Self-configuration Using Artificial Neural Networks

Self configuration is one of the major properties of self-managing systems that requires the real time processing while adding or removing any existing file or component to maintain the proper working state of the system. In order to achieve self configuration capability, artificial neural networks based self-management technique is proposed in this paper. Artificial Neural Networks (ANN) are capable to solve real-time complex problems that may not be resolved trivially by other learning techniques. In this paper, we propose a self-managing algorithm for autonomic system based on ANN. A prototype of self-configuration using ANN is implemented using autonomic forest fire application. The performance results show that ANN is an effective technique in case of dynamic learning in general and autonomic computing in special.