Change search
ReferencesLink to record
Permanent link

Direct link
Instant Feedback Loops – for short feedback loops and early quality assurance
Blekinge Institute of Technology, Faculty of Computing, Department of Communication Systems.
2016 (English)Independent thesis Advanced level (degree of Master (Two Years)), 20 credits / 30 HE creditsStudent thesis
Abstract [en]

Context. In recent years, Software Quality Assurance (SQA) has become a crucial part of software development processes. Therefore, modern software development processes led to an increase in demand for manual and automated code quality assurance. Manual code quality reviews can be a time-consuming and expensive process with varying results. Thus, automated code reviews turn out to be a preferred alternative for mitigating this process. However, commercial and open-source static code analyzer tools often offer deep analysis with long lead time.

Objectives. In this thesis work, the main aim is to introduce an early code quality assurance tool, which features a combination of software metrics. The tool should be able to examine code quality and complexity of a telecommunication grade software product such as source code of specific Ericsson product by Ericsson. This tool should encapsulate complexity and quality of a software product with regards to its efficiency, scope, flexibility, and execution time.

Methods. For this purpose, the background section of the thesis is dedicated to in-depth research on software metrics included in well-known static code analyzers. Then, development environment, under investigation source code of Ericsson product, and collected software metric for evaluation were presented. Next, according to each software metric’s characteristics, point of interest, and requirement, a set of steps based on a Susman’s action research cycle were defined. Moreover, SWAT, a suitable software analytics toolkit, employed to extract conducted experiment data of each software metric from a static analyzer code named Lizard in order to detect most efficient software metrics. Outcome of conducted experiment demonstrates relationship of selected software metrics with one another.

Results. The chosen software metrics were evaluated based on a variety of vital factors especially actual data from number of defects of specific Ericsson product. Highly effective software metrics from investigations in this thesis work were implemented as a new model named hybrid model to be utilized as an early quality assurance.

Conclusions. The proposed model, which consist of well-performing software metrics, demonstrate an impressive performance as an early code quality indicator. Consequently, the utilized model in this master thesis could be studied in a future research to further investigate the effectiveness and robustness of this tool an early quality assurance. 

Place, publisher, year, edition, pages
2016. , 63 p.
Keyword [en]
Software Quality Assurance, software metrics, static code analyzer, Ericsson, Lizard and SWAT.
National Category
Telecommunications
Identifiers
URN: urn:nbn:se:bth-13379OAI: oai:DiVA.org:bth-13379DiVA: diva2:1045996
External cooperation
Ericsson AB
Subject / course
ET2580 Master's Thesis (120 credits) in Electrical Engineering with emphasis on Telecommunication Systems
Educational program
ETATE Master of Science Programme in Electrical Engineering with emphasis on Telecommunication Systems
Presentation
2016-09-27, J3208 Claude Shannon, Blekinge Tekniska Högskola, Karlskrona, 12:30 (English)
Supervisors
Examiners
Available from: 2016-11-16 Created: 2016-11-11 Last updated: 2016-11-16Bibliographically approved

Open Access in DiVA

fulltext(6035 kB)18 downloads
File information
File name FULLTEXT02.pdfFile size 6035 kBChecksum SHA-512
a3c020f65cfbc80baf882d6ed6b87b3d57d26c981cea302f0081e0162b6cdb685553d868cd786fa4e6761568a637ffda8742692fd1ca2f2db77293e5dcfec274
Type fulltextMimetype application/pdf

By organisation
Department of Communication Systems
Telecommunications

Search outside of DiVA

GoogleGoogle Scholar
Total: 18 downloads
The number of downloads is the sum of all downloads of full texts. It may include eg previous versions that are now no longer available

Total: 162 hits
ReferencesLink to record
Permanent link

Direct link