High & low-level features modelling of nodes in WSNs using SystemC

Performing an adequate modeling of sensors in contemporary sensor networks can be difficult due to the need to include characteristics of high and low level of the entire network into a single software model. This paper presents a novel approach for modeling the nodes in a sensor network, as well as its integration into the network using a programmable parametric structure. The proposed approach was developed in SystemC language considering the properties of this language which perfectly fits the needs of both the hardware description of the nodes, as the complex algorithms that can run on them. The proposed model allows to include several physical node features such as data formats, connections between components and memory, and real limitations in runtime, among others. The developed model is applied to a geoelectrical prospection network in order to demonstrate in practice its advantages and application possibilities, and considering that in such kind of networks nodes act as sensors or actuator, depending on the operation mode in which it is working. Two classic network topologies, chain and bus, are used to test the model with a set of different parameter values. Performance metrics are used to compare the network topologies and their interaction with the node set size, in order to verify the validity of the proposed approach.