Özet:

Abstract Intelligent towns and cities, medical care, and automation in industries have all reaped countless benefits from the Internet of Things (IoT) devices quick proliferation. Although anomaly detection is essential for safeguarding the integrity and dependability of IoT systems, the widespread implementation of connected devices also brings new security problems. Bird Swarm-Optimized artificial neural networks (BSO-ANN) are a unique anomaly detection framework presented in this study for IoT contexts. The BSO approach enables the model to look for an ideal network configuration that improves anomaly detection accuracy by mimicking the collective actions of birds in a swarm. The BSO-ANN approach's ability to detect anomalies was assessed using the UNSW-NB15 dataset. The findings show that the BSO-ANN algorithm detects several types of irregularities in IoT systems with impressive precision, recall, accuracy, and f1-measure parameters. This work can act as a basis for creating sophisticated anomaly detection methods to defend IoT networks against new security risks.

Anahtar Kelimeler: