Deep-transfer learning framework in SDN for gateway ports security

Abstract

Gateway port security with validation method is an important part in Software Defined Networks (SDN). Using soil-adjusted vegetation index (SAVI) and SDN in one salable network such as Internet of Things (IoT) and Wireless Sensor Network (WSN) meets new challenges in infrastructure and architecture. The main challenges are complexity of deploying SAVI and obtaining security data at the access layer. The use of combined deep-transfer learning model to allow the validation of the appropriate source address at downstream network input ports in SDN. This research aims to provide a secure gateway structure with the maximum privacy of data in the IoT-SDN environment. The use of encryption mechanisms including Datagram Transport Layer Security (DTLS) and Frequency shift keying (FSK) is considered as the main approach used in the Message Queuing Telemetry Transport (MQTT) protocol structure for sending and receiving data. Improving the quality of services includes delays, throughput, and some others to manage the resources of the network. The results represented that the proposed approach has better performance in comparison to others.