Ppg Radar Based Vital Signs

Radio frequency (RF) and radar-based vital signs monitoring represents a fundamentally different approach to contactless physiological measurement compared to optical methods like photoplethysmography (PPG) and remote PPG (rPPG). While optical PPG detects blood volume changes through light absorption modulation, radar-based methods detect the mechanical displacement of the body surface caused by cardiopulmonary activity. Each heartbeat produces a sub-millimeter displacement of the chest wall, and each breath produces a displacement of 5-15 mm. Radar systems can detect these minute movements at distances of several meters, through clothing, in total darkness, and even through walls.

This article provides a comprehensive technical overview of RF and radar approaches to vital signs monitoring, comparing their capabilities and limitations to optical PPG, and examining the emerging convergence of these modalities in multi-sensor health monitoring systems.

Physical Principles of Radar Vital Signs Detection

The Doppler Effect and Phase Modulation

The fundamental mechanism underlying radar vital signs detection is the modulation of the reflected electromagnetic wave by body surface displacement. When a radar signal reflects from a moving surface, the reflected signal experiences a phase shift proportional to the displacement. For a target at distance d(t) from the radar, the received signal phase is:

phi(t) = 4 * pi * d(t) / lambda

where lambda is the radar wavelength. For a 60 GHz radar (lambda = 5 mm), a 0.3 mm heartbeat-induced chest displacement produces a phase shift of approximately 0.75 radians, which is readily measurable with modern radar receivers.

This is fundamentally different from PPG, which measures optical absorption changes due to blood volume pulsations. Radar measures the mechanical consequence of cardiac contraction (chest wall motion), while PPG measures the hemodynamic consequence (peripheral blood volume pulse). This distinction means that radar and PPG provide complementary information about cardiovascular function.

Signal Characteristics

The radar vital signs signal contains both respiratory and cardiac components. The respiratory component dominates in amplitude, with chest displacement of 5-15 mm producing strong phase modulation. The cardiac component is typically 10-50 times weaker (0.2-0.5 mm displacement), making its extraction a significant signal processing challenge. This amplitude disparity is analogous to the challenge of extracting the cardiac-frequency PPG signal from larger motion artifacts, though the noise sources differ.

The radar vital signs signal also contains body motion artifacts from limb movement, postural shifts, and involuntary muscle contractions. These random body movements (RBM) can be orders of magnitude larger than the vital signs components, presenting challenges similar to those encountered in wearable PPG during physical activity.

Radar Architectures for Vital Signs Monitoring

Continuous Wave (CW) Doppler Radar

CW Doppler radar is the simplest architecture for vital signs detection. It transmits a single-frequency continuous wave and analyzes the phase of the reflected signal. The key advantage is simplicity: CW radar can be implemented with minimal hardware complexity and achieves excellent phase sensitivity for displacement measurement.

Li et al. (2006) demonstrated one of the earliest CW radar vital signs systems at 2.4 GHz, successfully detecting both respiration and heart rate at 1.5 meters distance (DOI: 10.1109/TMTT.2006.884190). The system achieved respiratory rate accuracy within 0.5 breaths per minute and heart rate accuracy within 2.5 BPM compared to reference contact sensors.

The primary limitation of CW radar is the lack of range discrimination. All reflections from the environment contribute to the received signal, and the vital signs of multiple people at different distances cannot be separated. The "null point problem" is another CW-specific issue: at certain radar-to-subject distances, the chest displacement aligns with a point of zero phase sensitivity, causing signal dropout. Quadrature receivers (I/Q architecture) and optimal demodulation techniques (e.g., arctangent demodulation) mitigate the null point problem but do not fully eliminate it.

FMCW (Frequency-Modulated Continuous Wave) Radar

FMCW radar transmits a frequency-swept signal (chirp) and achieves range resolution through beat frequency analysis. This enables separation of targets at different distances, allowing isolation of the vital signs signal from a specific individual in a multi-person environment. Range resolution is determined by the sweep bandwidth: delta_R = c / (2 * B), where B is the bandwidth. A 4 GHz bandwidth at 60 GHz yields range resolution of approximately 3.75 cm.

Texas Instruments' IWR1443 and AWR1642 77 GHz FMCW radar chips have become widely used research platforms for vital signs monitoring due to their commercial availability, low cost, and integrated antenna arrays. These devices achieve range resolution of approximately 4 cm with 4 GHz bandwidth and can simultaneously track vital signs of multiple individuals at different ranges.

