The concept of a fault and damage tolerant communications network has been proposed previously for applications requiring high reliability. This concept is now being considered for an advanced transport aircraft to connect sensors, actuators and subsytems to a central computer instead of redundant multiplexed data buses. The reliability of such a network has been evaluated for a baseline configuration in which all sensors, actuators and communications subsystems are quadruply redundant. The processing overhead associated with managing the network redundancy, uncovering and identifying faults in the nodes and links of the network and repairing the network in the presence of faults and damage are also estimated. These estimates are derived from execution times of actual algorithms running on a quad redundant computer that interfaces with the simulated network.
|Number of pages
|Proceedings of the American Control Conference
|Published - Dec 1 1984
All Science Journal Classification (ASJC) codes
- Electrical and Electronic Engineering