An intelligent forecasting system in Internet of Agriculture Things sensor network

Irrigation refers to the process of supplying water to the soil via pumps and spraying water across the field. In conventional agriculture, the installation and operation of an irrigation system depend exclusively on the farmers' knowledge. The Food and Agriculture Organisation (FAO) has forecasted that by 2030, emerging nations will increase their irrigated areas by 34 %, although water use will rise by only 14%. Thus, the necessity of constantly monitoring water flow and volume, rather than depending on people's approximations, is highlighted by this variation. Therefore, this research proposes an efficient prediction system for intelligent irrigation in the Internet of Agriculture Things (IoAT) sensor network. In the initial phase of the suggested approach, agricultural sensor data is transmitted to the total variation regularisation in the alternate direction method of multipliers (ADMM) way to mitigate the adverse effects caused by noisy samples. Subsequently, probabilistic clustering is utilised to address the missing entries. During the second phase of the suggested algorithm, the Lagrangian L1 point is extracted from the sensory data, which is followed by the extraction of the maximum plausible row used for forecasting. The experimental evaluation is conducted on the massive amount of agriculture sensor data sets and cross-validation using several matrices proves its efficacy over competing approaches.