Abstract
In this paper, we present and evaluate a hardware multi-sensor architecture for objects / events detection, especially designed to enable low power and higher precision in Wireless Sensor Networks. The proposed architecture is based on fuzzy logic approach in order to improve mapping between QoS requirements for a communication processes and available hardware resources in WSN nodes. It improves the accuracy in taking decision about a system that has imprecise information. Each node introduces two types of sensors, scalar and visual. Scalar sensors perform detection and localization of objects, while the other visual is responsible to monitoring objects. A set of fuzzy rules to analyze signals collected by different sensors is used to activate rest of the sensing circuits (Le., data processing and communication) only when the event of interest in the environment is available.
We discuss in depth the internal hardware architecture of the proposed solution which is planned to reach high performance running in FPGAs circuit. Synthesis results and relevant performance comparisons with related works are presented.