The implementation of Cognitive Wireless Sensor Network using rule-based and supervised approaches

This work proposes an adaptive learning based wireless sensor network which relies on cognitive computational process to provide a dynamic capability in configuring the network, which is called Cognitive Wireless Sensor Network (CogWSN). The network is formed by sensor nodes that equipped with cognitive modules allowing them to aware of their operating environment. The concept of CogWSN and its framework are defined. The decision process consists of Observe, Plan, Implement, and Evaluate phases. The architecture introduces Transceiver, Transducer, and Power Supply virtual modules and the modules are coordinated by CogWSN´s decision process together with the required operation by user. Three types of CogWSNs are designed for the implementation: Rule-based CogWSN, Supervised CogWSN, and other combination work. Comparison for these modules is carried out through some case studies on power transmission and communication slots allocation.