IMM estimator for maneuvering target tracking with Improved Current Statistical Model

Abstract

Target tracking is one of the most important applications of Wireless Sensor Networks (WSN). It is the state estimation of a moving object based on sensor information. Interacting Multiple Model (IMM) estimator is a self adaptable, multiple model filters with less computational complexity where estimation is done by probabilistically integrating different models. This paper proposes IMM estimator with Improved Current Statistical Model (ICSM). The main feature of this estimator is ability to track the maneuvering target according to its current acceleration so that it can track accurately as its original motion. IMM is implemented with two models namely Constant velocity (CV) and ICSM to track the uniform and maneuvering targets respectively. The estimator is examined for different motions and simulation result is compared with the traditional IMM estimator. The accuracy of the proposed tracking has been validated for missing event cases and different scenarios.