A secure fuzzy extractor based biometric key authentication scheme for body sensor network in Internet of Medical Things

Abstract

Body sensor network (BSN) is largely utilized in IoMT to attain easier access of patient’s data remotely without much cost by connecting the various bio-sensors. However, there is a security threat in accessing the BSN because of hacking issues. Hence, In order to secure the most sensitive details of the patient, secured fuzzy extractor combined with fuzzy vault is developed in this approach with an objective of providing more security using bio-metric key authentication scheme. Initially, preprocessing is done to remove the noise in ECG signal using the adaptive filtering. Secured Fuzzy extractor is designed for extracting the features like QRS, PR and QT interval and the private key generation for authentication process is based on these features only. Because of unique features of each ECG, private key cannot be hacked easily. Along with private key values, a set of random chaff points are generated using polynomial construction principle and are stored with a checksum vector in a separate fuzzy vault set. In the authentication phase, the data in the fuzzy set will be checked with checksum values to detect missing data in the communication. The IP address of the device will be used as the public key for estimating the bit rate of sensors at decoding phase. The security of the system depends mainly on the hash function. In our proposed method, the hash variable value is independent of the hash functions to enhance the network security. This variation in hashing does not affect the latency or delay of our proposed technique. The proposed fuzzy extractor based biometric key authentication scheme achieved enhanced results such as 40% reduced data loss, 20% reduced energy consumption and reduced delay when compared to previous encoding techniques.