In practical surveillance, an intruder could walk around a target house with a Wi-Fi scanner to probe its interior structure or detect where people are. This technique leverages how Wi-Fi signals attenuate differently when passing through various materials like walls, furniture, and human bodies. By collecting signal strength data from multiple points around a perimeter, one can develop a surprisingly accurate map of interior spaces without physical access.

Doing this in real time is challenging due to the processing requirements and need for multiple measurement points. Real-time applications require significant computational power to process the signal data quickly enough to be useful. Additionally, obtaining comprehensive coverage requires either multiple receivers strategically placed or a mobile receiver that can rapidly collect measurements from different positions, both presenting logistical complexities in field operations.

RSSI monitoring is simple (just measuring power), but it can be noisy and low in resolution. Environmental factors like interference from other electronic devices, multipath propagation, and signal reflections can significantly degrade measurement accuracy. The wavelength of Wi-Fi signals (typically several centimeters) also fundamentally limits the spatial resolution achievable, making it difficult to distinguish small objects or precise details of interior structures without more sophisticated signal processing techniques.

Advanced implementations use many measurements and computations (like synthetic aperture radar algorithms) to form images. These sophisticated approaches combine multiple RSSI readings with statistical models and machine learning algorithms to filter out noise and enhance resolution. Some cutting-edge systems incorporate multiple frequency bands, phased array antennas, and time-of-flight measurements alongside RSSI data to create more detailed three-dimensional representations of hidden spaces and moving objects within them.

CSI-based sensing involves heavy computation (AI models) and is less plug-and-play than RSSI methods. Modern implementations typically require specialized Wi-Fi cards capable of providing CSI data along with high-performance GPUs to run the complex neural networks in real-time. This makes the barrier to entry higher for casual users and increases deployment costs significantly compared to simpler radio-based sensing approaches.

It represents a state-of-the-art technique for through-wall sensing, pushing the envelope from just detecting motion to classifying what the motion is (e.g., gestures, postures) with high granularity. Recent research has demonstrated capabilities like identifying specific individuals by their gait patterns, detecting falls with over 95% accuracy, and even monitoring breathing rates for health applications - all through walls and without requiring the subject to wear any devices.

Researchers envision using this for privacy-preserving monitoring (since it doesn't use cameras) in smart homes. Potential applications include eldercare systems that can detect if someone has fallen without invasive cameras, security systems that can distinguish between pets and people, and smart home controls that respond to gestures rather than voice commands. The non-visual nature makes this technology particularly valuable for bedrooms, bathrooms, and other private spaces where camera monitoring would be inappropriate.

In the wrong hands it could be a very intrusive surveillance tool, capable of detailed monitoring without visible equipment. Unlike cameras which can be spotted or covered, Wi-Fi sensing is completely invisible to the subject. With sufficient training data, malicious actors could potentially identify individual people, monitor activities in private spaces, or track movement patterns throughout buildings without consent. This raises significant ethical questions about informed consent and appropriate regulation, especially as the technology becomes more accessible to non-researchers.

A stationary person is effectively invisible to these systems (aside from the slight breathing motion). The moment they start to move – even just an arm or a shuffle – they introduce a time-varying signal. This creates distinguishable patterns in the received waveforms that sensing systems can detect. Even minute movements as small as 1-2 cm can register as changes in phase or frequency shifts, depending on the system's sensitivity and the wavelength of the signals being used.

This could be as obvious as a spike in Doppler spectrum (a peak at a few Hz corresponding to footstep cadence) or as subtle as a change in the complex channel coefficients. Different movements generate characteristic signatures – walking typically produces rhythmic patterns at 0.5-2 Hz, while rapid hand gestures might show brief bursts at 3-5 Hz. The amplitude, duration, and frequency distribution of these signatures can reveal not just the presence of movement, but also its type, speed, and sometimes even the direction relative to the sensor.

Many systems translate these changes into visualizations. For example, Doppler-time heatmaps (often called micro-Doppler signatures) can show different patterns for different activities. Other common visualization techniques include Range-Doppler maps that show both distance and velocity, spectrogram waterfall displays that reveal the evolution of movement over time, and point cloud reconstructions for multi-sensor systems. These visual representations help researchers and operators interpret the often complex signal relationships and extract meaningful information about human presence and behavior.

A through-wall test showed distinct Doppler traces for actions like walking, waving, or nodding. In another case, a CNN model was used to classify human motions from CSI spectrograms with high accuracy. Advanced machine learning approaches have demonstrated the ability to differentiate between multiple people, identify specific gestures like sign language elements, and even detect emotional states from gait patterns. Military and security applications have developed libraries of movement signatures that can distinguish between civilian and potential threat behaviors, while healthcare monitoring systems can identify warning signs like falls or irregular movement patterns that might indicate medical emergencies.

However, detection is difficult – a passive Wi-Fi radar setup gives off no signal, so the target wouldn't know it's happening. Unlike active systems that emit their own signals, passive systems only receive and analyze existing signals, making them virtually undetectable through conventional means. This stealth capability makes passive Wi-Fi sensing particularly concerning from both a security and privacy standpoint.

This makes enforcement of any such prohibition tricky without specialized RF sensing equipment on the victim's side. Most consumers lack access to radio frequency analyzers or spectrum analyzers that could detect unauthorized signal monitoring. Even with such equipment, distinguishing legitimate network traffic from surveillance activity requires technical expertise beyond most individuals' capabilities.

Technically, these methods involve intercepting radio signals that emanate from someone's private property. Wi-Fi signals naturally pass through walls and can travel significant distances beyond property boundaries, creating ambiguity about where legal boundaries lie. Since these signals are broadcast through public airspace, there's an ongoing debate about whether intercepting them constitutes an invasion of privacy if no encryption is broken.

Wi-Fi signals, when not carrying content (like just observing their presence or waveform), occupy a gray area. Most existing legislation focuses on prohibiting the interception of communication content rather than the analysis of signal characteristics. This distinction creates significant legal uncertainty as new sensing technologies emerge that can derive meaningful information from signal patterns without actually decoding the encrypted data being transmitted.

In summary, Wi-Fi-based human tracking blurs the line between digital and physical privacy. It exploits radio waves that were originally just meant for communication and repurposes them for sensing. This transformation of everyday connectivity infrastructure into surveillance tools happens largely without public awareness or consent, creating a new category of privacy concerns that transcends traditional digital security considerations.

The technology is rapidly advancing – from detecting a simple presence to extracting detailed pose information – and the legal system will need to play catch-up. Research labs are constantly improving algorithms that can discern increasingly subtle movements, while commercial applications are already beginning to appear in smart homes, security systems, and health monitoring devices. This acceleration creates a widening gap between technological capabilities and regulatory frameworks.

In the meantime, the onus is on users and policymakers to be aware of these capabilities. Companies developing these technologies should implement privacy-by-design principles and transparent data practices. Users need education about what their devices might be monitoring, while policymakers must develop informed regulations that protect privacy without stifling beneficial applications in healthcare and safety.

Finding the right balance between legitimate uses and privacy protection will be an ongoing challenge as the technology continues to evolve. Beneficial applications in elder care, home security, and health monitoring must be weighed against the potential for surveillance abuse, unauthorized tracking, and erosion of physical privacy. International standards and clear ethical guidelines will be crucial in establishing boundaries that both protect individual rights and allow for technological innovation.

Environmental factors like wall materials, distance, and interference play a big role in performance. Concrete and metal surfaces cause significant signal attenuation, while wooden structures and drywall are more permeable to Wi-Fi. Signal penetration decreases exponentially with wall thickness, and presence of metal objects, pipes, or electrical wiring can create unpredictable reflections and dead zones. Even humidity levels in the air can influence signal propagation in subtle but measurable ways.

Real-world demonstrations have shown that a person's motion behind a wall can be detected reliably, and even multiple people can be tracked with some precision. This capability has been verified across different environments including homes, offices, and industrial settings. Detection accuracy typically reaches 85-95% for single-person scenarios and 70-80% for multiple occupants. The technology can distinguish between humans and pets with reasonable accuracy due to differences in movement patterns and signal reflection characteristics.

More futuristic, researchers have mapped entire human poses through walls using Wi-Fi, hinting at a world where wireless networks effectively become sensors everywhere. These systems can now identify basic postures like sitting, standing, or lying down with accuracy approaching 80% in controlled environments. Advanced algorithms combining machine learning with signal processing can even detect subtle movements like breathing patterns or gesture recognition. Research from leading universities demonstrates potential applications in healthcare monitoring, where fall detection or abnormal movement patterns can trigger alerts without invasive cameras.

As this field (sometimes dubbed "Wi-Fi sensing" or "wireless vision") matures, it will be important to strike a balance between its beneficial uses and the potential for abuse. The technology offers remarkable possibilities for eldercare, building security, and emergency response scenarios where traditional sensors would be impractical. However, the passive nature of the surveillance makes it particularly concerning from a privacy perspective. Unlike cameras that might be visible, Wi-Fi sensing gives no indication that monitoring is taking place. This creates new challenges for policymakers and technologists to develop appropriate guardrails, consent mechanisms, and technical limitations to prevent misuse while allowing innovation to continue.