Artificial Intelligence enabled Dynamic Bandwidth Allocation Optimization in IoT for improved QoS

Abstract

The Internet of Things (IoT) is a collection of intelligently connected devices that are connect to the internet. Every day, as technology advances, an IoT device outnumbers humans on the world. The Internet of Things (IoT) systems would connect not only physical users and devices, but also virtual networks and devices that have already been implemented in various scenarios using new networking, computing, and integrated system technologies. With the rapid advancement of technology, the number of devices involved grows rapidly. IoT systems have crucial issues in supplying sufficient bandwidth to IoT Devices for Improving Quality in Services provided in the form of QoS (Quality of Service) parameters, depending on the volume of devices and data. The proposed work is an artificial intelligence enabled dynamic bandwidth allocation optimization in IoT for improved QoS comprising of IoT devices, router, bandwidth data, KM-PSO clustering, Deep Alternative Neural Network (DANN), Dynamic Bandwidth Allocation Optimization, and QoS Estimation. It provides a scalable solution for allocating the bandwidth for the IoT enabled devices for improving QoS parameters. The work uses K-Means-Particle Swarm Optimization (KM-PSO) for generating the optimum number of clusters for improving the accuracy of the proposed work. Once the optimal number of clusters are generated, a Deep Alternative Neural Network (DANN) used in the present invention detects the traffic from the IoT devices, predicts the required bandwidth to be allocated. The Artificial Intelligence enabled Dynamic Bandwidth Allocation Optimization (AIEDBAO) uses Fuzzy Logic Controller (FLC) for allocating Bandwidth based on the traffic rules generated. The improved QoS parameters achieved with the present invention are Bandwidth Consumption of 34.5Mbps, Packet Delivery Delay (PDD) of 32ms, Packet Loss of 4% and Throughput of 99% at supply of 50Mbps