Cost constrained reliability maximization of software systems

Architecture-based techniques have been largely used for the reliability assessment of software systems. However, these techniques also enable the exploration of cost/reliability tradeoffs and evaluation of a set of competing archiectural alternatives. This paper presents an optimization framework based on an evolutionary algorithm (EA) which can be used to explore cost$/$reliability tradeoffs based on software architecture. Evolutionary algorithm was used as an optimization technique because of the discontinuous search space, usually nonlinear but monotonic relation between the cost and reliability of individual modules comprising the software, and complex software architectures giving rise to nonlinear dependencies between individual module reliabilities and the overall application reliability. We illustrate the use of the EA using a case study, where the results of the EA are compared with those obtained from exhaustive enumeration. A comparison of the time taken by the EA to generate an optimal solution with the time taken by exhaustive search to generate an optimal solution indicates that the EA can be used to obtain optimal designs with much greater efficiency than exhaustive search.