The detected speed was 13.4 m/s, which very well matches to the reading from the speedometer of the car.

 

ISARImageOfWalkingMan

 

Radar Micro-Doppler Signatures

 

Radar micro-Doppler signatures can be used to identify human activities, such as walking, running or jumping, to monitor and predict human behaviors, and to detect a person at a distance.  Human gait analysis is useful in biomedical engineering, sports medicine, physiotherapy, medical diagnosis, and rehabilitation.  Compared with the visual perception of the human body motion, the radar micro-Doppler method is not affected by distance, variations in lighting, deformations of clothing, and occlusions on the appearance of human body segments.

 

In the micro-Doppler signature of a walking human, each forward leg swing appears as large peaks, and the left and right leg-swing completes one gait cycle.  The body torso motion that is the stronger component underneath the leg swings tends to have a slightly saw-tooth shape because the body speeds up and slows down during the swing as shown. It is possible to classify and identify bodies and their movements based on radar micro-Doppler signatures for human gaiting.

 

The SDR-KIT 980B can be used as a micro-Doppler radar.  The micro-Doppler radar using FMCW waveforms can provide more than 10 dB SNR on a 1 square meter RCS at a distance up to 50 meters.

 

RADAR MICRO-DOPPLER SIGNATURES OF A WALKING PERSON 

 

In this trial, the SDR-KIT 980B with 400 MHz bandwidth was used to capture an approaching walking person starting at a distance about 10 m.

 

MicroDopplerOfWalkingMan

 

 

RADAR MICRO-DOPPLER SIGNATURES OF A RUNNING PERSON

 

In this trial, the SDR-KIT 980B with 400 MHz bandwidth was used to capture an approaching running person starting at a distance about 10 m.

 

MicroDopplerOfRunningMan

 

 
RADAR MICRO-DOPPLER SIGNATURES OF MODEL HELICOPTER 

 

In this trial, the SDR-KIT 980B with 400 MHz bandwidth was used to capture a model helicopter with rotating rotor blades at a distance less than 10 m.

 

MicroDopplerOfHelicopter

 

 
RADAR MICRO-DOPPLER SIGNATURES OF A QUADCOPTER

 

In this trial, the SDR-KIT 2400AD2 with 500 MHz bandwidth was used to measure both radial and angular velocity of a drone quadcopter with rotating rotor blades at a distance less than 0.7 m and a 90-degree radar depression angle. Antennas spacing was set at 5λ.

 

drone-experiment