Musa, J., Iannino, A. & Okumoto K. (1987) Software Reliability, Measurement, Prediction, Application. McGraw Hill Book Company.  Home/Search   Document Not in Database   Summary   Related Articles   Check

....faults can be deU99[6 and re[ vez TheR we canofte find thatthe fault de6 8 RR time intez al be M longe andthe correA onding probability isde9 U[6 R asthe teeMM go e on. From thisres[zM the so calle software relgrowth model have be de e e to de[ AR e the failure[ ccurre[ phere[ inte 9U phase [1] [3] The software re9RR[6zA y growth mode have se e advantage to support ade U8M[ making inthe de e8zA[ t manage[6 t of software products. For instance it isuse forthe software pro manage to de00MM[6 the optimaltime to stop software teez8 and todezz e the syste tousezR This probleo calle ....

J.D. Musa, A. Iannino, and K. Okumoto, Software Reliability, Measurement, Prediction, Apdiction,y McGraw-Hill, N.Y., 1987.

....or enumerable that has remained untested so far. This process makes actual testing more directed and hence more efficient than random testing. Malaiya, von Mayrhauser and Srimani [mvs92] show that this non random process leads to a defect finding behavior described by the logarithmic growth model [mus87]. Their analysis gives an interpretation for the model parameters. The coverage growth of an enumerabletype depends on the detectability profile of the type and the test selection strategy. If the defect coverage growth in practice is described by the logarithmic model, it is likely that the ....

J.D. Musa, A Iannino, K. Okumoto, Software Reliability, Measurement, Prediction, Application, McGraw-Hill, 1987.

....as the software in operational usage. Indeed, this is an assumption generally made for software reliability models [9] If the input selection during testing phase is different in distribution from that in operation, some adjustment should be made to account for the differences. Musa et al. [17] introduce a concept termed test compression factor for this purpose. In contrast with real operational use, input states for software during testing phase are generally not repeated or repeated with much lower frequency. Thus, actual test inputs are more effective in revealing errors than random ....

....with real operational use, input states for software during testing phase are generally not repeated or repeated with much lower frequency. Thus, actual test inputs are more effective in revealing errors than random sampling according to operational usage patterns. An simple example was given in [17] to illustrate the concept of test compression factor, which is quoted below. Assume that a program has only two input states, A and B. Input state A occurs 90 percent of the time; B, 10 percent. All runs take 1 CPU hr. In operation, on the average, it will require 10 CPU hr to cover the input ....

[Article contains additional citation context not shown here]

J.D. Musa, A Iannino, K. Okumoto, Software Reliability, Measurement, Prediction, Application, McGraw-Hill, 1987.

....based on an operational profile. An operational profile is a set of relative frequencies of occurrence of the operations associated with the software during its use in the field [Mus93] Interpretation of many software reliability models assumes failure detection based on operational profiles [Mus87]. Since this assumption is usually violated during early software testing phases (for example, during unit testing and integration testing) assessment and control of software quality growth during non operational testing stages is difficult and open to interpretation. Research supported in part ....

....will depend on how often and in what manner the system, or a component, is used. This suggests that the data also needs to be collected on the software usage, size, and any other relevant metric. Distinction needs to be made between unique failures, repeated failures, and the underlying faults [Mus87]. 2.1.2 Failure intensity Failure intensity is a classical SRE metric [Mus87] It can be defined as the rate of change of the mean value function, or the number of failures per unit time. The mean value function is the average cumulative number of failures at a point in time. In the context of the ....

[Article contains additional citation context not shown here]

J. Musa, A. Iannino and K. Okumoto, Software Reliability (Measurement, Prediction, Application), McGraw-Hill 1987

....assumptions about the software failure process so that the model becomes mathematically tractable. Goel [12] has given the typical assumptions made by the models along with the limitations. Many reliability models have been proposed [11, 16, 20, 22, 27] Surveys of different models are given in [8, 12, 23]. To employ a model for reliability prediction, value of some of the parameters need to be specified. These are typically determined by analyzing the past failure data of the software. Some of the well known models are briefly described below. The model proposed by Jelinski and Moranda [16] is one ....

....failures and is proportional to the current error content (number of faults remaining) of the program being tested. This model is very simple to use. It is also fairly accurate for some data sets, but sometimes leads to inaccurate predictions. The Basic Execution Time model proposed by Musa [23] makes assumptions similar to the above model except that the process modeled is the number of failures in specified execution time intervals. There are a finite number of faults in the beginning of the test phase, and the times between failures are exponential, the failure rate being uniform ....

[Article contains additional citation context not shown here]

J.D. Musa, A. Iannino and K. Okumoto, Software Reliability, Measurement, Prediction, Application, McGraw-Hill, 1987.

No context found.

Musa, J., Iannino, A. & Okumoto K. (1987) Software Reliability, Measurement, Prediction, Application. McGraw Hill Book Company.

No context found.

. J. Musa, A. Iannino, and K. Okumoto, Software Reliability, McGraw Hill, 1987.

No context found.

J. D. Musa, A. Iannino, K. Okumoto, Software Reliability, Measurement, Prediction, Application, McGraw-Hill, New York, 1987.

No context found.

J.D. Musa, A. Iannino and K. Okumoto,Software Reliability (Measurement, Prediction, Application), McGraw-Hill 1987

Online articles have much greater impact   More about CiteSeer.IST   Add search form to your site   Submit documents   Feedback  

CiteSeer.IST - Copyright Penn State and NEC