Reasoning techniques for analysis and refinement of policies for service management
File(s)DTR05-7.pdf (1.44 MB)
Published version
Author(s)
Type
Report
Abstract
The work described in this technical report falls under the general problem of developing
methods that would allow us to engineer software systems that are reliable and would offer a
certain acceptable level of quality in their operation. This report shows how the analysis and
refinement of policies for Quality of Service can be carried out within logic by exploiting forms
of abductive and argumentative reasoning. In particular, it provides two main contributions. The
first is an extension of earlier work on the use of abductive reasoning for automatic policy
refinement by exploiting the use of integrity constraints within abduction and its integration with
constraint solving. This has allowed us to enhance this refinement process in various ways, e.g.
supporting parameter values derivation to quantify abstract refinement to specific policies ready
to be put in operation, and calculating utility values to determine optimal refined policies. The
second contribution is a new approach for modelling and formulating Quality of Service policies,
and more general policies for software requirements, as preference policies within logical
frameworks of argumentation. This is shown to be a flexible and declarative approach to the
analysis of such policies through high-level semantic queries of argumentation, demonstrated
here for the particular case of network firewall policies where the logical framework of
argumentation allows us to detect anomalies in the firewalls and facilitates the process of their
resolution. To our knowledge this is the first time that the link between argumentation and the
specification and analysis of requirement policies has been studied.
methods that would allow us to engineer software systems that are reliable and would offer a
certain acceptable level of quality in their operation. This report shows how the analysis and
refinement of policies for Quality of Service can be carried out within logic by exploiting forms
of abductive and argumentative reasoning. In particular, it provides two main contributions. The
first is an extension of earlier work on the use of abductive reasoning for automatic policy
refinement by exploiting the use of integrity constraints within abduction and its integration with
constraint solving. This has allowed us to enhance this refinement process in various ways, e.g.
supporting parameter values derivation to quantify abstract refinement to specific policies ready
to be put in operation, and calculating utility values to determine optimal refined policies. The
second contribution is a new approach for modelling and formulating Quality of Service policies,
and more general policies for software requirements, as preference policies within logical
frameworks of argumentation. This is shown to be a flexible and declarative approach to the
analysis of such policies through high-level semantic queries of argumentation, demonstrated
here for the particular case of network firewall policies where the logical framework of
argumentation allows us to detect anomalies in the firewalls and facilitates the process of their
resolution. To our knowledge this is the first time that the link between argumentation and the
specification and analysis of requirement policies has been studied.
Date Issued
2005-01-01
Citation
Departmental Technical Report: 05/7, 2005, pp.1-64
Publisher
Department of Computing, Imperial College London
Start Page
1
End Page
64
Journal / Book Title
Departmental Technical Report: 05/7
Copyright Statement
© 2005 The Author(s). This report is available open access under a CC-BY-NC-ND (https://creativecommons.org/licenses/by-nc-nd/4.0/)
Publication Status
Published
Article Number
05/7