Improving weather dependent zone specific irrigation control scheme in IoT and big data enabled self driven precision agriculture mechanism

Precision agriculture involves manipulation of variations in field productivity, maximization of income, scale backing of wastes, and minimizing of the impact on surroundings using automated machine-controlled information assortment and documentation. This work focuses on the efficient control of farm irrigation by exploiting the capabilities of Internet of Things (IoT) and Big Data-based Decision Support System (DSS) to generate adequate valve control commands. Three varieties of prediction techniques such as Deep Neural Network (DNN), Random Forest (RF) and Resilient Back-Propagation Neural Network model are tested to predict soil Moisture Content (MC) in one hour advance by considering 6 numbers of different sensors. The real-time data collection is performed using the proposed IoT node deployment strategy tested in the field. An integrated IoT-based DSS framework is proposed to accumulate 17 numbers of soil and environmental parameters to predict future variation of soil MC in 1 h advance. Further, Structural Similarity (SSIM) Index is used to visualize and maintain uniform MC all over the agriculture area during the entire cropping period. Site and zone specific irrigation control scheme is tested in the test site using fuzzy logic-based weather dependent model. The complete system architecture, deployment strategy and performance of the proposed IoT-based DSS mechanism is discussed in this article.