A novel poisoned water detection method using smartphone-embedded Wi-Fi technology and machine learning algorithms

Water is crucial for our health, and ensuring its quality and cleanliness is a vital ongoing research area. One innovative method we explored involves using Wi-Fi signals to check water quality. By analyzing how these signals change when passing through the water, we can identify whether the water is clean or poisoned. In this study, we trained various machine learning (ML) algorithms to make this distinction effectively. The Wi-Fi access points (WAPs) signal is acquired via equivalent smartphone-embedded Wi-Fi chipsets, and then channel state information (CSI) measures are extracted and converted into feature vectors to be used as input for ML classification algorithms. The measured amplitude and phase of the CSI data are selected as input features into four classifiers k-nearest neighbor (k-NN), support vector machine (SVM), long short-term memory (LSTM), and Ensemble. The experimental results show that the model is capable to differentiate toxic water from clean water with a classification accuracy of 91.5% using LSTM and 92% when GentleBoost-Ensemble is applied to distinguish poisoned water from clean water.