An Innovative Intrusion Detection System Model Aimed at Drone Nodes Networks Threatened by DoS Attacks

Abstract

A very important branch of IoT is ad-hoc mobile networks, where sensor networks move in a given space and have been created to operate without a specialized infrastructure. However, there is a branch of this technology that involves unmanned vehicles, and can be divided into two sub-branches: Vehicle Ad-hoc Networks and Flying Vehicle Ad-hoc Networks. There, end node security becomes paramount. This paper´s objective proposes detection time as a metric to measure the impact that a Denial-of-Service, (DoS), attack could have, even with an Intrusion Detection System, (IDS), operating on the network. Furthermore, the importance of developing an IDS that revolves around false positives, and how this could affect the entire network system is emphasized. Likewise, a model is proposed and described to detect DoS attacks from the security approach of the end node, whereas, instead of starting to track the attack, the supposed node being attacked is secured, protecting it without interrupting its operations and subsequently confirming the attack to be identified. In the future, we intend to explain the correlation between time detection and security.

Keywords

• safety before detection
• iot
• machine learning
• intruder detection system
• drone networks