An Efficient and Secure Authentication Protocol with Deep Learning Based Key Generation toward Securing Healthcare Data in IoT

Due to the emergence of technologies in the growing world, the IoT is termed as a novel concept. The combination of internet with smart devices provides convenient user-related services. The attack toward confidential data may leak data and can cause serious security issues. Meanwhile, the classical security techniques cannot be utilized in a direct manner. Thus, novel techniques are needed for safeguarding the IoT from security issues. This paper designs secure authentication model for IoT in the healthcare. Several entities, such as gateway node, sensor node and medical professional are included for performing the secure authentication. The proposed protocol is devised by providing a mathematical model that makes usage of hashing function, encryption algorithm, XOR, Chebyshev polynomial, passwords, secret keys, and other security operations. Here, Secret key is generated with the deep neuro fuzzy network (DNFN) using Extreme Learning Machine (ELM) based sub-key generation. The proposed technique outperformed with less computation time of 33,484,154,235 ns, smallest memory of 51.750 MB, and highest detection rate of 94.9%.