Estimating Directional Data From Network Topology for Improving Tracking Performance

Slavisa Tomic,Rui Dinis,Marko Beko,Paulo Montezuma

doi:10.3390/jsan8020030

Abstract

This work proposes a novel approach for tracking a moving target in non-line-of-sight (NLOS) environments based on range estimates extracted from received signal strength (RSS) and time of arrival (TOA) measurements. By exploiting the known architecture of reference points to act as an improper antenna array and the range estimates, angle of arrival (AOA) of the signal emitted by the target is first estimated at each reference point. We then show how to take advantage of these angle estimates to convert the problem into a more convenient, polar space, where a linearization of the measurement models is easily achieved. The derived linear model serves as the main building block on top of which prior knowledge acquired during the movement of the target is incorporated by adapting a Kalman filter (KF). The performance of the proposed approach was assessed through computer simulations, which confirmed its effectiveness in combating the negative effect of NLOS bias and superiority in comparison with its naive counterpart, which does not take prior knowledge into consideration.

Highlights

Tracking a mobile target is important in many practical applications, such as autonomous surveillance, search-rescue, robotic navigation, and wildlife monitoring, to name a few [1,2,3,4,5,6,7,8,9,10,11,12,13]
By taking a closer look at Equation (20), one notices that it contains exactly what we aspired for: maximum likelihood (ML) part related to the measurement model and prior probability distribution function (PDF) part related to prior knowledge
We proposed a novel approach for tracking a moving target in adverse NLOS environments by means of combined received signal strength (RSS) and time of arrival (TOA) measurements

Summary

Introduction

Tracking a mobile target is important in many practical applications, such as autonomous surveillance, search-rescue, robotic navigation, and wildlife monitoring, to name a few [1,2,3,4,5,6,7,8,9,10,11,12,13]. Besides the referred schemes designed for RSSand RSS-AOA-based target tracking, several other classical approaches (fundamentally nothing else but alterations of the KF) can be found in the literature nowadays Among others, these include the extended KF (EKF) [39,40,41,42,43,44,45], which does not require any presumptions regarding the linearity of the state or measurement models. Instead of employing a high number of particles [51] or calculating the Jacobian matrix [52], which can considerably elevate the computational burden, the proposed approach is based on linearizing the measurement models by the use of estimated AOA information, which makes it very light in terms of computational cost.

Problem Formulation

Target Tracking

Performance Results

Conclusions and Future Work