Tutorial @ CPSWeek 2011

Sensor Management for Tracking

This half-day tutorial will be held on the morning of Monday, April 11. The complete schedule can be found here.

 

Presenters

Prof. Venugopal V. Veeravalli, University of Illinois at Urbana-Champaign

Dr. George Atia, University of Illinois at Urbana-Champaign

 

Abstract

Large sensor networks collecting data in dynamic environments are typically composed of a distributed collection of nodes with limited energy and processing capabilities. Hence, it is imperative to efficiently manage the sensors' resources to prolong the lifetime of such networks without sacrificing performance. This tutorial seeks to address the general problem of joint sensor resource management for tracking applications. The fundamental goal is to dynamically manage and control the sensing assets and modalities in order to efficiently track the objects' physical information states (e.g., target locations, speeds, types, number) subject to constraints on various resources such as energy, communication, and computation.

 

The tutorial is structured into four major parts. In part I, we introduce the mathematical models for object(s) dynamics, the sensing system (including a description of the sensing modalities) and the observation models. Since sensor management for tracking is essentially a problem of filtering with dynamic sequential decision making under uncertainty, in part II of this tutorial we aim to provide an overview of non-linear filtering and Markov Decision Processes with special emphasis on models where the state is only partially observable to the controllers, due to noise and model uncertainties, natural limitations of the measurement devices, or incomplete data about the surroundings. In Part III, we address various problems in sensor management, including the problems of sensor scheduling, sensor sleep management, and multi-modal control. The goal is to design efficient control policies that yield the best tradeoffs between tracking performance and resource usage. Exact solutions are generally intractable even for the simplest models due to the dimensionality of the information and control spaces. Hence, in part IV of this tutorial, we discuss approximate solution techniques to design control policies ranging from simple myopic policies, through information theoretic approaches, to approximation techniques which use special surrogates to the optimal value function such as the observable-after-control and the point-based type policies.

 

Biographies

Venugopal V. Veeravalli received the Ph.D. degree in 1992 from the University of Illinois at Urbana-Champaign, the M.S. degree in 1987 from Carnegie-Mellon University, Pittsburgh, PA, and the B. Tech. degree in 1985 from the Indian Institute of Technology, Bombay, (Silver Medal Honors), all in Electrical Engineering. He joined the University of Illinois at Urbana-Champaign in 2000, where he is currently Professor in the department of Electrical and Computer Engineering, Research Professor in the Coordinated Science Laboratory, and Director of the Illinois Center for Wireless Systems (ICWS). He served as a program director for communications research at the U.S. National Science Foundation in Arlington, VA from 2003 to 2005.  He has previously held academic positions at Harvard University, Rice University, and Cornell University, and has been on sabbatical at MIT, IISc Bangalore, and Qualcomm, Inc. His research interests include distributed sensor systems and networks, wireless communications, detection and estimation theory, and information theory.  He is a Fellow of the IEEE and was on the Board of Governors of the IEEE Information Theory Society from 2004 to 2007.  He was an Associate Editor for Detection and Estimation for the IEEE Transactions on Information Theory from 2000 to 2003, and an associate editor for the IEEE Transactions on Wireless Communications from 1999 to 2000. Among the awards he has received for research and teaching are the IEEE Browder J. Thompson Best Paper Award, the National Science Foundation CAREER Award, and the Presidential Early Career Award for Scientists and Engineers (PECASE). He is a distinguished Lecturer for the IEEE Signal Processing Society for 2010-2011.

 

George Atia is currently a postdoctoral researcher with the Coordinated Science Laboratory at the University of Illinois at Urbana-Champaign. He received his Ph.D. degree in Electrical and Computer Engineering from Boston University, Massachusetts, in 2009. He received the B.Sc. and M.Sc. degrees, both in Electrical Engineering, from Alexandria University, Egypt, in 2000 and 2003, respectively. He is the recipient of many awards, including the outstanding graduate teaching fellow of the year award in 2003-2004, the 2006 College of Engineering Dean's Award at the Science and Engineering Research Symposium, and the best paper award at the International Conference on Distributed Computing in Sensor Systems (DCOSS) in 2008. His main research interests are in wireless communications, network information theory, distributed systems, compressive sensing, and detection and estimation theory.

 

Intended audience

The aim of this tutorial is to present to a broad audience the mathematical foundations, tools and algorithms for sensor management for tracking applications. The intended audience includes graduate students, academic researchers as well as industry participants who are interested in the general areas of tracking and stochastic control.

 

Tutorial Structure

Part I:   Mathematical Models

·        Object dynamics

·        Sensing modalities

·        Observation models

Part II:  Necessary background

·        Overview of non-linear filtering theory

·        Markov decision processes and POMDPs

Part III: Problems

·        Sensor scheduling

·        Sensor sleep management

·        Multi-modal control

Part IV: Solution techniques