SPS SA-TWG Webinar: Maximum a Posteriori Penalty Function (MAP-PF) Multitarget Tracking
Date: 25 April 2024
Time: 10:00 AM ET (New York Time)
Speaker(s): Dr. Kristine Bell
This webinar is the next in a series by the IEEE Synthetic Aperture Technical Working Group (SA-TWG)
Abstract
Traditional multiple target tracking techniques partition the track estimation problem into three steps: generation of detections from sensor data, association of detections to targets, and target state estimation. A detection is an observation that consists of a target-present call and a measurement of one or more target sensor parameters (e.g. range, angle, Doppler), which are related to the target state parameters (e.g. position and velocity in Cartesian coordinates) by a fixed, known, possibly nonlinear transformation. The detection step typically consists of processing the sensor data to form a detection surface, finding peaks in the detection surface that exceed a threshold, and then extracting target sensor parameter estimates from the locations of the peaks.
The maximum a posteriori penalty function (MAP-PF) approach formulates the track estimation problem directly from the sensor data using the maximum a posteriori (MAP) estimation criterion and the penalty function (PF) method of nonlinear programming to obtain a tractable solution. The result is a two-step estimation process that is similar to traditional methods, except the processes are coupled via the penalty function and the data association step of traditional approaches is simplified. In the sensor parameter measurement step, the penalty function uses the current target states to guide the estimator to extract one measurement for each target. In the track estimation step, the penalty function acts like the measurement likelihood function.
In this talk, we present the MAP-PF approach and present a variety of passive sonar examples.
Biography
Kristine Bell (Metron, Inc.) is a Distinguished Fellow at Metron, Inc. and also holds an Affiliate Faculty position in the Statistics Department at George Mason University (GMU). Her technical expertise is in the area of statistical signal processing and multi-target tracking and her current focus is on cognitive and fully adaptive radar, sonar, and electronic warfare systems. She received the B.S. in Electrical Engineering from Rice University in 1985, and the M.S. and Ph.D. from GMU in 1990 and 1995. From 1996-2009, Dr. Bell was an Associate/Assistant Professor in the Statistics Department and C4I Center at GMU. During this time she was also a visiting researcher at the US Army Research Laboratory and the US Naval Research Laboratory. Dr. Bell has served on the IEEE Dennis J. Picard Radar Technologies Medal Selection Committee, the IEEE Jack S. Kilby Signal Processing Medal Selection Committee, the IEEE Aerospace and Electronic Systems Society (AESS) Fellow Evaluation Committee, and the AESS Radar Systems Panel, where she was the chair of the Student Paper Competition Committee. She was the chair of the IEEE Signal Processing Society’s Sensor Array and Multichannel (SAM) Technical Committee. She received the GMU Volgenau School of IT & Engineering Outstanding Alumnus Award in 2009 and the IEEE AESS Harry Rowe Mimno Best Magazine Paper Award in 2021. She is a Fellow of IEEE.
