Wireless Potato Root Tuber Sensing 2.0 (WPS 2.0)

This dataset is an extended version of the WPS dataset [1], which is proposed for underground potato tuber sensing. Compared with the previously published dataset, this dataset contains additional received signal strength (RSS) data collected in a greenhouse, in a double-tuber scenario, and in dynamic environments. To collect extensive RSS data, we perform the following measurement campaigns. First, sixteen TI CC2531 sensor nodes are deployed in a greenhouse for data collection, and 26 tubers placed at 32 positions are used in thie experiment. Second, we perform experiments in a double-tuber scenario, where six tubers are used. During data collection, two tubers are randomly selected and burried at two predefined positions within the sensing area. Third, we perform data collection in dynamic environments. In the initial envrionmental condition, RSS data are collected from 26 tubers placed at a fixed position. Subsequently, the environmental layout is altered three times, and RSS data are recorded after each layout change. Furthermore, to analyze variations in RSS data under different soil conditions, we collect data from three conditions of soils, including loose soil, compacted soil, and soil with altered mineral composition. To study the effects of soil moisture, we perform additional long-term experiments for data collection. In the first experiment, RSS data are collected with soil moisture gradually decreased, while in the second experiment, RSS data are collected with soil moisture increased through irrigation. In addition, to demonstrate performance in outdoor environments, we construct an outdoor underground tuber planting scenario with two ridges. We bury 18 potato tubers evenly into two ridges and collect one minute of RSS data for each tuber. More details on potato tubers and the file structure are provided in "ReadMe.txt". [1] Wang, Tao, et al. "See-through Soil: Underground Root Tuber Sensing with RF Sensor Networks." IEEE Transactions on Geoscience and Remote Sensing (2025).