Data Privacy Preservation and Authentication Scheme for Secured IoMT Communication Using Enhanced Heuristic Approach with Deep Learning

Abstract

Internet of Medical Things (IoMT) is a promising field that is widely used in healthcare applications nowadays. The IoMT is an extension of the Internet of Things (IoT) that is utilized for the processing, generation, collection, and evaluation of medical data. However, the most concerning issue in IoMT is regarding the privacy and protection of the IoMT data. Because privacy preservation of information is the major concern in IoT-based healthcare systems. Hence, there is a need for a trustworthy end-to-end system, which can tolerate even insider attacks. The core aim of this research work is to promote the latest secure technique for transmitting the data from sink nodes to the server and also to ensure the security level of communication between them. Here, the gathered sink node is authenticated by deep hybrid methods of combining the Auto Encoder (AE) and Bidirectional Long Short-Term Memory (BiLSTM). This hybrid model is named as Auto Encoder-Bidirectional Long Short-Term Memory (AE-Bi-LSTM), in which the parameters are tuned by the Improved Chimp Optimization Algorithm (IChoA). Hence, data privacy preservation makes to secure communication in the networks. The proposed privacy-prevention model is utilized for maintaining communication between sink nodes and servers. The communication is done through three different steps (a) Sanitation, (b) Optimal key generation, and (c) Restoration. Moreover, from the overall result analysis, the accuracy and precision rate of the designed IChoA-AE-Bi-LSTM approach are 95.58 and 91.93%. The experimental analysis shows that the designed model offers a better authentication scheme and guarantees data privacy than traditional models. Moreover, applying a high amount of energy will reduce the lifetime of the sensor nodes. Here, privacy preservation is a challenging task due to fraudulent attacks from third-party service providers. In the future, advanced techniques will be implemented with lightweight encoded mechanisms.