Introduction
As smart environments expand, the number of connected devices in homes and offices is skyrocketing. Managing access to these devices—from smart toasters to office equipment—presents a significant usability and security challenge. Traditional authentication methods like passwords are impractical for screenless devices, while biometric hardware like fingerprint scanners is often too expensive and power-intensive for low-cost IoT nodes.
TouchAuth introduces a "touch-to-access" paradigm. It leverages the human body as an antenna to capture ambient electrical noise from indoor wiring, enabling secure authentication through a simple touch. By comparing induced body electric potentials (iBEP) measured by a wearable "authenticator" and a smart "authenticatee," the system verifies that both are being touched by the same authorized individual.
System Architecture and Protocol
The TouchAuth system consists of two primary components: an authorized wearable (e.g., a smartwatch) and a target smart object. Both devices are equipped with standard Analog-to-Digital Converter (ADC) pins to sample the iBEP signal.
The authentication process follows a structured protocol:
- Initialization & Secure Channel: The devices establish a TLS connection and synchronize their clocks.
- iBEP Sampling: Both devices simultaneously sample the ambient electrical noise through the user's skin.
- Secure Exchange: Using a commitment-based "H2H" protocol, the devices exchange their sampled data. This prevents man-in-the-middle attacks where an adversary might attempt to replay or forge signals.
- Correlation Analysis: The system calculates the Absolute Pearson Correlation Coefficient (APCC) between the two signals. A high correlation indicates that the signals originated from the same human body, leading to an "accept" decision.
The Body Antenna Effect
The fundamental principle behind TouchAuth is the Body Antenna Effect. Indoor environments are permeated by 50/60Hz electric fields (EF) emitted by AC power lines. When a person touches an electrode, their body acts as a large conductor, significantly altering the local EF and inducing a measurable potential.
Because the human body is a conductor, it acts as an equipotential surface. Two sensors touching the same person will observe highly correlated potential fluctuations relative to the body's reference, even if their own "floating grounds" differ. This spatial sensitivity ensures that even individuals standing close to each other will have distinct iBEP signatures, providing a robust layer of security against accidental cross-authentication.
Key Results and Performance
TouchAuth was evaluated with 12 participants across 8 diverse indoor environments, including kitchens and laboratories. The system demonstrated high accuracy and efficiency:
- Speed and Accuracy: With only 1 second of sampling, the system achieves a 94.2% True Acceptance Rate (TAR). Extending the sampling time to 5 seconds increases the TAR to 98.9%, with a False Acceptance Rate (FAR) of just 2%.
- Robustness: The system maintains performance in the presence of high-power appliance noise (e.g., heaters, blenders) and active cellular interference.
- Efficiency: Compared to physiological signals like ECG or PPG, which often require 30–60 seconds of data, TouchAuth provides near-instantaneous authentication using ubiquitous, low-cost hardware.
Limitations and Future Outlook
While highly effective indoors, TouchAuth relies on the presence of ambient electric fields from building wiring, making it unsuitable for outdoor use. Additionally, the signal correlation decreases as the distance between sensors on the body increases (e.g., wrist to head), which serves as a natural security boundary but limits certain wearable configurations.
Future research will focus on further miniaturizing the sensing electrodes and integrating TouchAuth into a broader range of consumer electronics, potentially replacing traditional pairing and login methods with a simple, intuitive touch.
Summary updated