Intelligent Cyber-Security System for IoT-Aided Drones Using Voting Classifier

Rizwan Majeed,Muhammad Umer,Muhammad Faheem Mushtaq,Michele Nappi,Nurul Azma Abdullah

doi:10.3390/electronics10232926

Rizwan Majeed, Muhammad Umer + Show 3 more

Open Access

https://doi.org/10.3390/electronics10232926

Copy DOI

Abstract

Developments in drones have opened new trends and opportunities in different fields, particularly in small drones. Drones provide interlocation services for navigation, and this interlink is provided by the Internet of Things (IoT). However, architectural issues make drone networks vulnerable to privacy and security threats. It is critical to provide a safe and secure network to acquire desired performance. Small drones are finding new paths for progress in the civil and defense industries, but also posing new challenges for security and privacy as well. The basic design of the small drone requires a modification in its data transformation and data privacy mechanisms, and it is not yet fulfilling domain requirements. This paper aims to investigate recent privacy and security trends that are affecting the Internet of Drones (IoD). This study also highlights the need for a safe and secure drone network that is free from interceptions and intrusions. The proposed framework mitigates the cyber security threats by employing intelligent machine learning models in the design of IoT-aided drones by making them secure and adaptable. Finally, the proposed model is evaluated on a benchmark dataset and shows robust results.

Highlights

Pervasive environments have become increasingly popular over the last two decades, especially considering the importance of pervasiveness and the smartness of the objects present in an environment such as a building, a town, a playground, a shopping mall, etc
The results are displayed for the mobile system consisting of the security status of the drones and the Internet of Things (IoT) network identified using machine learning
The analysis of the results reveals that the machine learning models as well as a simple deep learning model have shown considerable results for intrusion detection on the drone dataset

Summary

Introduction

Pervasive environments have become increasingly popular over the last two decades, especially considering the importance of pervasiveness and the smartness of the objects present in an environment such as a building, a town, a playground, a shopping mall, etc. A benefit of these small-sized drones is that they can enter a building and can hover inside a building for the sake of monitoring and surveillance in various fields of life, such as the surveillance of industrial areas [1,2,3] and for disaster management, military uses [4], search and rescue [5,6] shipping and delivery [7], precision agriculture [8,9], and many other applications. The security and privacy of these drones are an open challenge [13] In recent times, another emerging field of research has been the attempt to make drones smart by adding a few sensors that a small drone may be sufficiently capable of carrying. A set of devices such as sensors, transmitters, and cameras for many different and complex applications can make drones more useful and effective [14]

Objectives

Methods

Results

Conclusion