Search

The lack of range information for passive measurements complicates sensor registration, requires one to address the arctangent function's singularities, and prevents one from applying existing polar to Cartesian conversion mappings. Consequently, a coordinate conversion method is developed for passive measurements using an integral representation. In addition, the existing coordinate conversion methods are shown to differ only by a constant multiplying each converted measurement and in the eigenvalues of the corresponding covariance matrix.

The general frequency modulation tracker (GFMT) for one or more GFM signals in noise using the expectation/conditional maximization (ECM) algorithm that is an extension of the expectation-maximization (EM) algorithm is developed. Several simulated examples are presented to illustrate the algorithm's performan

A maximum likelihood method is presented for training probabilistic neural networks (PNN's) using a Gaussian kernal, or Parzen window. The proposed training algorithm enables general nonlinear discrimination and is a generalization of Fisher's method for linear discrimination. Important features of maximum likelihood training for PNN's are: 1) it economizes the well known Parzen window estimator while preserving feed-forward NN architecture, 2) it utilizes class pooling to generalize classes represented by small training sets, 3) it gives smooth discriminant boundaries that often are 'piece-wise flat' for statistical robustness, 4) it is very fast computationally compared to back-propagation, and 5) it is numerically stable. The effectiveness of the proposed maximum likelihood training algorithm is assessed using nonparametric statistical methods to define tolerance intervals on PNN classification performance.

In active sonar and radar target tracking, measurements consist of position and often also include range rate. Tracking algorithms use these measurements over time to estimate target state comprising position, velocity and, where applicable, turn rate. In most cases there is an underlying assumption in the tracking algorithm that the target is a "point target" (i.e. the target has no physical extent). Another common assumption is that at most one measurement per scan originates from the target. For certain combinations of transmitted waveform and target type, the resolution of the waveform is such that the target is "over-resolved" (i.e. the sensor resolution is high enough that closely spaced scatter centers can be resolved). For such cases the point target assumption must be replaced with an extended target assumption. This work provides a methodology to exploit the extended nature of the target for the case of a rigid target whose spatial characteristics are fixed with respect to the line of motion. By employing a combination of the expectation maximization (EM) algorithm and allowing more than one measurement per scan to originate from the target, a technique is developed that uses a single scan of raw measurements that include range, bearing and range rate to provide an estimate of target position, velocity, heading and turn rate. This single scan estimate is then used in a nearly constant turn rate extended Kalman filter to provide a multi-scan estimate of the target state. [ABSTRACT FROM AUTHOR]

This paper presents a modified version of the probabilistic data association filter (PDAF) algorithm for single target tracking which allows for measurement covariances to vary within, as well as between, time scans. This modification enhances the ability to track targets in situations where the measurement errors within a time scan are highly variable. The potential for improved tracking is demonstrated by comparing the performance of the modified and the standard algorithm in tracking simulated intensity detector measurements. [ABSTRACT FROM AUTHOR]