Information Technology Reference
In-Depth Information
services should be able to tolerate a range of violation in their quality descrip-
tion. However, handling this toleration need to be done with special care. An
importantissuetoaddressinSBAsisto what extent the QoS parameters of a
Web Service are satisfiable. The answer to this issue could be a basic for making
adaptation decisions. However, this issue has not been addressed adequately in
the literatures. Evaluating the extent of parameter satisfaction is necessary to
help the selection of best adaptation strategy.
As an initial step to this, in [3] we provided conditions under which QoS
changes are acceptable. We used a temporal logic namely Allen's Interval Algebra
(AIA) [2] to formally specify the non-functional properties of web services. We
then used the AIA to reason about changes of quality parameters and their
evolution. In this paper, we extend [3] and propose a fuzzy approach to support
service adaptation and evolution. We define
fuzzy parameters
for QoS property
description of Web Services. Fuzzy parameters could be considered as fuzzy sets
and measured based on their value of membership. Satisfaction degree of fuzzy
parameters is measured according to their actual distance of the agreed quality
ranges in the contract. The goal of this paper is to provide flexibility for service
specification by applying fuzzy parameters. Using a fuzzy approach allows us to
deal with reasoning on the quality violations that is approximate rather than
accurate. At the end, we propose different categories of adaptation that perform
based on the satisfaction degree. Experimental results show the effectiveness
of using the fuzzy approach over the non-fuzzy one in making decisions for
adaptation.
The remainder of the paper is structured as follows. Section 2 describes the
major related work. In Section 3 we present a definition for QoS property descrip-
tion of services through introducing fuzzy parameters. In Section 4 we specify
satisfaction functions for each parameter to measure to what extend the QoS is
achieved with respect to the existing contract. We explain the decision making
mechanism in Section 5 that works based on the satisfaction degree. Section
6 provides experimental result using a simulator and evaluates the effective-
ness of the proposed approach. Section 7 concludes the paper and discusses our
future work.
2 Related Work
Deviation of quality ranges from the existing contract may produce a system
failure and bring dissatisfaction for customers. To this end, the evolution and
adaptation of web services are becoming two important issues in reacting to
the various changes in order to provide the agreed QoS stated in the contract.
Recently, many adaptation strategies and methods have been proposed in the
literature. However, most of the work in service adaptation concentrates on the
technical issues and definition of mechanisms for adaptation rather than con-
sidering QoS perspective. A list of adaptation strategies for repair processes in
SBAs is provided in [6] and [1]. For example, [7] proposed a service replacement
approach for adaptive Web Service composition and execution, while Canfora
