Scholarlink

<p>Smart car Autonomous Vehicular Systems (AVS) face significant threats from GPS spoofing, which can dangerously manipulate a vehicle’s perceived location. This paper presents a robust and enhanced mitigation framework that combines machine learning-based detection, cryptographic safeguards, and real-time alert systems. The proposed methodology begins with system initialization and GPS signal reception, followed by a data consistency check that flags suspicious signals. Statistical analysis and supervised machine learning, developed in prior work, identify spoofed signals based on key features such as PDOP, HDOP, and VDOP, with PDOP emerging as a primary indicator. The Extra Trees Regressor model demonstrates superior performance in handling these complex interactions, with visualizations such as heatmaps and scatter plots offering deeper insights. To further enhance reliability, an MPU6050 accelerometer-gyroscope module is integrated to continuously monitor real-time motion and orientation. This enables inertial measurements to cross-verify GPS data, especially in spoofing scenarios or environments with signal obstructions. When spoofing is confirmed, the system triggers LED indicators and sends alerts via GSM, while normal data is hashed using SHA-256, encrypted with RSA, and securely transmitted. This comprehensive approach ensures high integrity, real-time spoofing detection, and system resilience, significantly enhancing safety and operational reliability in AVS and other GPS-dependent applications.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

Mitigating GPS spoofing in AVS: SHA and RSA algorithms with proteus simulation for real-time detection

  • 期刊论文

  • 发表日期：

    2025年6月30日

查看全文
  • B Poornima,
  • Lalitha Surya Kumari

摘要

Smart car Autonomous Vehicular Systems (AVS) face significant threats from GPS spoofing, which can dangerously manipulate a vehicle’s perceived location. This paper presents a robust and enhanced mitigation framework that combines machine learning-based detection, cryptographic safeguards, and real-time alert systems. The proposed methodology begins with system initialization and GPS signal reception, followed by a data consistency check that flags suspicious signals. Statistical analysis and supervised machine learning, developed in prior work, identify spoofed signals based on key features such as PDOP, HDOP, and VDOP, with PDOP emerging as a primary indicator. The Extra Trees Regressor model demonstrates superior performance in handling these complex interactions, with visualizations such as heatmaps and scatter plots offering deeper insights. To further enhance reliability, an MPU6050 accelerometer-gyroscope module is integrated to continuously monitor real-time motion and orientation. This enables inertial measurements to cross-verify GPS data, especially in spoofing scenarios or environments with signal obstructions. When spoofing is confirmed, the system triggers LED indicators and sends alerts via GSM, while normal data is hashed using SHA-256, encrypted with RSA, and securely transmitted. This comprehensive approach ensures high integrity, real-time spoofing detection, and system resilience, significantly enhancing safety and operational reliability in AVS and other GPS-dependent applications.