An Improved IDS Method Using the Giza Pyramids in IoT Networks

The Internet of Things (IoT) is an intelligent global network that connects physical and virtual objects, leading to significant advancements across various fields. However, its rapid growth and increasing adoption have brought about serious security challenges. The resource limitations in IoT nodes, designing comprehensive and efficient security mechanisms in this area is a big challenge. As a result, implementing a security mechanism like an intrusion detection system (IDS) is crucial for identifying and monitoring unauthorized access attempts, ensuring protection against security threats and vulnerabilities in IoT networks. Machine learning serves as a powerful technology capable of accurately detecting cyberattacks and safeguarding systems from harm. The large volume of data generated within IoT environments poses a challenge to the real-time performance of IDS, requiring swift and precise responses to hacking and malicious activities. This paper proposes an approach to enhance the performance of support vector machines (SVMs) using the Giza pyramid construction algorithm, aiming to improve both accuracy and speed in detecting intrusions while reducing detection errors. The findings demonstrate that the proposed model outperforms classifiers like k-nearest neighbor and ensemble in terms of error rates across training, testing, and overall datasets.

کلیدواژه ها:

Internet of Things ، Security ، Intrusion Detection System ، Support Vector Machines