Source URL: https://mobile.slashdot.org/story/25/07/22/2112203/humans-can-be-tracked-with-unique-fingerprint-based-on-how-their-bodies-block-wi-fi-signals?utm_source=rss1.0mainlinkanon&utm_medium=feed
Source: Slashdot
Title: Humans Can Be Tracked With Unique ‘Fingerprint’ Based On How Their Bodies Block Wi-Fi Signals
Feedly Summary:
AI Summary and Description: Yes
Summary: Researchers from La Sapienza University in Rome have developed “WhoFi,” a novel system that leverages the distortion of Wi-Fi signals caused by human bodies to identify individuals across different locations, achieving an impressive accuracy of up to 95.5%. This innovation, grounded in deep learning techniques, highlights the potential of using Wi-Fi signals for biometric identification while raising privacy concerns.
Detailed Description:
The development of the WhoFi system signifies a groundbreaking approach in the realm of biometric identification, utilizing Wi-Fi signals—a domain that has traditionally been overlooked. Key points about WhoFi and its implications include:
– **Biometric Identification via Wi-Fi**: WhoFi exploits the unique distortions in Wi-Fi signals, which are altered by the physical characteristics of individuals, allowing for person-specific identification.
– **Channel State Information (CSI)**: The system relies on Channel State Information, which captures detailed data about the amplitude and phase of electromagnetic transmissions. This data is altered by humans, resulting in unique “signatures” that can be analyzed.
– **Deep Neural Network Training**: The researchers used a deep neural network to process the distortions and achieve high accuracy. Their technique outperformed a previous system known as EyeFi, which only achieved 75% accuracy.
– **Privacy Implications**: While the results highlight a novel identification technique, they also raise important privacy concerns. The use of physical characteristics captured through Wi-Fi may pose risks if applied without appropriate safeguards.
– **Technological Viability**: The study positions Wi-Fi signals as a credible and potentially privacy-preserving biometric modality, suggesting a future direction for research in signal-based re-identification systems.
Potential implications for professionals in security, privacy, and compliance include the need to assess how such technologies could impact user privacy and the regulatory landscape concerning biometric data. Organizations must remain vigilant about incorporating emerging technologies while upholding privacy standards and regulations.