Error Recovery in the Time-Triggered Paradigm with FTT-CAN

Area of Science:

• Computer Science
• Network Engineering
• Embedded Systems

Background:

• Data networks face interference, risking critical system failure.
• Time-triggered networks offer detection but slow recovery, requiring overprovisioning.
• Flexible Time-Triggered (FTT) paradigm balances detection and recovery with less bandwidth.

Purpose of the Study:

• Introduce a novel error recovery method for time-triggered Controller Area Network (CAN) using FTT.
• Design a traffic shaper for regulating message retransmissions.
• Achieve reliability goals while minimizing system interference and preserving schedulability.

Main Methods:

• Developed a traffic shaper server for FTT-CAN error recovery.
• Modeled worst-case recovery scenarios using a Poisson process for fault arrival rates.
• Evaluated the method through extensive simulations on practical and random systems.

Main Results:

• The FTT-CAN method achieves predefined reliability goals.
• Error recovery mechanism interference is limited, preserving system schedulability.
• Demonstrated a two-orders-of-magnitude reduction in average recovery bandwidth compared to traditional methods.

Conclusions:

• FTT-CAN offers an efficient solution for timely error recovery in time-triggered CAN networks.
• The proposed traffic shaping approach significantly reduces bandwidth overhead.
• This method enhances the resilience of critical distributed systems.