Characterization of Ethernet Voltage Profiles for Network Health Monitoring
Automotive Ethernet is increasingly being adopted in modern vehicles to support high-bandwidth, real-time communication among sensors, controllers, and intelligent systems. While physical-layer diagnostic techniques are well established for traditional automotive networks such as Controller Area Networks (CAN), comparable voltage-based diagnostic approaches for Automotive Ethernet remain relatively new and challenging due to higher data rates, differential signaling, and increased physical-layer complexity. This research aims to characterize the physical-layer voltage behavior of Automotive Ethernet as a foundation for future real-time network health monitoring and diagnostics. The study focuses on developing baseline voltage profiles of healthy Ethernet communication links under varying normal traffic and operating conditions. An experimental platform integrating a Speedgoat Performance Real-Time Target Machine with high-resolution voltage measurement instrumentation will be used to acquire and analyze Ethernet waveforms, with particular emphasis on 1000BASE-T1 Automotive Ethernet. Voltage characteristics and signal integrity metrics will be extracted to identify repeatable signatures and quantify the normal variability associated with different traffic and operating conditions. The resulting voltage profiles will establish reference operating envelopes for healthy Ethernet communication, providing a basis for distinguishing normal signal variations from those associated with future degradation and physical-layer faults. By enabling electrical changes in the communication link to be identified before they develop into communication failures, this work can support early fault detection, predictive maintenance, and improved network reliability and resilience. The developed framework will also provide a foundation for future degradation and fault characterization, automated diagnostics, and machine-learning-based Ethernet network health prediction, ultimately helping reduce unexpected network failures, system downtime, and maintenance costs in connected and automated vehicles.
abanimilhim@mail.fresnostate.edu
California State University, Fresno
This research will produce a validated, high-fidelity baseline characterization of Automotive Ethernet physical-layer voltage behavior under normal operating and traffic conditions. Key outputs will include baseline voltage profiles and reference operating envelopes for healthy Ethernet communication, a curated experimental dataset of physical-layer voltage measurements, and a real-time data acquisition and analysis framework for voltage characterization. The project will also establish an experimental methodology for collecting, processing, and evaluating Automotive Ethernet voltage signals with an emphasis on 1000BASE-T1 communication. These research outputs will provide the technical foundation needed to distinguish normal voltage variations from future degradation-related changes and support subsequent development of voltage-based network health monitoring, diagnostics, and predictive maintenance methods for transportation and industrial communication networks.
The anticipated impact of this research is improved safety, reliability, and maintainability of transportation systems that increasingly depend on Automotive Ethernet for communication among sensors, controllers, and automated vehicle functions. By establishing high-fidelity baseline voltage profiles and normal operating envelopes, this research will provide the foundation for future diagnostic technologies capable of identifying physical-layer degradation before it progresses to intermittent communication errors or complete network failure. Early detection of deteriorating cables, connectors, and other communication-link issues could enable condition-based and predictive maintenance rather than relying primarily on reactive troubleshooting after a failure occurs.
Implementation of voltage-based Ethernet health monitoring could reduce unexpected communication failures, vehicle downtime, diagnostic time, and maintenance costs while improving the reliability and durability of increasingly complex vehicle electronic systems. This capability is particularly important as connected and automated vehicles rely on high-speed Ethernet networks to exchange safety-critical sensor and control information. The developed characterization and real-time analysis framework could ultimately support onboard network health monitoring, maintenance decision-making, and the development of more resilient vehicle communication architectures. The research may also inform future testing and diagnostic practices for Automotive Ethernet and provide a technical basis for subsequent development of commercial diagnostic tools and intellectual property related to physical-layer network health monitoring.
-
Contact Us
San José State University One Washington Square, San Jose, CA 95192 Phone: 408-924-7560 Email: mineta-institute@sjsu.edu
