Soil monitoring and evaluation system using EDL‐ASQE: Enhanced deep learning model for IoT smart agriculture network. (24th May 2021)

• Record Type: Journal Article
• Authors: Sumathi, P; Subramanian, R; Karthikeyan, VV; Karthik, S
• Is Part Of: International journal of communication systems. Volume 34:Number 11(2021)
• Page Start: n/a
• Page End: n/a
• Publication Date: 2021-05-24
• Subjects: agriculture IoT network -- enhanced deep learning -- soil moisture -- soil quality -- weight factor; Telecommunication systems -- Periodicals; 621.382
• DOI: 10.1002/dac.4859
• Languages: English
• ISSNs: 1074-5351

Abstract

Summary: The enormous growth of the Internet of Things (IoT) network provides abundant support to agriculture and development, which states the future scope of IoT‐based agriculture. In a recent scenario, agriculture IoT can be integrated with sensors, communication protocols, and microcontrollers for automated process executions to increase productivity. Moreover, deep learning effectiveness produces appropriate results and solves several real‐time issues related to agriculture‐based advancements. The proposed system presents the design of an IoT network communication system to estimate the soil conditions. Soil quality is an important factor in modernized agriculture, productivity enhancement, and hydrological cycles. By the soil quality analysis, the accurate prediction is very significant for sensible usage of resources. An enhanced deep learning model for IoT network‐based automated soil quality evaluation observes the complex soil features and meteorological factors with those concerns. Here, the real‐time samples are collected from the local area sensor network for analysis. The deep learning model is developed with big data fitting ability for soil quality prediction. The weight factors (W.F.) are derived for measuring the soil quality accurately. The proposed IoT network‐based agriculture structure allows a flexible approach to different types of crops and implementation in agricultural areas. Experimental results obtained in the laboratory and onsite confirmed the performance and reliability of the system. The result evaluations are carried out based on precision, accuracy, and processing time, and results show that the model achieves better results than compared models. Abstract : The architecture contains an input layer, an output layer, and multiple hidden layers, in which the terminals are fully connected. The numbers of hidden layers are determined based on the data set ranges. Based on that, the activation functions are also derived. And the mathematical computations are described.