Doppler-bearing Tracking Research Articles

Unscented Particle Filter Approach for Underwater Target Tracking

This research aims to find an appropriate approach to improve system accuracy in the Doppler-bearing tracking (DBT) problem for target estimation. The topic of DBT problem is to achieve a target trajectory using bearing and frequency measurements. The difficulty of DBT problem comes from the nonlinearity terms exposed in the measurement equations. The unscented particle filter approach is proposed to estimate the accuracy in the target motion parameters (TMP). This approach requires the observer maneuver so that the target trajectory is observable. Although in recent research papers, DBT has been proven to work efficiently without observer maneuver, TMP is unknown to the observer, and consequently, there is a need for observer maneuver. So, the algorithm is simulated with observer following s-maneuver and without any maneuver executed by the observer, and results are compared. The effectiveness of the solution and results are determined by using MATLAB simulation. It is shown that the truthfulness of the outcome is superior when the observer performs s-maneuver while compared to that without observer maneuver.

Underwater surveillance in non-Gaussian noisy environment

The aim of this paper is to evaluate the performance of different filtering algorithms in the presence of non-Gaussian noise environment for tracking underwater targets, using Doppler frequency and bearing measurements. The tracking using Doppler frequency and bearing measurements is popularly known as Doppler-bearing tracking. Here the measurements, that is, bearings and Doppler frequency, are considered to be corrupted with two types of non-Gaussian noises namely shot noise and Gaussian mixture noise. The non-Gaussian noise sampled measurements are assumed to be obtained (a) randomly throughout the process and (b) repeatedly at some particular time samples. The efficiency of these filters with the increase in non-Gaussian noise samples is discussed in this paper. The performance of filters is compared with that of Cramer-Rao Lower Bound. Doppler-bearing extended Kalman filter and Doppler-bearing unscented Kalman filter are chosen for this work.

Geometric-Polar Tracking From Bearings-Only and Doppler-Bearing Measurements

Bearings-only tracking (BOT) and Doppler-bearing tracking (DBT) perform target motion analysis from measurements of bearings, Doppler and bearings, respectively. The state equations are simple for both the target and observer in the rectangular coordinates. Hence, BOT and DBT mostly work in that coordinate. By developing the trigonometric relations between target and observer postions over time, this paper first derives the pseudolinear equations for direct range and bearing estimation. It then provides a generalized total least squares (GTLS), and a maximum likelihood solution. Polar tracking has two advantages over rectangular. First, for applications that require range and bearing, polar output avoids a coordinate conversion at every output instant. Second, polar equations often have a smaller bias than rectangular. Consequently, polar estimates can give a closer (to the optimum) initialization for iterative algorithms. Simulation results have validated the theoretical development, confirming the near optimality of the GTLS and maximum likelihood solutions.

Doppler-Bearing Tracking in the Presence of Observer Location Error

<para xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"> This paper investigates the Doppler-bearing tracking problem in the presence of random observer position error at one or two active observers. We first derive the CRLBs for the target and observer location parameters when the noise is Gaussian. The CRLB results indicate that the observer position uncertainty always leads to degradation in the target location estimation accuracy in both the two-observer and single-observer tracking scenarios. In the two-observer case, the Doppler-bearing measurements may be able to improve the observer positions while in the single-observer case, they cannot. This paper then studies the MSE of the target parameter estimate when the observer positions are assumed accurate but in fact have error. We show that for the two-observer case, the MSE is larger than the CRLB of the target parameter vector and thus, it is necessary for an algorithm to take the observer position error into account to achieve the optimum performance. On the other hand, for the single-observer case, the MSE is identical to the CRLB, indicating that the identification of the observer position can be eliminated without sacrificing the estimation accuracy of the target position. Simulations confirm the theoretical results. </para>

MANEUVERING TARGET DOPPLER-BEARING TRACKING WITH SIGNAL TIME DELAY USING INTERACTING MULTIPLE MODEL ALGORITHMS

MANEUVERING TARGET DOPPLER-BEARING TRACKING WITH SIGNAL TIME DELAY USING INTERACTING MULTIPLE MODEL ALGORITHMS

Parallelisation of the particle filtering technique and application to Doppler-bearing tracking of maneuvering sources

Particle filtering may cope with non-linear and/or non-Gaussian models. As a counterpart, this technique suffers from an heavy computation cost and cannot always satisfy the real time constraints of applications. A data parallel algorithm is proposed to achieve real-time particle filtering. Extensive results are presented in the passive SONAR tracking context for maneuvering targets.

Asymptotic theory for a two-step pseudo-linear Doppler-bearing tracker

This paper presents an analysis of a new pseudo-linear estimator for Doppler-bearing tracking, DBPLE. The parameters are estimated with a two step estimator, which consists of a bearings-only step followed by a Doppler-only step, both steps being explicit and linear. The computationally more expensive maximum likelihood estimator, MLE, solves a non-linear weighted least squares problem using bearing and frequency measurements simultaneously. There are three main approximations in the DBPLE. One of these, the two-step approach, is investigated in this paper using asymptotic maximum likelihood theory. Effects from the other approximations are studied in the simulations. If the own-ship is manoeuvring the two-step approach is severely degraded in performance. An efficient algorithm is mentioned in the paper.

Passive Doppler-bearing tracking using a pseudo-linear estimator

In this paper an explicit pseudo-linear estimator for Doppler-bearing tracking is proposed. It overcomes the problems with the bias of earlier pseudo-linear estimators and with the nonlinear frequency measurement equation by using another representation than the Cartesian one and by using the logarithm of the frequency, respectively. It is fast, inherently stable, and easy to implement. The Doppler shift is, for a nonmoving own-ship, determined by the target velocities, while the bearing rate is determined by the same velocities divided by the range. The special representation in this paper uses this difference in behavior to give fast and bias free estimation of the range. Instead of iterating a weighted feast squares problem using bearing and frequency measurements simultaneously, the true bearings and the course are estimated in a bearings only step followed by a frequency only step, which estimates range and frequency. The range estimate then gives the speed estimate. Modifications for scenarios with multiple emitted frequency lines and/or for frequency lines that disappear during parts of a scenario are shown.< <ETX xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">&gt;</ETX>