{"title":"Sum of Two Exponentials Based Path Loss Model for Inter-Device Range Estimation using Stochastic Gradient Descent Method","authors":"Deepali Kushwaha, Ankur Pandey, Sudhir Kumar","doi":"10.1109/ANTS.2018.8710135","DOIUrl":null,"url":null,"abstract":"In this paper, we propose a sum of two exponentials based path loss model for inter-device range estimation using Stochastic Gradient Descent (SGD) method. We observe Bluetooth Received Signal Strength Indication (RSSI) data for short-range distance estimation. Bluetooth location accuracy is very high for short-range localization systems and hence it is widely used in gadgets. This paper proposes a new model for the relationship between distance and three parameters namely RSSI, signal-to-noise ratio (SNR) and the data rate of the Bluetooth signal. We consider four different environments for evaluating various path loss models. The best path loss model for all the parameters is then further used for estimating the distance. We also show that the SGD method outperforms the Gradient Descent (GD) method in terms of location accuracy and is computationally efficient.","PeriodicalId":273443,"journal":{"name":"2018 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS)","volume":"1 1","pages":"0"},"PeriodicalIF":0.0000,"publicationDate":"2018-12-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"2018 IEEE International Conference on Advanced Networks and Telecommunications Systems (ANTS)","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/10.1109/ANTS.2018.8710135","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
In this paper, we propose a sum of two exponentials based path loss model for inter-device range estimation using Stochastic Gradient Descent (SGD) method. We observe Bluetooth Received Signal Strength Indication (RSSI) data for short-range distance estimation. Bluetooth location accuracy is very high for short-range localization systems and hence it is widely used in gadgets. This paper proposes a new model for the relationship between distance and three parameters namely RSSI, signal-to-noise ratio (SNR) and the data rate of the Bluetooth signal. We consider four different environments for evaluating various path loss models. The best path loss model for all the parameters is then further used for estimating the distance. We also show that the SGD method outperforms the Gradient Descent (GD) method in terms of location accuracy and is computationally efficient.