MERPAL: multicollinearity regressive multilayer perceptron-based traffic-aware scheme for IoT-enabled smart cities

The evolution of smart cities, driven by advancements in information transmission and the fusion of sensor technologies with IoT, faces challenges from substantial data traffic affecting Quality of Service factors like throughput and causing data loss and delivery delays. In order to improve traffic-aware data transmission in IoT-based smart city applications, a novel deep learning technique called the Multicollinearity Deming Regression-based Deep Perceptive NeurAl Learning Classifier (MERPAL) method is introduced. The proposed Deep Multilayer Perceptive Classifier consists of multiple layers such as input, output, and a number of hidden layers for assessing the given input. The number of nodes taken as input is passed to the primary hidden layer, where the energy and bandwidth availability of the nodes are assessed. The second hidden layer then uses the Multicollinearity Deming Regression to evaluate the estimated energy level and bandwidth of the nodes. After that, the analyzed results are given to the soft step activation function in the third hidden layer. The activation function identifies the best nodes based on energy level and bandwidth availability for efficient data delivery and helps minimize the network traffic. The simulation results indicate that MERPAL demonstrates superior performance compared to existing methods, exhibiting a 7% enhancement in data packet delivery ratio, a 15% increase in throughput, along with a reduction of 20% in packet loss and 14% in delay.