Genetic machine learning approach for data fusion applications in dense Wireless Sensor Networks

Abstract

Wireless sensor networks (WSN) are being targeted for use in applications like security, resources monitoring and factory automation. However, the reduced available resources raise a lot of technical challenges. Self-organization in WSN is a desirable characteristic that can be achieved by means of data fusion techniques when delivering reliable data to users. In this paper it is proposed a genetic machine learning algorithm (GMLA) approach that makes a trade-off between quality of information and communication efficiency. GMLA is based on genetic algorithms and it can adapt itself dynamically to environment modifications. The main target of the proposed approach is to achieve self-organization in a WSN application with data fusion. Simulations demonstrate that the proposed approach can optimize communication efficiency in a dense WSN.