WorldCat Identities

Kastella, Keith

Overview
Works: 14 works in 24 publications in 1 language and 48 library holdings
Roles: Editor, Author
Publication Timeline
.
Most widely held works by Keith Kastella
Foundations and applications of sensor management by Alfred O Hero( )

7 editions published between 2006 and 2008 in English and held by 12 WorldCat member libraries worldwide

Foundations and Applications of Sensor Management presents the emerging theory of sensor management with applications to real-world examples such as landmine detection, adaptive signal and image sampling, multi-target tracking, and radar waveform scheduling. It is written by leading experts in the field for a diverse engineering audience ranging from signal processing, to automatic control, statistics, and machine learning. The level of treatment of the book is tutorial and self-contained. The chapters of the book follow a logical development from theoretical foundations to approximate approaches and ending with applications. The coverage includes the following topics: stochastic control foundations of sensor management; multi-armed bandits and their connections to sensor management; information-theoretic approaches; managed sensing for multi-target tracking; approximation methods based on embedded simulation; active learning for classification and sampling; and waveform scheduling for radar. An appendix is included to provide essential background on topics the reader may not have encountered as a first-year graduate student: Markov decision processes; information theory; and stopping times. Foundations and Applications of Sensor Management is an important reference for signal processing and control engineers and researchers as well as machine learning application developers
A relativistic approach to quantum-mechanical path integrals by Keith Kastella( )

1 edition published in 1980 in English and held by 1 WorldCat member library worldwide

Discrimination gain to optimize detection and classification by University of Minnesota( Book )

1 edition published in 1995 in English and held by 1 WorldCat member library worldwide

Sensor management by Avner Friedman( Book )

3 editions published in 1996 in English and held by 1 WorldCat member library worldwide

Several sensor management schemes based on information theoretic metrics such as discrimination gain have been proposed, motivated by the generality of such schemes and their ability to accommodate mixed types of information such as kinematic and classification data. On the other hand, there are many methods for managing a single sensor to optimize detection. This paper compares the performance against low signal noise ratio targets of a discrimination gain scheme with three such single sensor detection schemes: the Wald test, an index policy that is optimal under certain circumstances and an alert-confirm' scheme modeled on methods used in some existing radars. For the situation where the index policy is optimal, it outperforms discrimination gain by a slight margin. However, the index policy assumes that there is only one target present. It performs poorly when there are multiple targets while discrimination gain and the Wald test continue to perform well. In addition, we show how discrimination gain can be extended to multisensor / multitarget detection and classification problems that are difficult for these other methods. One issue that arises with the use of discrimination gain as a metric is that it depends on both the current density and an a priori distribution. We examine the dependence of discrimination gain on this prior and find that while the discrimination depends on the prior, the gain is prior independent
Sensor management using discrimination gain and interacting multiple model Kalman filters by University of Minnesota( Book )

2 editions published in 1998 in English and held by 1 WorldCat member library worldwide

This paper describes an algorithm using a discrimination-based sensor effectiveness metric for sensor assignment in multisensor multitarget tracking applications. The algorithm uses Interacting Multiple Model Kalman Filters to track airborne targets with measurements obtained from two or more agile-beam radar systems. Each radar has capacity constraints on the number of targets it can observe on each scan. For each scan the expected discrimination gain is computed for the sensor target pairings. The constrained globally optimum assignment of sensors to targets is then computed and applied. This is compared to a fixed assignment schedule in simulation testing. We find that discrimination based assignment improves track accuracy as measured by both the root-mean-square position error and a measure of the total covariance
A nonlinear filter for real-time joint tracking and recognition by University of Minnesota( Book )

1 edition published in 1998 in English and held by 1 WorldCat member library worldwide

Tracking algorithms for air traffic control applications by Keith Kastella( )

1 edition published in 1995 in English and held by 1 WorldCat member library worldwide

Foundations and Applications of Sensor Management. Signals and Communication Technology( )

1 edition published in 2008 in English and held by 0 WorldCat member libraries worldwide

Sensor management is an enabling technology for the next generation of agile, multi-modal, and multi-waveform sensor platforms to efficiently perform tasks such as target detection, tracking, and identification. In sensor management, the sequence of sensor actions, such as pointing angle, modality, or waveform, are selected adaptively based on information extracted from past measurements. This book presents the theory of sensor management with applications to real world examples such as adaptive mine detection, adaptive signal and image sampling, multiple target tracking, and radar waveform design. It is written by leading experts in the field for a diverse engineering audience ranging from signal processing, to automatic control, mathematical statistics, and machine learning. The level of treatment of the book is tutorial and self contained. The chapters of the book are grouped into three sections: theoretical foundations; approximate approaches; and applications. The book assumes the reader has a technical background at the level of a first year graduate student in one of the systems engineering disciplines, e.g. signal processing, control, or communications. An appendix is included on topics that the reader may not have seen as a first year graduate student such as: partially observable markov processes, statistical decision theory, information theory, and dynamic programming
Multiple Model Particle Filtering For Multi-Target Tracking( )

1 edition published in 2004 in English and held by 0 WorldCat member libraries worldwide

