Developing IoT Performance in Healthcare Through the Integration of Machine Learning and Software-Defined Networking (SDN)

Abstract

In this era of rapid growth of the Internet of Things (IoT) in healthcare, the networking solution must be robust and efficient for smooth data transfer and the network. It is very well established that Deep Learning (DL) is a subset of Machine Learning (ML) and has shown great promise in enhancing network performance through intelligent data analysis and making smart decisions. The wide deployment of Wireless sensor networks and Mobile ad hoc networks (MANET) is why they are used in healthcare applications, as they offer low-cost deployment and real-time data exchange. Nevertheless, WSNs suffer limitations in energy availability due to their inherent constraints. However, the energy consumption in a WSN routing is a significant problem because of resource limitations. Integrating ML in a Software Defined Networking (SDN) framework shows the possibility of network resource optimization and a sustainable network architecture. Typically, conventional routing annoys most about shortest path algorithms, where nodes are used unevenly, and critical nodes die early from battery depletion. This work proposes a new DRL-based algorithm for maximizing traffic distribution and maintaining energy consumption equilibrium across the nodes in WSNs to improve their lifetime. The model proposes dynamically changing routing paths to preserve the nodes’ loading with a limited power supply system to persist the network lifespan and thus improve efficiency. However, the approach adds some hops to the average packet transmissions per hop. Still, it delays the first node’s energy depletion long enough to provide sustained network operability in IoT-driven healthcare environments. The experimental results validate the effectiveness of the proposed method in furthering the network functionality and reliability for data transmission, thereby making it a viable solution for the next generation of IoT healthcare networks.