Who solved the optimal software release problems based on Markovian software reliability model?

In this paper, we answer the titled question for the typical software release problems based on a Markovian software reliability model. Under the assumption that the software fault-detection process is described by a Markovian birth process with absorption, like Jelinski and Moranda model (1972), we formulate the total expected software costs with two different release policies, and compare them in terms of the cost minimization. Also, we consider the other software cost model taking account of gain in reliability. It can be concluded throughout numerical examples that the existing optimal software release policy underestimates and overestimates the real optimal software release time and its associated cost function, respectively.