Data Reduction Technique for Smart Agriculture in Wireless Sensor Networks using Machine Learning

Abstract

Smart farming is an advanced approach that combines technologies like the IoT, automation, drones, and artificial intellect to boost both the harvest and quality of farm products while tumbling manual work. By applying a wireless sensor network, which delivers self-governing energy, tracks valve and shift functionality, and achieves remote locations, high-quality crops can be produced consistently throughout the year. Wireless sensor networks, designed to collect data from all sensors with low energy consumption and extensive communication ranges, are a fundamental element of IoT. This paper proposes a system for monitoring soil moisture, temperature, and humidity in small-scale farms. Transmitting large quantities of data can lead to high energy and bandwidth consumption on the sensor nodes. To mitigate this, a machine learning algorithm is introduced, aimed at reducing data using the Data Reduction Algorithm (MLDR). MLDR’s purpose is to efficiently gather commercial data. It works as a technique for dimensionality reduction, utilizing machine learning at the sensor network interface to minimize the amount of data sent to the central system while ensuring accuracy and relevancy."