Máquina de interferência autonônica distribuída para RSSF

Abstract

Wireless sensor networks are examples of resource constrained networks (RCNs) in which processing resources, storage and energy are limited. This type of network can provide context information to Ambient Intelligence systems, which tend to overload as it increases the amount of sensor nodes in WSN and/or the degree of heterogeneity of its data types captured. This paper proposes an autonomic distributed inference machine (MIAD) that uses fuzzy logic to increase the level of semantic information in the context of WSN, and self-configures sensing intervals, dissemination of sensor nodes and redundancy message context of the monitored area. Experimental tests carried out with temperature sensors and relative humidity show that MIAD embedded in sensor nodes provides an increase of 21.8% in dispatching context information relevant to fire risk and reduces the power consumption of WSN in 18.4%. MIAD has shown better results when compared to both the producer of a distributed application framework for WSN as well as an autonomic motor based on crisp rules responsible for self-configuration of WSN.