Decentralized Estimation and Control for Multisensor Systems
Decentralized Estimation and Control for Multisensor Systems
Data Fusion and Sensor Management: A Decentralized Information-Theoretic Approach
Data Fusion and Sensor Management: A Decentralized Information-Theoretic Approach
Scalable object-tracking through unattended techniques (SCOUT)
ICNP '00 Proceedings of the 2000 International Conference on Network Protocols
Energy-quality tradeoffs for target tracking in wireless sensor networks
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
Maximum mutual information principle for dynamic sensor query problems
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
Detection, classification, and collaborative tracking of multiple targets using video sensors
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
A distributed algorithm for waking-up in heterogeneous sensor networks
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
Acoustic target tracking using tiny wireless sensor devices
IPSN'03 Proceedings of the 2nd international conference on Information processing in sensor networks
| Hi-index | 0.00 |
In this paper we discuss the use of sensors sequentially for the purpose of target estimation and tracking. Such problems are motivated by scenarios where many of the sensors are inactive under normal situations with only a few active sensors "on guard". Upon the detection of a stationary or moving object of interest by an active sensor, this sensor subsequently triggers/wakes up/activates the inactive/sleeping sensors sequentially to get more measurement readings of the object (e.g., location, speed, temperature). Each activation of a sensor and the sensing and communication that follow will incur a cost. This process terminates when the error of the estimate based on the sequence of observations is satisfactory (relative to its cost), or when the moving object is out of range, or when the time period of interest has elapsed, and so on. The questions we are interested in answering are when to stop, when to use a sensor and when not to, and what is the optimal sequence of sensors to use for a given objective function. In this paper we present a framework with which the above problems are formulated into optimal sequential decision problems with energy constraints. Optimal results can then be obtained either through numerical methods or analytical derivation. Our preliminary results shQW that at least for some instances of such a formulation we can obtain well structured solutions as well as properties of these solutions that describe the sequential selection of sensors.