Informatics Publishing Limited

K. V. Murali Mohan ¹, R. Satya Prasad ², G. Sridevi ³

Affiliations
1 ECE, H. M. Institute of Technology and Science, Hyderabad - 501301, Andhra Pradesh, IN
2 CSE, Acharya Nagarjuna University, Guntur - 522 510, Andhra Pradesh, IN
3 CSE, KL University, Guntur – 522 502, Andhra Pradesh, IN

Abstract

Background/Objectives: There is a need to have tools/mechanisms to detect whether a software is reliable or not. Much time is spent while using Classical Hypothesis testing because the conclusions can be drawn only after collecting large amounts of data. Methods/Statistical Analysis: By adopting Sequential Analysis of Statistical science, such decision can be made quickly. In this paper, we proposed a new type of statistical science procedure, Sequential Probability Ratio Test (SPRT) applied for Burr Type XII model based on Time domain data. Findings: For the proposed Burr type XII model, we applied the SPRT methodology on five real time software failure datasets that were borrowed from different software services. The result exemplifies that the model has given a decision of rejection for all the datasets. Therefore our findings state that all used datasets are unreliable. Application/Improvements: Applying SPRT procedure on datasets, we can come to an early conclusion of reliable/unreliable software.

Keywords

Burr Type XII Model, ML Estimation, NHPP, Sequential Probability Ratio Test, Time Domain Data

Full Text

   

D. N. V. Syma Kumar ¹, R. Satya Prasad ², Venkateshwarla Rama Raju ³

Affiliations
1 Dept. of Computer Sci. and Engineering, SSN Engineering College, Ongole, IN
2 Dept. of Computer Sci. and Engineering, Acharya Nagarjuna University, Guntur, IN
3 Sree Veera Venkata Satyanarayana Engineering College, Ongole, AP, IN

Abstract

With the discovery of Internet, the new ways of creating and/or developing software engineering have arisen that make use of computational telematic tools, utilities and techniques, follow flexible methodologies and incorporate third (or other) party contributions into software. One of the pragmatic and paradigmatic examples of software engineering development that counts on these characteristics can be found in the phenomenon of (free) open source software (i.e., un-licensed, licensed free called "libre") being of special interest, those projects that are large in number of participants and in size of software (source-code).

Albeit, at first, these novel environments are less preventable than those conventional/ traditional ones, because, creation and hence development is done usually in a geographically distributed way, but there is no solid company behind the development that takes lead, traditional hierarchical structures are not followed or external contributions are hardly predictable, we have access to much information. The software product itself, and many by products that are created/ designed, during the development process (e.g., communication archives, bug track systems and versioning systems, among others). These data usually publicly available on the Internet, therefore, we can make the extensive and hence exhaustive analysis with massive data, and much of it is cumbersome (too hardly obtainable in industrial and other environments). In this paper, we elucidate aims and objectives which are goals and discuss the rudimentary theory, data sources, techniques and methodologies that are proposed in our empirical software engineering research on open source called LIBRE software.

Keywords

LIBRE Software, LISPs Software Projects, Libre Software Development (LSWD), Free Open Source Software (FLOSS) Projects, GNU Emacs, GNU Compiler Collection (GCC).

Full Text

     

R. Satya Prasad ¹, Bandla Sreenivasa Rao ², R. R. L. Kantham ³

Affiliations
1 Department of Computer Science and Engineering, Acharya Nagarjuna University, Guntur, IN
2 Department of Computer Science, VRS & YRN College, Chirala, IN
3 Department of Statistics, Acharya Nagarjuna University, Guntur, IN

Abstract

This paper consider an MMLE (Modified Maximum Likelihood Estimation) based scheme to estimate software reliability using exponential distribution. The MMLE is one of the generalized frameworks of software reliability models of Non Homogeneous Poisson Processes (NHPPs). The MMLE gives analytical estimators rather than an iterative approximation to estimate the parameters. In this paper we proposed SPC (Statistical Process Control) Charts mechanism to determine the software quality using inter failure times data. The Control charts can be used to measure whether the software process is statistically under control or not.

Keywords

Software Reliability, Statistical Process Control, Control Charts, NHPP, Exponential Distribution, MMLE, Inter Failure Times Data.

Full Text