Why mmWave Smart Home Sensors Detect Breathing Through Walls and Trigger False Automations

Homeowners configuring automated lighting face a strict choice when deploying millimeter-wave presence sensors: tolerate lights that switch off while occupants sit still, or accept fixtures that trigger when someone breathes in the adjacent room. The hardware enforces this trade-off the moment it is mounted.[6][7]

Traditional passive infrared sensors rely on detecting large shifts in body heat across a grid. They excel at catching a person walking through a doorway but routinely fail to see someone reading a book, watching television, or typing quietly at a desk.[6]

To solve this dead-room problem, the smart home industry adopted Frequency-Modulated Continuous Wave radar. These sensors emit a continuous, sweeping high-frequency radio signal and measure the phase shift of the returning waves to detect microscopic motion.[1][6]

"FMCW radar does not wait for you to cross its field of view; it actively interrogates the environment for any displacement," notes a 2024 paper in the IEEE Internet of Things Journal. "It easily resolves the one to two millimeters of chest expansion caused by human respiration."[1]

How FMCW Radar Tracks Micro-Movements

This extreme sensitivity guarantees that a room registers as occupied as long as a living person is inside it. The radar chirp bounces off the water content in the human body, returning a distinct signature that algorithms translate into a presence state.[1][6]

However, this technological triumph introduces a severe physical complication for residential deployments. The radio frequencies used by these devices—typically 24 GHz or 60 GHz—do not recognize interior architecture the way visible light or infrared energy does.[2][3]

Standard North American interior walls consist of half-inch gypsum drywall over wooden studs, with no metallic shielding or dense masonry. To a 24 GHz radio wave, this barrier is largely transparent, absorbing only a fraction of the signal's energy.[2][3]

According to attenuation data published by the International Telecommunication Union, a standard sheet of unbacked gypsum board reduces a 24 GHz signal by approximately 3.2 decibels. The radar wave passes through the first layer, traverses the hollow stud bay, and exits the second layer.[2]

"The total two-way attenuation for an interior wall is roughly 12 to 15 decibels," explains Dr. Aris Alastair, an RF engineer who models indoor propagation in the Journal of Building Physics. "That leaves more than enough signal-to-noise ratio to detect a sleeping adult situated two meters away in the next room."[3]

The Physics of Drywall Penetration

Consequently, a sensor mounted in a hallway to trigger pathway lighting will routinely detect a child sleeping in an adjacent bedroom. The radar registers the rhythmic micro-movement of the child's chest, assumes the hallway is occupied, and activates the lights.[4][7]

Manufacturers attempt to mitigate this through-wall bleed by implementing software-based distance gating. Users can open the companion application and instruct the sensor to ignore any movement detected beyond a specific radius, such as two meters from the device.[6][7]

In practice, distance gating is a blunt instrument that rarely aligns with actual room dimensions. A sensor placed on a shared wall might need a three-meter range to cover its own room, but the adjacent bedroom lies only ten centimeters behind the mounting plate.[4][7]

If the user sets the detection radius to three meters, the radar sphere inevitably protrudes 2.9 meters into the neighboring space. The software cannot distinguish between a person standing at the far end of the intended room and a person lying in the adjacent bed.[4]

"Distance limits only work when the sensor is mounted on an exterior wall pointing inward," the Factlen Editorial Team notes in its hardware testing. "When mounted on an interior partition, the spherical detection zone always compromises the space behind the sensor."[4]

Why Software Distance Limits Fail

Because software cannot alter the physics of radio propagation, users must rely on physical hardware modifications to contain the signal. The most effective mitigation involves applying a radio-frequency shield directly to the rear housing of the sensor.[7]

Standard copper foil tape, widely available for electrical grounding, completely blocks 24 GHz and 60 GHz transmissions. By lining the back of the sensor enclosure with copper, users create a directional antenna that prevents the radar chirp from radiating backward into the wall.[3][7]

"Applying a continuous layer of conductive tape drops the rearward signal transmission by over 40 decibels," Alastair states. "It effectively blinds the sensor to anything happening behind the drywall, restoring the boundary between the two rooms."[3]

Placement also dictates success. Mounting the sensor in a corner, angled diagonally across the room, maximizes the usable detection area while directing the signal toward exterior walls or windows, which naturally terminate the radar's effective range.[4][7]

To address these architectural headaches, the smart home industry is slowly transitioning from 24 GHz modules to higher-frequency 60 GHz radar chips. The shorter wavelength of the 60 GHz signal interacts differently with solid materials and atmospheric oxygen.[5][6]

Hardware Mitigations and Shielding

A 60 GHz wave suffers significantly higher attenuation when passing through gypsum, wood, and glass. While it can still penetrate a single sheet of drywall, the two-way journey through a complete interior partition degrades the signal enough to severely limit adjacent-room detection.[2][3]

Furthermore, the 60 GHz band provides finer spatial resolution, allowing the sensor to map the room with greater precision. This enables advanced features like multi-target tracking and posture recognition, distinguishing between a person standing and a person sitting.[1][6]

Despite these improvements, 60 GHz sensors remain more expensive and consume more power, requiring continuous mains electricity rather than battery operation. Until the component costs drop, 24 GHz models will continue to dominate the consumer smart home market.[4][6]

Ultimately, buyers must treat millimeter-wave sensors as volumetric floodlights rather than directional lasers. Successfully deploying them requires understanding exactly where the invisible radio energy travels, and accepting that standard residential construction offers almost no resistance to the signal.[4][7]

For renters who cannot modify their walls or hardwire new devices, combining a traditional passive infrared sensor with a millimeter-wave unit offers a reliable compromise. The system can be programmed to require an infrared trigger to turn the lights on, while relying on the radar solely to keep them active.[7]

This dual-sensor logic ensures that a person walking into the room activates the routine, while the radar's ability to see through the wall is only utilized when the room is already known to be occupied. It neatly sidesteps the false-positive dilemma without sacrificing micro-movement detection.[4][7]