Case Study for Signal Preprocessing for Enhanced Bluetooth Device Identification Using RF Fingerprints and SVM

Version: 20250616-------------------------------------------------------------------------Authors: Rene F. Santana-Cruz (1) Martin Moreno (2) Daniel Aguilar-Torres (1,3) Román A. Valverde-Domínguez (4) Rubén Vázquez-Medina (1)-------------------------------------------------------------------------Affiliations:(1) Instituto Politécnico Nacional, Centro de Investigación en Ciencia Aplicadad y Tecnología Avanzada, Unidad Querétaro 76090, Querétaro Mexico(2) Universidad Tecnológica de San Juan del Río, San Juan del Río, 76800 Querétaro, Mexico(3) Secretaría de Ciencia, Humanidades, Tecnología e Innovación, 03940 Mexico City, Mexico(4) Instituto Politécnico Nacional, Unidad Profesional Interdisciplinaria de Energía y Movilidad, 07738 Mexico City, Mexico-------------------------------------------------------------------------Contact: [e-mail] | [e-mail] -------------------------------------------------------------------------Related paper:<br>Santana-Cruz, R.F., Moreno, M., Aguilar-Torres, D., Valverde-Domínguez, R.A., &amp; Vázquez-Medina, R. (2025). Signal Preprocessing for Enhanced Bluetooth Device Identification Using RF Fingerprints and SVM. Future Internet, 17(6), 250. https://doi.org/10.3390/fi17060250 <br>-------------------------------------------------------------------------The RF signal dataset used for this study consists of 24 Bluetooth devices:- 16 smartphones (8 twin pairs) from the dataset by:<br>Uzundurukan, E., Dalveren, Y., &amp; Kara, A. (2020). A database for the radio frequency fingerprinting of Bluetooth devices. Data, 5(2), 55. https://doi.org/10.3390/data5020055<br>- 8 wearable devices from the dataset by:<br>Rusins, A., Tiscenko, D., Dobelis, E., Blumbergs, E., Nesenbergs, K., &amp; Paikens, P. (2024). Wearable Device Bluetooth/BLE Physical Layer Dataset. Data, 9(4), 53. https://doi.org/10.3390/data9040053<br>Each device contributed 150 RF signals captured under consistent experimental conditions.Sampling rates include 250 Msps (wearables), 5 Gsps, 10 Gsps, and 20 Gsps (smartphones).<br>------------------------------------------------------------------------DEVICE FOLDER NAMING CONVENTION:Each device folder follows the format:<br>[BR][MD].[XX].[FFF].GHz<br>where- BR = Two-letter brand code (e.g., IP for iPhone, SM for Samsung)- MD = Two-character model code (e.g., 5s, G4)- XX = Device sequence number (01 or 02 if it is a twin)- FFF = Frequency in MHz (e.g., 005 for 5 GHz)<br>Example:- IP5s.01.005.GHz -&gt; iPhone 5s, unit 1, recorded at 5 GHz- SMG4.02.005.GHz -&gt; Samsung G4, unit 2 (twin), recorded at 5 GHz<br>

版本：20250616-------------------------------------------------------------------------作者：Rene F. Santana-Cruz (1) Martin Moreno (2) Daniel Aguilar-Torres (1,3) Román A. Valverde-Domínguez (4) Rubén Vázquez-Medina (1)-------------------------------------------------------------------------机构：(1) 墨西哥国立理工学院（Instituto Politécnico Nacional）应用科学与先进技术研究中心克雷塔罗分部，墨西哥克雷塔罗州76090；(2) 圣胡安德里奥技术大学，圣胡安德里奥，墨西哥克雷塔罗州76800；(3) 科学、人文、技术与创新秘书处，墨西哥城03940；(4) 墨西哥国立理工学院能源与移动跨学科专业分部，墨西哥城07738-------------------------------------------------------------------------联系方式：[e-mail] | [e-mail] -------------------------------------------------------------------------相关论文：<br>Santana-Cruz, R.F., Moreno, M., Aguilar-Torres, D., Valverde-Domínguez, R.A., & Vázquez-Medina, R. (2025). 基于射频指纹（RF Fingerprints）与支持向量机（SVM）的蓝牙设备识别信号预处理. Future Internet, 17(6), 250. https://doi.org/10.3390/fi17060250 <br>-------------------------------------------------------------------------本研究使用的射频信号数据集包含24个蓝牙设备：- 16部智能手机（8对孪生设备），来源于以下数据集：<br>Uzundurukan, E., Dalveren, Y., & Kara, A. (2020). 蓝牙设备射频指纹（radio frequency fingerprinting）数据库. Data, 5(2), 55. https://doi.org/10.3390/data5020055<br>- 8个可穿戴设备，来源于以下数据集：<br>Rusins, A., Tiscenko, D., Dobelis, E., Blumbergs, E., Nesenbergs, K., & Paikens, P. (2024). 可穿戴设备蓝牙/BLE物理层数据集. Data, 9(4), 53. https://doi.org/10.3390/data9040053<br>每个设备在一致的实验条件下采集了150个射频信号。采样率包括250 Msps（可穿戴设备）、5 Gsps、10 Gsps和20 Gsps（智能手机）。<br>------------------------------------------------------------------------设备文件夹命名规则：每个设备文件夹遵循以下格式：<br>[BR][MD].[XX].[FFF].GHz<br>其中：- BR = 两位品牌代码（例如，IP代表iPhone，SM代表Samsung）- MD = 两位型号代码（例如，5s、G4）- XX = 设备序号（若为孪生设备则为01或02）- FFF = 频率（单位：MHz，例如005代表5 GHz）<br>示例：- IP5s.01.005.GHz → iPhone 5s，1号设备，5 GHz下记录- SMG4.02.005.GHz → Samsung G4，2号设备（孪生），5 GHz下记录<br>