Real-Time RISC-V-Based CAN-FD Bus Diagnosis Tool

Area of Science:

• Computer Engineering
• Embedded Systems
• Network Hardware

Background:

• Industrial-grade network diagnosis tools are inaccessible to researchers and students.
• Developing distributed embedded systems requires reliable and low-cost diagnostic solutions.
• Advancements in RISC-V extensions and Domain-Specific Architectures enable novel hardware development.

Purpose of the Study:

• To propose an open-source, real-time, RISC-V based Controller Area Network Flexible Data-rate (CAN-FD) bus diagnosis tool.
• To provide researchers and students with an affordable and dependable alternative for network diagnostics.
• To leverage RISC-V extensions for a dedicated CAN-FD communication unit within a System-on-Chip (SoC).

Main Methods:

• Developed a Real-Time RISC-V-based CAN-FD Bus Diagnosis Tool (RiscDiag CanFd).
• Integrated a CAN-FD Communication Unit controlled by a dedicated Instruction Set Architecture (ISA) Extension on an extended RISC-V core.
• Designed a System-on-Chip (SoC) with UDP communication via Ethernet and a Graphical User Interface (GUI) for FPGA deployment.

Main Results:

• Achieved a 0% frame loss rate during diagnosis.
• Demonstrated a timestamp precision error below 0.001%.
• Measured an acquisition cycle jitter under 10 nanoseconds for Ethernet data preparation.

Conclusions:

• RiscDiag CanFd offers a high-performance, open-source solution for real-time CAN-FD bus diagnosis.
• The tool meets the needs of researchers and students for developing distributed embedded systems.
• The proposed RISC-V based SoC provides a reliable and cost-effective platform for network analysis.