Optimizing IoT Data Transmission in Smart Agriculture: A Comparative Study of Reduction Techniques

Smart agriculture, driven by Internet of Things (IoT) technologies, has revolutionized traditional farming practices. The integration of sensors and devices in the agricultural environment has generated a vast amount of data for analysis and decision-making. However, managing and transferring this abundant data poses challenges, particularly in terms of data transfer and storage. Therefore, the purpose of this study is to investigate and compare different data reduction techniques for optimizing IoT data transmission in smart agriculture. These techniques aim to reduce the size or quantity of transmitted data samples, leading to reduced energy consumption, improved network efficiency, and minimized storage and processing requirements. The selected data reduction techniques include sampling, quantization, and deduplication. The evaluation of these techniques was conducted using a real-time dataset specifically designed for smart agriculture systems. The dataset comprises five features: ambient air temperature, ambient air moisture, soil temperature, soil moisture, and soil pH. The comparison of the techniques was based on three metrics: information loss, computation cost, and energy consumption, taking into account the size of each transmitted sample. Python programming language was employed to implement the three data reduction techniques. The results revealed that there exists a trade-off among information loss, computational cost, and energy consumption, which necessitates careful consideration when selecting a data reduction technique for IoT nodes in smart agriculture applications. This analysis provides valuable insights for making informed decisions that strike a balance between these factors, considering the specific requirements and constraints of each application.