Effective secure data transmission and attack detection using advanced encryption standard (AES) integrated with adaptive divergence weight firefly algorithm (ADWFA) and the bi-directional gated recurrent unit classifier algorithm over IoT-based WSN

Wireless Sensor Networks (WSNs) are used in many Internet of Things (IoT) applications and are highly susceptible to various cyberattacks such as denial of service (DoS) and other types of cyberattacks, which affect the availability and reliability of the network and compromise the privacy of the data transmitted over the network. This article proposes a novel framework called AES-ADWFA-BiGRU for secure and energy-efficient attack detection in IoT-based WSNs. The proposed framework uses Adaptive Divergence-Weight Firefly Algorithm (ADWFA) for optimal Cluster Head (CH) selection to improve network lifetime and energy efficiency; Advanced Encryption Standard (AES) for secure data transmission at the node and CH levels; and Bidirectional Gated Recurrent Unit (BiGRU) for accurate intrusion detection. K-Means Clustering (KMC) is used for preprocessing the network data; missing values are handled using the centroid-based imputation method of KMC. AES ensures secure data transmission and privacy of the data transmitted over the network; ADWFA improves network lifetime and energy efficiency; and BiGRU is used for accurate intrusion detection in WSNs. The proposed framework is implemented and validated on the NSL-KDD dataset; the results show improved accuracy, precision, recall, and F-measure compared to other methods; and the framework is effective for secure and efficient WSN applications based on the IoT paradigm.