Deep Learning Based Attack Detection and QoS Aware Secure Routing Protocol for SDN-IoT Network

Abstract

The IoT network and the software-based controller comprise the Software-Defined Network-based IoT (SDN-IoT). SDN-IoT is widely utilized in traffic control and management, smart buildings and homes, safety applications, health care monitoring and automation, and several application domains. The most challenging aspects of the SDN-IoT network scenario are efficient routing and security issues. Hence, a novel framework of SDN-IoT with deep learning-based attack detection and a QoS-aware secure routing protocol is proposed. The attack detection is employed initially for the user request using the Deep Belief Network (DBN). The detected malicious request is dropped, and the routing is developed for the normal data packet. Here, African Aquila Optimization is introduced for efficient routing based on secure QoS factors. The analysis of the proposed method based on the assessment measures such as end-to-end delay, energy consumption, network lifetime, packet delivery ratio, and throughput acquired the values of 1.86, 5.22, 1632.84, 0.9959, and 0.96, respectively.