Paper #5

Secure Autonomous Vehicle Localization Framework using GMCC and FSCH-KMC under GPS-Denied Locations

Mohammed Shafi Kundiladi, Sheik Masthan Shahul Abdul Rahim and Mohammed Shahal Rishad

Abstract: For safe and efficient navigation, self-driving cars rely on determining their position through a system called Autonomous Vehicle Localization (AVL). Traditional self-driving cars face challenges related to security and speed in finding their location. To address these problems, this article presents a more secure way, called the Secured Localization (SL), to locate the vehicle, even when signals are weak. First, vehicles are registered and logged in. If Global Positioning System (GPS) signals are available, they are processed securely utilizing Gini-Montgomery Curve Cryptography (GMCC); If GPS signals are not available, then the car uses nearby signal points to find its location. SL data and sensed data from the sensors, including Light Detection and Ranging (LIDAR), Radio Detection and Ranging (RADAR), and Camera are given to On Board Unit (OBU). Then, the vehicle’s position is matched with a pre-stored map for accurate navigation. Finally, various methods such as Fisher Score Chi-Hell Square based K Means Clustering (FSCH-KMC), Cosine Gramian-Kalman Filter (CG-KF), and Hadoop Distributed File System (HDFS) are applied to prioritize and refine the vehicle’s location for better navigation. The proposed framework is simulated and the results show an accuracy of 98%, precision of 98%, and recall of 99%, with improvements in security and faster location finding compared to previous systems.

Keywords: Cosine Gramian-Kalman Filter (CG-KF); Fisher Score Chi-Hell Square based K Means Clustering (FSCH-KMC); Gini-Montgomery Curve Cryptography (GMCC); Global Positioning System (GPS); Secured Localization (SL).