A Risk-Based Approach to Sensor Resource Management

We investigate the benefits of employing a suitable risk-based metric to determine in real-time the high level actions that an agile sensor should execute during a mission. Faced with a barrage of competing goals, a sensor resource manager must optimize system performance while simultaneously meeting all requirements. Numerous authors advocate the use of information-theoretic measures for driving sensor tasking algorithms, wherein the relative value of different sensing actions is calculated in terms of the expected gain in information. In this chapter, motivated by the sensor resource allocation problem in missile defense, we deviate from the information-based trend and propose an approach for determining sensor tasking decisions based on risk, or expected loss of defended assets. We present results of a missile defense simulation that illustrate the advantages of our risk-based objective function over its information-theoretic and rule-based counterparts.

[1]  Chee Chong,et al.  A rollout algorithm to coordinate multiple sensor resources to track and discriminate targets , 2006, SPIE Defense + Commercial Sensing.

[2]  M. Kalandros.  Covariance control for multisensor systems , 2002 .

[3]  Dimitri P. Bertsekas,et al.  Dynamic Programming and Optimal Control, Two Volume Set , 1995 .

[4]  Kenneth J. Hintz,et al.  Information theoretic approach to sensor scheduling , 1996, Defense, Security, and Sensing.

[5]  Andrew G. Barto,et al.  Reinforcement learning , 1998 .

[6]  Andrew W. Moore,et al.  Reinforcement Learning: A Survey , 1996, J. Artif. Intell. Res..

[7]  Kenneth J. Hintz,et al.  A measure of the information gain attributable to cueing , 1991, IEEE Trans. Syst. Man Cybern..

[8]  Warren B. Powell,et al.  Handbook of Learning and Approximate Dynamic Programming , 2006, IEEE Transactions on Automatic Control.

[9]  Richard S. Sutton,et al.  Reinforcement Learning: An Introduction , 1998, IEEE Trans. Neural Networks.

[10]  Michael A. Bender,et al.  Sensor resource management for an airborne early warning radar , 2002, SPIE Defense + Commercial Sensing.

[11]  Keith Kastella.  Discrimination gain to optimize detection and classification , 1997, IEEE Trans. Syst. Man Cybern. Part A.

[12]  Martin L. Puterman,et al.  Markov Decision Processes: Discrete Stochastic Dynamic Programming , 1994 .

[13]  John N. Tsitsiklis,et al.  Neuro-Dynamic Programming , 1996, Encyclopedia of Machine Learning.

[14]  Warren B. Powell,et al.  Hierarchical Decision Making , 2004 .

[15]  Bonaventure Intercontinental,et al.  ON DECISION AND CONTROL , 1985 .

[16]  F. Bolderheij.  Mission-driven sensor management , 2008 .

[17]  P. van Genderen,et al.  A risk-based object-oriented approach to sensor management , 2005, 2005 7th International Conference on Information Fusion.

[18]  Chris Kreucher.  An information-based approach to sensor resource allocation. , 2005 .

[19]  Keith D. Kastella,et al.  Information-based sensor management and IMMKF , 1998, Defense, Security, and Sensing.

[20]  Eugene S. McVey,et al.  Multi-process constrained estimation , 1991, IEEE Trans. Syst. Man Cybern..

[21]  K. Kastella,et al.  A Comparison of Task Driven and Information Driven Sensor Management for Target Tracking , 2005, Proceedings of the 44th IEEE Conference on Decision and Control.

[22]  Wayne W. Schmaedeke.  Information-based sensor management , 1993, Defense, Security, and Sensing.