Oppositional poor and rich optimization with deep learning enabled secure internet of drone communication system.

• Develop a secure communication system for internet of drones environment. • Derive OPRFS-ODFNN model for intrusion detection in IoD environment. • Present an oppositional poor and rich optimization based feature selection. • Apply optimal deep feed-forward neural network for intrusion detection. As a result of technological advancements and the need for continuous reduction in manufacturing costs, the concept of Internet of Things (IoT), consisting of Unmanned Aerial Vehicles (UAVs), has entered the industrial production units. IoT devices have not only penetrated the day-to-day activities of human beings, but also in defence. In recent years, there is a widespread application of Internet of Drones (IoD) in areas such as television and film shooting, meteorological monitoring, forest fire detection, agricultural monitoring, emergency rescue, etc. In this background, Intrusion Detection System (IDS) plays an important role to effectually secure the IoD network. The current research work develops an Opposition Poor and Rich Optimization-based Feature Selection with Optimal Deep Feed-forward Neural Network (OPRFS-ODFNN) model for intrusion detection in IoD communication system. The aim of the presented OPRFS-ODFNN technique is to accomplish enhanced security in IoD communication system. In order to achieve the objective, OPRFS-ODFNN model initially executes feature scaling as a pre-processing step. Then, OPRFS technique is applied for effective selection of the features. Moreover, Improved Mayfly Optimization (IMFO) is applied with ODFNN model for intrusion detection and classification processes. In order to validate the enhanced performance of the proposed OPRFS-ODFNN method, extensive simulations were conducted. [Display omitted] [ABSTRACT FROM AUTHOR]