Distributed Autonomic Inference Machine for Wireless Sensor Networks

Wireless Sensor Networks (WSNs) offer data to Intelligence Ambient system, but, due the big number of sensor nodes and data heterogeneity, it can be overload by them. This paper proposes MIAD, a distributed autonomic inference machine which uses fuzzy logic to make ambient context and to self-configure sensing and dissemination rates and minimize redundant context of WSN. Tests with Crossbow micaz motes and temperature and relative humidity sensors show that MIAD sends more relevant risk fire context messages to final system while it saves WSN energy. It presents better results than distributed WSN application without self-configuration and an autonomic engine based on crisp rules.