Research On Wireless Body Area Network Identity Authentication Based On Iris And ECG Signals

As one of the typical applications of the Internet of Things,wireless body area network plays an increasingly important role in people’s lives.Among them,telemedicine technology relies on wireless body area network and uses wearable or implantable medical sensors to allow users to remotely transmit physiological parameters to doctors for analysis and receive medical advice remotely without leaving home,which greatly facilitates users.However,it is precisely because these sensors transmit or store a large number of users’ health privacy data.Once invaded by illegal users,it will pose a serious threat to the safety of users’ lives.Allowing only legitimate users to access is a prerequisite for ensuring the security of wireless body area networks.Therefore,the authentication mechanism between medical sensors that master the user’s first hand privacy data is a key research content when designing a wireless body area network.However,medical sensors generally have the disadvantage of limited resources,and traditional identity authentication mechanisms are often infeasible in wireless body area network scenarios.How to design an appropriate authentication mechanism for wireless body area networks is a challenging issue.Therefore,this master degree thesis designs a new wireless body area network secondary identity authentication mechanism that combines iris features and ECG signal features,and carried out the following work.Firstly,This master degree thesis studies the security protection of wireless body area networks based on iris signals,and points out that the reason why few researchers use iris feature to protect wireless body area network is that the length of traditional iris code is too long,and in the resource-constrained wireless body area network not applicable.This master degree thesis proposes a new iris coding method: Iris MappingBlock Logic Coding(IMBLC),which combines interval mapping and block logic operation ideas in this field for the first time.Through experimental analysis,compared with traditional methods,IMBLC method can reduce the coding length by 98.3% at most,and has good random performance.Subsequently,this master degree thesis studied the characteristic coding of ECG signals in wireless body area network security protection,reproduced and analyzed six existing ECG signal coding methods,providing a reference for future researchers.On this basis,this master degree thesis also proposes a new ECG feature coding method:IPI Fusion Coding(IPIFSC).Experiments show that the IPIFSC method can still maintain excellent randomness under the premise of significantly shortening the coding time.The performance is significantly better than the existing 6 encoding methods.Finally,this master degree thesis combines relatively static iris features with dynamic ECG signal features to design a wireless body area network two-level identity authentication mechanism,and implements a key agreement mechanism based on fuzzy commitment.Due to the introduction of ECG signals,this mechanism solves the problem that iris feature recognition is still feasible after death.After experimental analysis,the false rejection rate FRR(False Rejection Rate)of the identity authentication mechanism proposed in this master degree thesis has reached 0.032481,and the false acceptance rate has reached 0.032481.FAR(False Acceptance Rate)is close to 0,the recognition performance is excellent,and it can resist 6 threats such as brute force attacks and replay attack,and has good security performance.