Contribution of Controller Area Networks Controllers to Masquerade Failures
Authors:
- 22 March 2006
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 310-316
Abstract
This paper scrutinizes faults in a CAN controller that may result in masquerade failures, and suggests an even parity mechanism to detect them with minimum hardware overhead. To do this, a CAN controller is modeled by VHDL at behavioral level and is exploited to setup a CAN-based network composed of two nodes. A total of 5,500 faults are injected into essential parts of one of the controllers. The results show that about 3.44% of faults terminate in masquerade failures. The results, also, show that register bank in the CAN controller are the most sensitive portions in which 92.10% of faults occurring in the register bank result in masquerade failures. The even parity mechanism detects about 96.31% of all masquerade failures.Keywords
This publication has 10 references indexed in Scilit:
- Software detection mechanisms providing full coverage against single bit-flip faultsIEEE Transactions on Nuclear Science, 2004
- An efficient TDMA start-up and restart synchronization approach for distributed embedded systemsIEEE Transactions on Parallel and Distributed Systems, 2004
- Dependability analysis using a fault injection tool based on synthesizability of HDL models2004
- Reliable real-time communication in can networksIEEE Transactions on Computers, 2003
- CAN emulation in a time-triggered environment2002
- A Comparison of CAN and TTPAnnual Reviews in Control, 2000
- Time redundancy based soft-error tolerance to rescue nanometer technologies1999
- Scheduling messages on controller area network for real-time CIM applicationsIEEE Transactions on Robotics and Automation, 1997
- Impact of technology trends on SEU in CMOS SRAMsIEEE Transactions on Nuclear Science, 1996
- FERRARI: a tool for the validation of system dependability properties1992