User Behavior Based Secure Smart Wearable Device Pairing

Smart wearable devices have been widely used in various fields,including social,medical,financial,etc.,which leads to frequent information sharing between devices.Since applications in these fields are likely to involve privacy and sensitive information,secure device pairing is required to establish a secure data transmission channel before information sharing between devices,guaranteeing the security of the transmitted information.Different from traditional mobile devices,the secure smart wearable device pairing is usually limited by input and output interface and network environment,which increases the difficulty of the interface interaction and pre-deployment.User behavior based secure smart wearable device pairing schemes needs only simple human-computer interactions,and is plug-and-play,which can solve above problems.However,most of the existing researches are devoted to secure device pairing on the same individual,which is not suitable for multi-party applications such as sharing schedules,exchanging social business cards,transmitting photos and streaming media,submitting health data and sports information,and even trading.To this end,committed to secure smart wearable device pairing on different individuals,we design secure device pairing schemes based on handshake.In this paper,the threat model is first proposed.Then,two different secure device pairing schemes are designed,which are based on the proposed threat model and respectively focus on device-friendliness and userfriendliness.The device-friendly device pairing scheme aims to reduce computing overhead,thereby reducing the burden of the device;while the user-friendly device pairing scheme is designed to shorten user interaction time and simplify user operations.Specific contributions of this paper are summarized as follows:1.For device-friendliness,a secure smart wrist-worn-device pairing scheme based on fuzzy cryptography is proposed.Specifically,the device first preprocesses the collected handshake acceleration and extracts features from it.Then,pre-association is designed which helps device adaptively initiate pairing attempt with the optimal device to avoid repeated pairing attempts,thereby reducing computing consumptions.Finally,key negotiation is performed based on fuzzy cryptography,implementing the secure device pairing.A new perturbation vector based fuzzy cryptography PVFC is designed,which uses perturbation vectors to recover the noise data,therefore has high calculation efficiency.The simulation experiment results of the PVFC-based secure device pairing scheme show that PVFC has better performance in computing and success rate compared with previous fuzzy cryptography;the PVFC-based secure device pairing scheme has high accuracy and is secure under the proposed threat model.2.For user-friendliness,a secure device pairing scheme based on optimal feature selection and fuzzy commitment is proposed.Specifically,first,the time series of handshake acceleration magnitude is taken as the feature sequence and be quantified with the random thresholds.Then,the features are analyzed,and an optimal feature selection algorithm is proposed to filter features that are prone to bias in the quantization process.Besides,the negotiation of a symmetric key between devices based on fuzzy commitment is proposed.Finally,the identities of the paired devices are verified through multiple verifications.Both experiment and theory prove that the scheme shows high accuracy and security.And compared with existing schemes,the time of the handshake that our scheme needs is shorter,therefore the scheme is more user-friendly.