Energy Efficient Underground Wireless Sensor Networks for Disaster Analysis in Hill Stations

Abstract

The emerging growth of wireless sensor networks (WSN) systems provides various benefits over the data transmission at different environments. Remote monitoring systems are important to make continuous evaluation of environments, healthcare, safety, emergency response and smart analysis. Underground wireless sensor networks (UGWSN) are increasing in recent innovations for detection of various emergency condition and to provide early alert to the peoples. Even though UGWSN models face various challenges in terms of installations and maintenance, dynamically changing environment, sensor life span affected by environment, noise etc, power constraints act as important problem in having sustainability. The proposed approach considers an energy efficient underground wireless sensor network (UGWSN) using AI-optimized predictive neural computing (AI-OPNC) models early detection and analysis of disasters leads in hill regions. The primary goal enforced on detection of disaster triggers such as earth quakes, land slide dynamics and to analyse the occurrence stamps using AI-OPNC model, on the other hand the network health is monitored through network health keeper. The proposed system performance is measure in two phases such as network performance in terms of energy efficiency, data loss etc and the performance of AI-OPNC in terms of predictive analysis through accuracy. The proposed system achieved 95% accuracy; energy efficiency increases 10% with respect to data loss suppressed 8% using UGWSN nodes.