Implementation of Voter by using Artificial Bee Colony Algorithm for Fault Tolerant System

Abstract

In a fault tolerant system, voting algorithm plays a major role in decision making. Inexact majority voting, majority voting, and weighted voters are most popular voting algorithms. In each of these techniques, redundant modules produce unpredictable outputs, and finding an agreement among these outputs is the functionality of voter. In this proposed system, voter is implemented using soft computing technique called ABC algorithm. The basic ABC algorithm is a very modest, adaptable and clear model. There are two categories of searchers in the hive of ABC; namely, unemployed searchers and employed searchers. Based on the intelligent performance of honey bee swarm, ABC is one of the optimization algorithms. The behaviours of bees are used to optimize weights of Triple Modular Redundant system, where each module plays an imperative role in the selection of voter. For evaluating the performance of voting algorithms, some performance measures have been defined; they are, safety and availability measures. ABC algorithm has been proposed to improve the safety and availability of the existing system.