Alizadeh et al. (2019) demonstrated that 77 GHz FMCW radar could detect heart rate with MAE of 1.8 BPM and respiratory rate with MAE of 0.5 breaths per minute at 1 meter distance, with performance degrading gracefully to MAE of 3.5 BPM for heart rate at 3 meters (DOI: 10.1109/JSEN.2019.2898621). The range gating capability of FMCW eliminates most environmental clutter and enables vital signs monitoring of specific individuals in multi-occupant rooms.

UWB (Ultra-Wideband) Impulse Radar

UWB radar transmits extremely short pulses (sub-nanosecond) with very wide bandwidth (typically 3-10 GHz), achieving centimeter-level range resolution. UWB radar has two unique advantages for vital signs monitoring: very high range resolution for precise target localization, and the ability to penetrate non-metallic walls and barriers due to the low frequency components of the wideband signal.

Through-wall vital signs detection using UWB radar has been demonstrated for search-and-rescue applications, where buried or trapped individuals' breathing can be detected through rubble or collapsed structures. Ren et al. (2016) demonstrated respiratory detection through a 20 cm thick concrete wall at 5 meters total distance using a 0.5-3 GHz UWB radar (DOI: 10.1109/JSEN.2015.2505738). Heart rate detection through walls is more challenging due to the much smaller cardiac displacement, but has been achieved at shorter ranges with advanced signal processing.

The low center frequency of UWB radar means that the cardiac-induced phase modulation is proportionally smaller (phase shift scales inversely with wavelength), making heartbeat detection more challenging than with higher-frequency radars. Algorithmic advances including variational mode decomposition and ensemble empirical mode decomposition have improved cardiac signal extraction from UWB data.

Millimeter-Wave (mmWave) Radar

Millimeter-wave radar operating at 60 GHz or 77 GHz offers the highest phase sensitivity for vital signs detection due to short wavelengths (4-5 mm). The same 0.3 mm cardiac displacement that produces 0.1 radians of phase shift at 5 GHz produces 1.5 radians at 77 GHz, yielding dramatically improved cardiac signal SNR.

The 60 GHz band is of particular interest because it falls within an unlicensed ISM band in many jurisdictions and is subject to high atmospheric oxygen absorption (~15 dB/km), which naturally limits range and reduces interference in dense deployments. Google's Soli project and the subsequent Pixel 4 and later devices incorporated 60 GHz radar for gesture recognition, and the same hardware can be repurposed for vital signs monitoring. Mercuri et al. (2019) demonstrated 60 GHz radar vital signs monitoring with heart rate MAE of 1.1 BPM during stationary conditions (DOI: 10.1109/TMTT.2019.2930572).

The primary limitation of mmWave radar is shorter range due to higher path loss and atmospheric absorption. Beam focusing with antenna arrays partially compensates for this, and modern mmWave radar chips integrate phased array antennas that can electronically steer the beam toward a specific subject.

WiFi-Based Vital Signs Sensing

An alternative RF approach to vital signs monitoring exploits existing WiFi infrastructure rather than dedicated radar hardware. WiFi sensing analyzes the Channel State Information (CSI) of standard 802.11 WiFi signals to detect body motion, including the subtle movements caused by respiration and heartbeat.

CSI-Based Vital Signs Extraction

When WiFi signals propagate between a transmitter and receiver, they travel through multiple paths (direct, reflected, scattered). The CSI describes the complex gain of each subcarrier in the OFDM WiFi signal and encodes the complete multipath channel. Body motion modifies the reflection and scattering characteristics, modulating the CSI.

Liu et al. (2015) demonstrated respiratory monitoring using commodity WiFi hardware (Intel 5300 NIC), achieving respiratory rate accuracy within 1 breath per minute in 94% of measurements (DOI: 10.1145/2789168.2790121). Heart rate extraction from WiFi CSI is more challenging due to the smaller cardiac-induced body motion, but Wang et al. (2017) achieved heart rate accuracy within 2 BPM under favorable conditions using advanced CSI processing with the Fresnel zone model (DOI: 10.1145/3131672.3131693).

Advantages and Limitations

WiFi sensing has the unique advantage of requiring no additional hardware in environments with existing WiFi access points. It enables ambient vital signs monitoring without any device worn or installed specifically for health purposes. However, WiFi sensing has significantly lower accuracy than dedicated radar, particularly for heart rate. The WiFi signal was not designed for displacement sensing, and its sensitivity to cardiac-level body motion (0.2-0.5 mm) is marginal. Multi-person separation is also challenging because WiFi CSI captures the aggregate channel response from all reflectors.

