Structural Models for Failure Detection of Moore Finite-State Machines

Abstract

Fault detection is an important task in the design of fault-tolerant state machines. Structural models of Moore finite-state machines are proposed to detect multiple faults in various elements of the finite-state machine and prevent their negative impact on the controlled object. The structural models under consideration make it possible to detect invalid input and output vectors both in each state and for the entire automaton, invalid code of the current and next state of the automaton, and invalid transitions between states. The costs of implementing the proposed structures by area range on average from 3 to 26%, and the speed of the machine either does not change or even increases by an average of 24–30%. Estimates of the area and performance of the proposed structural models of finite-state machines are given, and recommendations for their practical use are given. It is shown that the choice of a suitable structure allows us not to increase the area, and in some cases even leads to an increase in the performance of the finite-state machine.