Please use this identifier to cite or link to this item:
https://ptsldigital.ukm.my/jspui/handle/123456789/513440Full metadata record
| DC Field | Value | Language |
|---|---|---|
| dc.contributor.advisor | Ravie Chandren Muniyandi, Prof. | - |
| dc.contributor.author | Mahdi Sahlabadi (P64620) | - |
| dc.date.accessioned | 2023-10-16T04:36:41Z | - |
| dc.date.available | 2023-10-16T04:36:41Z | - |
| dc.date.issued | 2020-06-02 | - |
| dc.identifier.other | ukmvital:130945 | - |
| dc.identifier.uri | https://ptsldigital.ukm.my/jspui/handle/123456789/513440 | - |
| dc.description | Software architecture evaluation searches for problems and potential risks in an architecture to assist the architect for assessing required quality requirements. Although a plethora of systematic software architecture evaluation methods were proposed by researchers, practitioners refrain from applying these methods since they are heavyweight. Alternatively, the lightweight architectural evaluation methods have been proposed to satisfy the practitioners’ concerns, but they are not also interested in using these methods. Hence, it is essential to develop a lightweight architectural evaluation framework based on the stakeholders’ concerns. In order to reach this goal, three objectives are defined. Firstly, the features and lightweight factors improving the use of architectural evaluation methods in industry are recognized. The features are identified based on practitioners’ concerns by the analysis of architecture’s definitions from stakeholders, and new researches’ trends by reviewing the literature of architectural evaluation methods. Then, the Lightweightness factors are obtained from the analysis of five most common used lightweight methods and ATAM as the most well-known heavyweight method. Subsequently, in order to address these features and factors, process mining is selected as the powerful tool and technique that deals with a large scale complexity. Secondly, the architectural evaluation framework is proposed. There are four steps to achieve this objective which are proposing the skeleton of the framework with respect to the features and factors, adapting process mining on the framework, assessing the discovered architectural models’ quality in order to choose the best model, and applying process mining techniques to analyze the security and performance of architectures. Thirdly, the proposed framework applicability, repeatability and efficacy are evaluated. The framework is verified by three longitudinal case studies. Moreover, it is validated by three industrial cases and experts review. Therefore, it is confirmed that the research achieves the lightweight, ad-hoc and agile friendly framework that analyzes the architecture security and performance of heterogeneous and distributed system. The architectures are presented in service oriented, Pipe and Filter, and component based architecture styles analyzed by process mining and Petri Net in the early and late stages of software development. Indeed, the framework defines the architecture, architectural events and UML to Petri Net transforming methods in order to collect the planned and implemented architecture. Additionally, the framework uses processes mining to discover the planned architecture from synthesized data, and the implemented architectures from the software log files. Accordingly, the framework defines log files and transforms data to log files. Despite the grand challenges, process mining is adapted in the framework for the architectural evaluation. In addition, the framework introduces Timed Hierarchical Colored Petri Nets that simulates the activities of the planned architectures. The framework opts the best architectural discovery algorithm out of Flower, Alpha, ILP, Heuristic, and Inductive based on the twelve quality criteria. Ultimately, the framework defines the criteria for the performance analysis, methods for security analysis and SA deviation detection. Moreover, the abnormal architectural activities are recognized by Linear Temporal Logic (LTL). The framework has been also implemented on six case studies. The results of the evaluation succeeded in getting the current method to be used for these case studies. Moreover, the experts agree with the advantages and efficacy of the framework to achieve its goal. In the verification stage, heuristic model is selected as the best architectural model. Then, the bottlenecks and architectural deviations are detected by using Petri Net models. Finally, the testing of framework on multiple case studies approves the feasibility of applying process mining on the architectural evaluation. Meanwhile, it eases architectural evaluation work at every stage of the software development. In addition, it is a platform independent solution that is able to effectively detect more security and performance issues in the complex and heterogeneous architecture with less time and effort.,Ph.D | - |
| dc.language.iso | eng | - |
| dc.publisher | UKM, Bangi | - |
| dc.relation | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | - |
| dc.rights | UKM | - |
| dc.subject | Universiti Kebangsaan Malaysia -- Dissertations | - |
| dc.subject | Dissertations, Academic -- Malaysia | - |
| dc.subject | Software architecture | - |
| dc.subject | Security evaluation | - |
| dc.subject | Computer networks -- Security measures | - |
| dc.title | Lightweight framework for performance and security evaluation of software architecture using process mining | - |
| dc.type | Theses | - |
| dc.format.pages | 265 | - |
| Appears in Collections: | Faculty of Information Science and Technology / Fakulti Teknologi dan Sains Maklumat | |
Files in This Item:
| File | Description | Size | Format | |
|---|---|---|---|---|
| ukmvital_130945+Source01+Source010.PDF Restricted Access | 4.19 MB | Adobe PDF |  View/Open |
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.