This paper addresses the problem of tracking multiple moving targets by recursively estimating the joint multitarget probability density (JMPD). Estimation of the JMPD is done in a Bayesian framework and provides a method for tracking multiple targets which allow nonlinear target motion and measurement to state coupling as well as non-Gaussian target-state densities. We utilize an implementation of the JMPD method based on particle filtering (PF) techniques. The details of this method have been presented elsewhere 1. One feature of real targets is that they are poorly described by a single kinematic model Target behavior may change dramatically i.e. targets may stop moving or begin rapid acceleration. To address this fact we evaluate the use of the adaptive target tracking strategy known as the interacting multiple model (IMM) algorithm. The IMM uses multiple models for target behavior and adaptively determines which model(s) are the most appropriate at each time step based on sensor measurements. We demonstrate the applicability of the IMM to a PF-based multitarget tracker in two settings. First we consider the traditional application of tracking targets that switch between kinematic modes. The target motion used is field data recorded during a military battle simulation and includes multiple modes of target behavior. Our investigation is distinguished from prior efforts in that it is concerned with multiple targets and real target motion data and utilizes a PF implementation. Second we present a nontraditional reinterpretation of the multiple model filter as multiple models on the state of the filter rather than on the state of the target. We find that this strategy is able to automatically detect model violations and compensate by altering the filter model which results in improved target tracking
Sensor Management Research( )

2 editions published between 1996 and 1997 in English and held by 0 WorldCat member libraries worldwide

This grant is supporting development of mathematical foundations for sensor management systems. This year's accomplishments are in three areas: Extension of a Kalman-filter based discrimination metric to interacting multiple model filters; extension of sensor management based on Joint Multitarget Probabilities to incorporate multiple sensor modes and target classification; and development of fast methods to solve the Fokker-Planck equation for real-time non-linear filtering applications. To support sensor management representations of multi-target probability densities must be developed that model the uncertainty between quantities such as the number of targets, their locations and their class. To solve this problem and study it in a simple setting, the notion of Joint Multitarget Probabilities for detection, tracking, and target classification was developed and tested. In certain cases the time-evolution of these probabilities is characterized by a partial differential equation called the Fokker-Planck equation leading to a nonlinear filter. Several prototype nonlinear filters using the Alternating Direction Implicit scheme to solve the Fokker-Planck equation in real-time were formulated. In these applications it seems to offer significant improvement in estimation performance at a supportable cost in computational load
Sensor Management and Nonlinear Filtering Research( )

1 edition published in 1998 in English and held by 0 WorldCat member libraries worldwide

This grant is supporting development of mathematical foundations for sensor management and nonlinear filtering. The accomplishments so far are in two areas: (1) The use of Interactive Multiple Model Kalman Filters (IMMKF) with a metric called discrimination gain (DG); and (2) the use of nonlinear filtering, (NLF) in the tracking of target elevation for objects flying close to a reflecting surface. In the case of IMMKF, we demonstrate, using simulated data, that IMMKF can be used to compute the information gain when multiple sensors observe a collection of maneuvering airborne targets. In the case of NLF, we demonstrate the feasibility of using NLF methods for altitude tracking in multipath
Comparison of Sensor Management Strategies for Detection and Classification( )

1 edition published in 1996 in English and held by 0 WorldCat member libraries worldwide

Several sensor management schemes based on information theoretic metrics such as discrimination gain have been proposed, motivated by the generality of such schemes and their ability to accommodate mixed types of information such as kinematic and classification data. On the other hand, there are many methods for managing a single sensor to optimize detection. This paper compares the performance against low signal noise ratio targets of a discrimination gain scheme with three such single sensor detection schemes: the Wald test, an index policy that is optimal under certain circumstances and an 'alert/confirm' scheme modeled on methods used in some radars. For the situation where the index policy is optimal, it outperforms discrimination gain by a slight margin. However, the index policy assumes that there is only one target present. It performs poorly when there are multiple targets while discrimination gain and the Wald test continue to perform well. In addition, we show how discrimination gain can be extended to multisensor / multitarget detection and classification problems that are difficult for these other methods. One issue that arises with the use of discrimination gain as a metric is that it depends on both the current density and an a priori distribution. We examine the dependence of discrimination gain on this prior and find that while the discrimination depends on the prior, the gain is prior independent
Adaptive Multi-modality Sensor Scheduling for Detection and Tracking of Smart Targets( )

1 edition published in 2004 in English and held by 0 WorldCat member libraries worldwide

This paper considers the problem of sensor scheduling for the purposes of detection and tracking of "smart" targets. Smart targets are targets that are able to detect when they are under surveillance and react in a manner that makes future surveillance more difficult. We take a reinforcement learning approach to adaptively schedule a multi-modality sensor so as to most quickly and effectively detect the presence of smart targets and track them as they travel through a surveillance region. An optimal scheduling strategy, which would simultaneously address the issue of target detection and tracking, is very challenging computationally. To avoid this difficulty, we advocate a two stage approach where targets are first detected and then handed off to the tracking algorithm
Emerging Applications in Probability (Sensor Management)( )

1 edition published in 1995 in English and held by 0 WorldCat member libraries worldwide

This grant from the Air Force Office of Scientific Research supported research in Sensor Management related to the IMA 1993-94 academic year program "EMERGING APPLICATIONS OF PROBABILITY". It provided partial support for residency of Keith Kastella, an industrial researcher at Unisys Government Systems, to pursue research on the use of discrimination gain optimization for sensor management in Air Traffic Control. Manufacturing, Robotics, Remote Sensing and Defense Applications. Grant AF/F49620-94-1-0275 supported the publication of a technical research report submitted by Dr Kastella for inclusion in the IMA Preprint Series. This paper has been submitted to the IEEE Transactions on Systems, Man and Cybernetics
 
moreShow More Titles
fewerShow Fewer Titles
Audience Level
0
Audience Level
1
  Kids General Special  
Audience level: 0.60 (from 0.46 for A relativi ... to 0.99 for Emerging A ...)

Foundations and applications of sensor management
Languages
English (24)

Covers