Design of fault-tolerant distributed control systems

Real-time control systems, involving continuous measuring and monitoring of physical quantities, need often to take into account fault-tolerance capabilities to guarantee a correct behavior (even in the presence of a given amount of hardware faults and software errors) or, at least, to provide gracefully degradable performances. In this paper, hardware dimensioning, the optimum allocation of the computation, and the fault-tolerance issues are afforded contemporaneously, with specific attention to the design of dedicated distributed control systems. A single optimization frame is defined to identify a globally optimum solution with respect to these conflicting goals.