Improvement of E-Contracts Accomplishments by Self-Adaptive Fuzzy Architecture
Anderson Francisco Talon, Edmundo R. M. Madeira · 2015
The growth of web services has been increasingly noticed. Monitoring these services ensures the Quality of Service. Monitoring is the basis for violations verification, but it can also be used to predict a possible violation. The paper proposes a self-adaptive fuzzy architecture for web service monitoring, which attempts to predict a possible failure of the provider. Initially two non-functional features were monitored: response time and availability. The prediction is based on services historical data that are analyzed by a fuzzy system. With this prediction, there is a self-configuration on the system that changes service priority, making the provider processes high priority before low priority services. This prediction also helps the self-optimization of the system. It can be observed by the decrease of average response time and by the increase of average availability. It is not always possible to predict a failure, and after a failure, the system is capable of self-healing using recovering actions. Results show improvements of e-contracts accomplishments. With the fuzzy system, we observed an increase of 40.41% in e-contracts accomplishments, and a decrease of 42.64% in average response time (according to the environment), showing that the proposed architecture is promising.