Data Fusion and Filtering Techniques

Abstract

This chapter presents two major approaches of data fusion: the least squares (LS) approach and the Bayesian framework. The LS approach is an algorithm that estimates the position of a mobile station (MS) by minimizing the squared error between the actual measurements observed in the wireless channel and the expected measurements resulting from the estimated position. In the Bayesian framework, the position is determined as an estimator that minimizes the mean square error between the actual measurements and the expected measurements. The chapter also provides three major implementations of the Bayesian framework: the Kalman filter (KF), the particle filter (PF) and grid‐based methods. The KF is a group of methods that assumes the measurements to be corrupted by white Gaussian noise, the PF is a Monte Carlo type of algorithm that does not constrain the noise component of the Gaussian distribution, and the grid‐based methods assume that the state space is discrete and finite.