An intelligent attention based deep convoluted learning (IADCL) model for smart healthcare security.

In recent times, there has been rapid growth of technologies that have enabled smart infrastructures-IoT-powered smart grids, cities, and healthcare systems. But these resource-constrained IoT devices cannot be protected by existing security mechanisms against emerging cyber threats. The aim of the paper is to present an improved security for smart healthcare IoT systems by developing an architecture for IADCL. The proposed system employs publicly available datasets such as CIC-IDS 2017, CIC-IDS 2018, CIC-Bell DNS 2021, and NSL-KDD to present a robust detection framework. IRKO selects features, reducing the feature dimensions and hence isolating the most relevant attributes. The AConBN classifier then accurately classifies normal and intrusion traffic. Afterwards, optimization in the classification process is done by the SA-HHO algorithm, which provides the optimal weight values. Results are such that the IADCL framework detects cyberattacks with a high degree of accuracy, and the performance evaluations are made based on a number of key performance metrics. Conclusively, the proposed system has very good potential to protect smart healthcare IoT devices from cyber threats.