Sensor Data Fusion Using Mutual Information Algorithm

Abstract

Traffic flow prediction is one of the congestion avoidance methods in highways. According to previous studies, no comprehensive model has been proposed for traffic flow prediction which can prevent congestion in many different traffic conditions. Using data fusion to reduce prediction error is an interesting idea to solve this problem. In this paper, a new hybrid algorithm based on mutual informat ion for traffic flow prediction will be proposed and compared with various types of previous hybrid algorithms and predictors. The Mutual Information (MI) algorithm is used to calculate the interdependency of data, so we expect this new hybrid algorithm to have high precisio n in comparison with others. Simulations will be implemented based on real data in MATLAB environment

Hybrid detection algorithm for online faulty sensors identification in wireless sensor networks

“…The basic properties of MI are as follows: MI is a measure of correlation or dependence between any two RVs [37]. MI is always positive. If any two RVs are independent, then . MI is symmetric [40]. [35].…”