Data congestion control framework in Wireless Sensor Network in IoT enabled intelligent transportation system

Abstract

Intelligent Transportation System (ITS) holds an inevitable concern in road safety and efficient transportation. Data communication is enforced by wireless sensor nodes and is compatible with traffic monitoring and control capabilities. Congestion in such a system will carry off serious constraints and effects on the intelligent transportation system. Congestion problems can severely limit the performance of Wireless Sensor Network (WSN)-based IoT, resulting in higher packet loss ratios, longer delays, and lower throughputs. To resolve such constraints, a novel particle swarm optimization algorithm-based Dynamic deep neural network (DDNN-PSO) is proposed. To enhance the DDNN performance, its weight parameters are optimized using the PSO algorithm. The performance analysis of the proposed DDNN-PSO is performed by estimating the Delivery ratio, Packet delay, Throughput, Overhead, and Energy consumption with the existing Genetic Algorithm based DNN (DNN-GA) and DNN techniques. The experimental findings show that the proposed DDNN-PSO surpasses models such as DNN and DNN-GA. The proposed method has an overall performance of 5.69% and 8.01% better than DNN-GA and DNN respectively.

Keywords

• Intelligent transportation system
• Deep neural network
• Congestion avoidance
• Particle swarm optimization
• WSN-Based IoT