Enhancement in localization routing and security based on internet of things

Abstract

Due to rapid development of Wireless Sensor Networks (WSNs) based communication in many areas such as surveillance of traffic, industry monitoring, healthcare system and military affairs etc. In order to address this type of communication, Internet of Things (IoT) has been widely used due to their properties. The working of the IoTis similar to a heterogeneous wireless WSNs, MANETs, Zig-Bee, WI-FI, and RFID. To make better communication in IoTnetworks, data forwarding routing mechanism plays an imperative responsibility for Device-to-Device (D2D) secure data packet transmission, but most of the research fails to deal with security problems. Nowadays, to handle with security issues, trust based routing is a difficult task during the D2D communication. So, in this research, we try to solve the existing routing and security problem by designing a Secure and Energy Efficient Trust Aware (SEETA) routing mechanism with Particle Swarm Optimization (PSO) based Convolutional Neural Network (CNN) to enhance thelocalization, routing and security factors forIoT network. Here, SEETA routing protocol is capable to detect the fail/malicious nodes based on their properties within the network by utilizing the PSO based CNN as a classifier. In order to address the detection probleminIoT communication, in this research, we planned to utilize the concept of artificial intelligence approach which helps to identify the nodes behaviours during communication and decide to discover a secure and efficient route for data transmission. We proposed two different scenario for the development of simulation model such as by utilizing Artificial Neural Network (ANN) and second is the model with CNN. In this research, we find out the performance of the IoTnetwork using SEETA with PSO based optimized CNN is better as compare to model with ANN. We also précises the trusted routing with maintenance in order to counter the adversaries which follow certain attack or malicious patterns along with optimized CNN. When the proposed IoT network is simulated on different network conditions, QoS parameters is calculated and compared with a few other state-of-art methods and we obtained the proposed IoT network with optimized CNN, achieves the best performance in terms of Throughput, Loss Rate, Energy Consumption Rate, Number of Alive Nodes, End to End Delay and Detection Rate. newline