A comparative analysis on traditional wired datasets and the need for wireless datasets for IoT wireless intrusion detection

<span>IoT networks mostly rely on wireless mediums for communication, and due to that, they are very susceptible intrusions. And the tiny nature, processing complexity, limited storage capacities, IoT require reliable intrusion detection systems (IDS). Although there many IDS types of research available in literature, most these suitable wired network environments, benchmark datasets used works relying such as KDD Cup’99 NSL-KDD. distinct nature give more emphasis data link layer physical layer. These concerns not given much attention traditional body knowledge. Therefore, this research, an system is developed using a newly dataset (NaBIoT) literature with focusing common related attacks, layers taken into consideration. The evaluated by comparing various machine learning algorithms terms evaluation metrics accuracy, F1 score, false positive, negative. Moreover, compared against NSL-KDD evaluate need datasets. NaBIoT show its effectiveness detecting Besides simulation performed different combinations features conclude that certain primary does all perform detection. This can reduce time mainly creating models. results have proposed some critically important be eliminating features. </span>