ENERGY-EFFICIENT IOT ARCHITECTURE USING EDGE INTELLIGENCE FOR SCALABLE ENVIRONMENTAL MONITORING APPLICATIONS DEPLOYMENT

Abstract

As the number of environmental monitoring applications continues to expand rapidly, the need for scalable and energy efficient Internet of Things (IoT) architectures with continuous sensing, real-time analytics and low power operations grows.The proliferation of environmental monitoring applications has increased the demand for energy efficient and scalable IoT architectures that allow for continuous sensing, real-time analytics and low power operation. Existing cloud-based IoT systems have issues like latency, high energy consumption and network congestion, which can hinder large-scale deployments. In this paper, an energy-efficient IoT architecture with the incorporation of edge intelligence solutions is proposed for scalable applications to monitor the environment. The proposed framework is designed as a layered system: sensing, edge, communication, and cloud, to allow for processing to be done at the edge, and to optimise the use of resources when doing so. Lightweight machine learning models, such as Random Forest (RF), Convolutional Neural Network (CNN), and Long Short-Term Memory (LSTM) are embedded into Edge intelligence for making local decisions, anomaly detection and predictive analytics. To minimize transmission overhead and extend the life of sensor nodes, energy optimization methods like duty cycling, adaptive communication protocols, data aggregation, edge-based filtering and compression methods are used. The framework aims to facilitate the environmental monitoring in smart city and industrial applications in a reliable, real-time and sustainable manner.