Interval Fuzzy Type-2-Based Model for Web Services’ QoS Compliance Monitoring

QoS compliance monitoring is an important component in the running of web services as it is used among others for identifying problems, evaluating services during selection process, and deciding whether to continue the subscription or not. The main problem in carrying out this monitoring is web services environment contains uncertainties due to the dynamic and unpredictable behaviors of its network. Existing models perform the monitoring based on crisp requirements. This crisp-based requirements do not have the ability to represent the uncertainties. Furthermore, requestors are generally do not have the knowledge on the realistic QoS values to be specified in requirements. Existing models also use crisp method to perform the monitoring, which is less accurate and precise to handle the uncertainties. Hence, in this research, interval fuzzy type-2 (IT2) method is experimented as the web services’ QoS compliance monitoring model. The aim of this research is to minimize the effect of uncertainties in existing crisp-based monitoring models. Another aim of this research is to introduce an approach to construct IT2 membership functions and rules using data clustering method. A web services’ QoS compliance monitoring model known as WeSQMon that utilizes IT2 method, fuzzy clustering and clustering optimization is therefore proposed. WeSQMon is introduced in two different types i.e. single input with single output (SISO) and multiple input with single output (MISO). WeSQMon is experimented to perform web services’ QoS monitoring based on fuzzy linguistic-based requirements specification. Other than that, WeSQMon is also measured in terms of accuracy and precision under the condition of uncertainties. This is performed by comparing the compliance monitoring results against 120 synthetic models that represent uncertainties. The results show that WeSQMon is able to perform web services’ QoS compliance monitoring based on fuzzy linguistic-based requirements specification. Moreover, WeSQMon improves the QoS compliance monitoring accuracy and precision where it outperforms the crisp and FT1 models. These results show that an IT2-based model can minimize the effects of uncertainties in web services’ QoS compliance monitoring.