Abstract

The paper deals with the cost-benefit analysis of a system composed of two non-identical units A and B arranged in series configuration with an identical cold standby unit corresponding to unit-A. System failure occurs when either unit-A (including its redundancy) or unit-B are in total failure mode. Failure and repair times of the units are independent random variables of discrete nature and follow geometric distributions with different parameters. The various system effectiveness measures are obtained using the regenerative point technique1. The reliability of software can be assessed by analyzing the pattern of failure occurrence at the software testing stage. To assess the reliability of Software, the Rayleigh Class software reliability growth model is proposed by assuming the Binomial pattern of failure occurrence. Also, the Bayes estimates for the number of inherent failures and the scale parameter are proposed using the non-informative, inverted gamma prior, respectively. The performance of proposed Bayes estimators are compared with corresponding MLEs based on relative efficiencies obtained by the Monte Carlo simulation technique for studying the performance of both proposed estimators.

Author: Rajesh Singh, Kailash R. Kale and Pritee Singh

Received on: May, 2023

Accepted on: January, 2024