Enhancement of IoT device security using an Improved Elliptic Curve Cryptography algorithm and malware detection utilizing deep LSTM

Internet of things (IoT) has become more popular due to the development and potential of smart technology aspects. Security concerns against IoT infrastructure, applications, and devices have grown along with the need for IoT technologies. Enhanced system security protocols are difficult due to the diverse capabilities of IoT devices and the dynamic, ever-changing environment, and simply applying basic security requirements is dangerous. Therefore, this proposed work designs a malware detection and prevention approach for secure data transmission among IoT gadgets. The malware detection approach is designed with the aid of a deep learning approach. The initial process is identifying attack nodes from normal nodes through a trust value using contextual features. After discovering attack nodes, these are considered for predicting different kinds of attacks present in the network, while some preprocessing and feature extraction strategies are applied for effective classification. The Deep LSTM classifier is applied for this malware detection approach. Once completed malware detection, prevention is performed with the help of the Improved Elliptic Curve Cryptography (IECC) algorithm. A hybrid MA-BW optimization is adopted for selecting the optimal key during transmission. Python 3.8 software is used to test the performance of the proposed approach, and several existing techniques are considered to evaluate its performance. The proposed approach obtained 95% of accuracy, 5% of error value and 92% of precision. In addition, the improved ECC algorithm is also compared with some existing algorithm which takes 6.02 s of execution time. Compared to the other methods, the proposed approach provides better security to IoT gadgets during data transmission.