Abstract (EN):
Although the Controller Area Network (CAN) is widely used as the communication subsystem of many distributed control systems, it presents dependability limitations that are inherent to its bus topology. In particular, it lacks of the tailored mechanisms to avoid that the errors generated by a single fault jeopardizes the communication capabilities of many nodes, possibly causing a general failure of the system. In order to overcome this limitation, we developed an innovative star topology, called CANcentrate, whose hub includes advanced error-containment mechanisms. This paper explains how to model both CAN and CANcentrate by means of stochastic Petri Nets, in order to assess the error-containment improvement achieved by CANcentrate over CAN, and further shows some first results. ©2006 IEEE.
Language:
English
Type (Professor's evaluation):
Scientific