With better computing facilities now available, there is an ever-increasing need to ensure that elegant theoretical results on hardware reliability are computationally available. This book discusses those aspects which have relevance to computing systems and those where numerical computation was a problem. It is also well known that nearly 70% of the cost goes into software development and hence software reliability assumes special importance. The book not only gives an extensive review of the literature on software reliability but also provides direction in developing models which are flexible and can be used in a variety of testing environments. Besides, several alternative formulations of the release time problem are discussed along with variants such as allocation of testing effort resources to different modules of the software, or the testing effort control problem. Software reliability has now emerged as an independent discipline and requires a strong partnership between computer scientists, statisticians and operational researchers. This aspect is broadly highlighted in the book.
Preliminary concepts and background; replacement policies with minimal repairs; problems with applications to computing systems; software reliability growth models based on NHPP; release policies; numerical computations in renewal and reliability theory.