ABSTRACT
This chapter proposes an enhanced Radial Basis Neural Network Function (RBFNN)-based positioning system to integrate Inertial Navigation System and Global Positioning System (GPS). Positioning systems used in Wireless Sensor Networks deployed for gathering data in smart-cities’ Precision Agriculture, viz. smart farming and crop harvesting is a challenging problem where wireless nodes equipped with sensors and GPS modules are subject to several risks. GPS provides positioning, velocity, and time information with consistent and acceptable accuracy when there is direct line of sight to four or more satellites. The major benefit of RBFNN is its ability to utilize Radial Basis Function (RBF) networks without identifying the number of neurons in its hidden layer. The presented system is simulated and tested using real measurements from an inertial sensor and GPS mounted on a land vehicle. The proposed system architecture compromises two modes of operation: The training mode, and the prediction mode.
