Using machine learning for security issues in cognitive IoT
Wafa Alzuabi, Wael Mohamed Elmedany, Mhd Sharif Saeed · IET conference proceedings. · 2024
Cognitive learning is progressively prospering in the field of Internet of Things (IoT). With the advancement in IoT, data generation rate has also increased, whereas issues like performance, attacks on the data, security of the data, and inadequate data resources are yet to be resolved. Recent studies are mostly focusing on the security of the data which can be handled by machine learning. Security and privacy of devices intrusion detection their success in achieving classification accuracy, machine deep learning with intrusion detection systems have greatly increased popularity. However, the need to store communication centralized server compromise privacy and security. Contrast, Federated Learning (FL) fits appropriately as a privacy-preserving decentralized learning technique that trains locally transfer the parameters the centralized instead of purpose current research provide thorough and application FL intrusion detection systems. Machine Learning (ML) and Deep Learning (DL) approaches, which may embed intelligence in IoT devices and networks, can help to overcome a variety of security challenges. The research includes a detailed overview of the application of FL in several anomaly detection domains. In addition, it increases understanding of ML and its application to the field of the Cognitive Internet of Things (CIoT). This endeavour also includes something crucial . The relevant FL implementation issues are also noted, revealing potential areas for further research. The researcher emphasised the flaws in current security remedies, which call for ML and DL methods. The report goes into great detail on how ML and DL are now being utilised to help handle various security issues that IoT networks are facing. Random Neural Networks that have been trained using data retrieved by Cognitive Packets make the routing decisions. A number of potential future directions for ML and DL-based IoT security research are also included in the study. The report concludes by outlining workable responses to the problem. The paper closes by offering a beginning point for future study, describing workable answers to the problem of FL-based intrusion detection system implementation.