Comparison: Radar vs. Optical PPG

The comparison between radar and optical PPG approaches reveals fundamental complementarity rather than direct competition.

What Radar Measures vs. What PPG Measures

Radar detects mechanical displacement of the body surface. PPG detects optical absorption changes caused by blood volume pulsations. These are related but distinct physiological phenomena. The cardiac-induced chest wall displacement reflects the overall mechanical activity of the heart, while the peripheral PPG pulse reflects the hemodynamic propagation of the pressure wave through the vascular system.

This distinction has important clinical implications. Radar can detect cardiac activity even in conditions where peripheral PPG signals are absent or severely degraded, such as during peripheral vasoconstriction, hypothermia, or severe hypotension. Conversely, PPG provides information about peripheral vascular health, blood oxygenation, and pulse wave morphology that radar fundamentally cannot access.

Performance Comparison

Complementary Fusion

The complementary strengths of radar and optical PPG suggest that multi-modal systems combining both modalities could outperform either alone. Radar provides robust cardiac timing information unaffected by optical interference, skin tone, or peripheral perfusion, while PPG provides blood oxygenation, detailed pulse wave morphology, and vascular health indicators.

Zhao et al. (2018) demonstrated a combined 60 GHz radar and camera rPPG system that achieved heart rate MAE of 0.8 BPM across diverse conditions, compared to 1.5 BPM for radar alone and 2.3 BPM for rPPG alone (DOI: 10.1145/3241539.3241544). The fusion approach used radar for robust cardiac timing and rPPG for waveform morphology, producing a composite signal superior to either individual modality.

Signal Processing for Radar Vital Signs

Clutter Removal

Static environmental reflections (walls, furniture, equipment) dominate the radar return signal and must be removed before vital signs extraction. Common clutter removal techniques include background subtraction (subtracting the mean signal to remove static components), high-pass filtering of the slow-time signal, and singular value decomposition (SVD) where the largest singular values correspond to static clutter.

Respiratory-Cardiac Separation

Because the respiratory signal is 10-50 times stronger than the cardiac signal, separating these components is critical. Simple bandpass filtering (0.1-0.5 Hz for respiration, 0.8-2.5 Hz for heart rate) works when the frequencies do not overlap. However, respiratory harmonics can fall within the cardiac band, and during exercise, respiratory rate may overlap with resting heart rate.

Advanced separation methods include empirical mode decomposition (EMD), variational mode decomposition (VMD), and wavelet decomposition. Li et al. (2013) demonstrated that VMD outperforms EMD for radar vital signs separation, achieving cardiac signal reconstruction with SNR improvement of 8-12 dB (DOI: 10.1109/TBME.2013.2257334).

Motion Artifact Handling

Random body movements create large signal perturbations that can mask vital signs components for seconds to minutes. Motion detection using signal power thresholds or acceleration statistics can identify and exclude motion-corrupted segments. More sophisticated approaches use separate radar channels or antenna elements for motion tracking and vital signs measurement simultaneously.

Clinical and Commercial Applications

Sleep Monitoring

Radar-based sleep monitoring is commercially deployed in products like the Google Nest Hub (2nd generation), which uses a Soli 60 GHz radar chip to track sleep stages through respiration patterns, body motion, and coughing events. Unlike wearable sleep trackers that use contact PPG and accelerometry, radar sleep monitors require no device to be worn, improving comfort and compliance.

Automotive Occupant Monitoring

The automotive industry is integrating 77 GHz radar vital signs monitoring for in-cabin occupant detection and driver health monitoring. European NCAP regulations increasingly require child presence detection systems, and radar provides a reliable, lighting-independent approach. Vital signs monitoring can detect driver drowsiness or medical emergencies through heart rate and respiratory abnormalities.

Healthcare Facility Monitoring

Continuous radar vital signs monitoring of hospitalized patients could reduce alarm fatigue from wired contact sensors and improve monitoring of ambulatory patients. Adib et al. (2015) demonstrated room-scale vital signs monitoring using a custom FMCW radar, tracking the respiration and heart rate of multiple patients simultaneously with clinically acceptable accuracy (DOI: 10.1145/2789168.2790124).

Future Directions

The integration of on-chip radar with edge AI processing is enabling increasingly compact and capable vital signs radar systems. Single-chip radar solutions from Texas Instruments, Infineon, and others integrate transmitter, receiver, antenna, and digital processing into packages smaller than a postage stamp, enabling embedding in consumer devices, light fixtures, and bathroom mirrors.

Multi-modal systems combining radar with optical sensing (both contact PPG and camera-based rPPG) represent the most promising direction for robust, comprehensive vital signs monitoring. The complementarity of mechanical (radar) and optical (PPG) sensing modalities provides resilience against individual failure modes, while sensor fusion algorithms can optimally weight each modality based on instantaneous signal quality.

The convergence of radar vital signs monitoring with digital biomarker frameworks opens new possibilities for continuous, unobtrusive health assessment that operates passively in the background of daily life, requiring no user interaction, no wearable device, and no conscious measurement session.

References

• Li et al. (2006) demonstrated one of the earliest CW radar vital signs systems at 2.4 GHz, successfully detecting both respiration and heart rate at 1.5 meters distance (DOI: 10.1109/TMTT.2006.884190). The system achieved respiratory rate accuracy within 0.5 breaths per minute and heart rate accuracy within 2.5 BPM compared to reference contact sensors.
• Alizadeh et al. (2019) demonstrated that 77 GHz FMCW radar could detect heart rate with MAE of 1.8 BPM and respiratory rate with MAE of 0.5 breaths per minute at 1 meter distance, with performance degrading gracefully to MAE of 3.5 BPM for heart rate at 3 meters (DOI: 10.1109/JSEN.2019.2898621). The range gating capability of FMCW eliminates most environmental clutter and enables vital signs monitoring of specific individuals in multi-occupant rooms.
• Through-wall vital signs detection using UWB radar has been demonstrated for search-and-rescue applications, where buried or trapped individuals' breathing can be detected through rubble or collapsed structures. Ren et al. (2016) demonstrated respiratory detection through a 20 cm thick concrete wall at 5 meters total distance using a 0.5-3 GHz UWB radar (DOI: 10.1109/JSEN.2015.2505738). Heart rate detection through walls is more challenging due to the much smaller cardiac displacement, but has been achieved at shorter ranges with advanced signal processing.
• The 60 GHz band is of particular interest because it falls within an unlicensed ISM band in many jurisdictions and is subject to high atmospheric oxygen absorption (~15 dB/km), which naturally limits range and reduces interference in dense deployments. Google's Soli project and the subsequent Pixel 4 and later devices incorporated 60 GHz radar for gesture recognition, and the same hardware can be repurposed for vital signs monitoring. Mercuri et al. (2019) demonstrated 60 GHz radar vital signs monitoring with heart rate MAE of 1.1 BPM during stationary conditions (DOI: 10.1109/TMTT.2019.2930572).
• Liu et al. (2015) demonstrated respiratory monitoring using commodity WiFi hardware (Intel 5300 NIC), achieving respiratory rate accuracy within 1 breath per minute in 94% of measurements (DOI: 10.1145/2789168.2790121). Heart rate extraction from WiFi CSI is more challenging due to the smaller cardiac-induced body motion, but Wang et al. (2017) achieved heart rate accuracy within 2 BPM under favorable conditions using advanced CSI processing with the Fresnel zone model (DOI: 10.1145/3131672.3131693).
• Zhao et al. (2018) demonstrated a combined 60 GHz radar and camera rPPG system that achieved heart rate MAE of 0.8 BPM across diverse conditions, compared to 1.5 BPM for radar alone and 2.3 BPM for rPPG alone (DOI: 10.1145/3241539.3241544). The fusion approach used radar for robust cardiac timing and rPPG for waveform morphology, producing a composite signal superior to either individual modality.
• Advanced separation methods include empirical mode decomposition (EMD), variational mode decomposition (VMD), and wavelet decomposition. Li et al. (2013) demonstrated that VMD outperforms EMD for radar vital signs separation, achieving cardiac signal reconstruction with SNR improvement of 8-12 dB (DOI: 10.1109/TBME.2013.2257334).
• Continuous radar vital signs monitoring of hospitalized patients could reduce alarm fatigue from wired contact sensors and improve monitoring of ambulatory patients. Adib et al. (2015) demonstrated room-scale vital signs monitoring using a custom FMCW radar, tracking the respiration and heart rate of multiple patients simultaneously with clinically acceptable accuracy (DOI: 10.1145/2789168.2790124).