Gait Device Pairing in Unconstrained Environments With Signal Morphing

People carry consumer devices, like smartphones and watches, as part of daily life to provide connectivity and data collection for various consumer applications. Consumers and manufacturers are becoming increasingly aware of data security risks of such devices, but security mechanisms for consumer devices to securely pair and communicate should not detract from core applications and user experience. As such, a method is needed to set up a secure body area network with low resource overhead while being transparent to the consumer. One solution that has gained popularity in research is gait-based key-sharing protocols. Our protocols were tested using over 3148 samples of gait data, totalling more than 15 hours of movement data gathered exclusively in uncontrolled environments. As a result, we ensure that our system can adapt to the dynamic/ever-changing life of consumers. Despite low correlation in data from these environments, applying optimization techniques allows us to map motion data between different body locations, improving our gait-based key-sharing protocol’s robustness. By leveraging different sensors and our signal mapping approach, we achieve an x4.93 increase in the Key Generation Rate (KGR) of our signals, resulting from an 87.98% and 34.00% Bit Agreement Rate (BAR) for intra and inter-body signals, respectively.