Natalia Vallejo Montoya, Daniel Rodriguez, Changzhi Li
{"title":"Wireless Power Transfer Sensing Approach for Milk Adulteration Detection Using Supervised Learning","authors":"Natalia Vallejo Montoya, Daniel Rodriguez, Changzhi Li","doi":"10.1109/RWS53089.2022.9719981","DOIUrl":null,"url":null,"abstract":"With the increasing demand for wireless sensors due to the growing Internet of Things (IoT) industry, it becomes desirable to use existing technologies to realize new sensing functions. As wireless power transfer (WPT) becomes a standard feature in smartphones, this paper studies the non-invasive classification of liquid solutions with different concentrations, based on the WPT technology already deployed in mobile devices. Average accuracies of up to 97.6% were achieved utilizing supervised machine learning for the classification of milk adulterated with different water volumes. For these experiments, milk concentrations of 100%, 80%, 60%, and 40% were used for classification. Additionally, singular value decomposition and boxplot analysis were used to reduce the radio frequency bandwidth needed for classification to 9.45 MHz, leading to a drastic reduction in hardware complexity and computational cost.","PeriodicalId":113074,"journal":{"name":"2022 IEEE Radio and Wireless Symposium (RWS)","volume":"54 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2022-01-16","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2022 IEEE Radio and Wireless Symposium (RWS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/RWS53089.2022.9719981","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
Abstract
With the increasing demand for wireless sensors due to the growing Internet of Things (IoT) industry, it becomes desirable to use existing technologies to realize new sensing functions. As wireless power transfer (WPT) becomes a standard feature in smartphones, this paper studies the non-invasive classification of liquid solutions with different concentrations, based on the WPT technology already deployed in mobile devices. Average accuracies of up to 97.6% were achieved utilizing supervised machine learning for the classification of milk adulterated with different water volumes. For these experiments, milk concentrations of 100%, 80%, 60%, and 40% were used for classification. Additionally, singular value decomposition and boxplot analysis were used to reduce the radio frequency bandwidth needed for classification to 9.45 MHz, leading to a drastic reduction in hardware complexity and computational cost.