Abuseta, Y M (2009) AutoTaSC : Model driven development for autonomic software engineering. Doctoral thesis, Liverpool John Moores University.

Preview

Text 505936.pdf - Published Version Download (29MB) | Preview

Abstract

Whilst much research progress has been achieved towards the development of autonomic software engineering tools and techniques including: policy-based management, modelbased development, service-oriented architecture and model driven architecture. They have often focused on and started from chosen object-oriented models of required software behaviour, rather than domain model including user intentions and/or software goals. Such an approach is often reported to lead to "misalignment" between business process layer and their associated computational enabling systems. This is specifically noticeable in adaptive and evolving business systems and/or processes settings. To address this long-standing problem research has over the years investigated many avenues to close the gap between business process modelling and the generation of enactment (computation) layer, which is responsive to business changes. Within this problem domain, this research sets out to study the extension of the Model Driven Development (MOD) paradigm to business/domain model, that is, how to raise the abstraction level of model-driven software development to the domain level and provide model synchronisation to trace and analyse the impact of a given model change. The main contribution of this research is the development of a MOD-based design method for autonomic systems referred to as AutoTaSC. The latter consists of a series of related models, where each of which represents the system under development at a given stage. The first and highest level model represents the abstract model referred to as the Platform Independent Model (PIM). The next model encapsulates the PIM model for the autonomic system where the autonomic capabilities and required components (such as monitor, sensor, actuator, analyser, policy, etc.) are added via some appropriate transformation rules. Targeting a specific technology involves adding, also via transformation rules, specific information related to that platform from which the Platform Specific Model (PSM) for the autonomic system is extracted. In the last stage, code can be generated for the specific platform or technology targeted in the previous stage, web services for instance. In addition, the AutoTaSC method provides a situated model synchronisation mechanism, which is designed following the autonomic systems principles. For instance, to guarantee model synchronisation each model from each AutoTaSC stage has an associated policy-based feedback control loop, which regulates its reaction to detected model change. Thus, AutaTase method model transformation approach to drive model query, view and synchronisation. The Auto'Iast? method was evaluated using a number of benchmark case-studies to test this research hypothesis including the effectiveness and generality of AutaTaSe design method.

Item Type:	Thesis (Doctoral)
Subjects:	Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software
Divisions:	Computer Science & Mathematics
Date Deposited:	14 Mar 2017 11:46
Last Modified:	03 Sep 2021 23:30
DOI or ID number:	10.24377/LJMU.t.00005927
URI:	https://researchonline.ljmu.ac.uk/id/eprint/5927

View Item