An Intelligent Framework for Signal Integrity and Communication Reliability in CAN-Based Autonomous Vehicle Systems

Abstract

The growing dependence of autonomous vehicles on continuous data exchange between electronic control units, sensors, and actuators has increased the need for reliable in-vehicle communication. The Controller Area Network (CAN) remains widely used for automotive communication because of its robustness, simplicity, and real-time message handling capabilities. However, increasing network traffic, electromagnetic interference, signal disturbances, transmission errors, and changing bus conditions can affect signal integrity and reduce communication reliability. Such disturbances may become critical in autonomous vehicle systems, where delayed or corrupted information can influence time-sensitive control decisions. This study proposes an intelligent framework for monitoring signal integrity and evaluating communication reliability in CAN-based autonomous vehicle systems. The framework combines communication-level measurements with an intelligent assessment mechanism to identify abnormal transmission behaviour and estimate the reliability of network conditions. Parameters including signal-to-noise ratio, bit and frame error characteristics, communication latency, bus load, and message loss are considered to represent the operational state of the CAN network. The proposed approach is designed to distinguish stable communication from degraded and potentially unreliable conditions while supporting early identification of communication anomalies. Its performance is evaluated under varying traffic loads and disturbance scenarios and compared with conventional monitoring approaches using reliability and detection-oriented performance measures. The framework is expected to provide a more adaptive method of assessing CAN communication quality than fixed-rule monitoring alone. The study contributes a structured approach for improving communication awareness and supporting dependable data exchange in autonomous vehicle architectures, with potential applicability to advanced driver assistance systems and other safety-critical automotive communication environments.