Smart IoT Application in Soil Moisture and Heat Level Prediction Using Sine Cosine-Horse Herd Optimized Deep Learning

Abstract

Agriculture assisted by Internet of Things (IoT) is termed as smart agriculture, which offers an increase in precision farming. Soil monitoring with IoT technology helps in the increase of agriculture by growing the yield through measuring accurate soil content information, like temperature, nutrition content, humidity, potential of hydrogen (PH), moisture and so on. In this research, the soil moisture and heat level is measured through an optimized deep learning technique namely, Sine Cosine Horse Herd optimization-based Deep Recurrent Neural Network (SCHHO-based Deep RNN). Here, the moisture and heat level is predicted using Deep RNN in which its weights are trained using SCHHO. In order to progress the effectiveness of prediction, the feature selection is done prior to prediction for choosing the appropriate features using weighted correlation coefficient. In addition, the gathered soil information is transmitted to the IoT nodes using SCHHO routing algorithm by considering fitness measures. Besides, the experimental outcome proves that the SCHHO-based Deep RNN algorithm provides better performance with the accuracy and precision of 0.918 and 0.908, respectively.