Sensor Selection for Minimizing Worst-Case Prediction Error

We study the problem of choosing the "best'' subset of ksensors to sample from among a sensor deployment of n > k sensors, in order to predict aggregate functions over all the sensor values. The sensor data being measured are assumed to be spatially correlated, in the sense that th...

Full description

Saved in:
Bibliographic Details
Published in:Information Processing in Sensor Networks; Proceedings: International Conference on Information Processing in Sensor Networks, 2008: St. Louis, MO pp. 97 - 108
Main Authors: Das, Abhimanyu, Kempe, David
Format: Conference Proceeding
Language:English
Published: Washington, DC, USA IEEE Computer Society 22.04.2008
IEEE
Series:ACM Conferences
Subjects:
Aggregates
Approximation algorithms
Biosensors
Chemical sensors
Clustering algorithms
Computer systems organization > Dependable and fault-tolerant systems and networks > Availability
Computer systems organization > Dependable and fault-tolerant systems and networks > Maintainability and maintenance
Computer systems organization > Dependable and fault-tolerant systems and networks > Reliability
estimation
Extraterrestrial measurements
Extraterrestrial phenomena
General and reference > Cross-computing tools and techniques > Measurement
General and reference > Cross-computing tools and techniques > Metrics
General and reference > Cross-computing tools and techniques > Reliability
Hardware > Communication hardware, interfaces and storage > Signal processing systems
Monitoring
Sensor phenomena and characterization
sensor placement
sensor selection
spatial correlation
Temperature sensors
estimation
spatial correlation
sensor selection
sensor placement
ISBN:9780769531571, 0769531571
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:We study the problem of choosing the "best'' subset of ksensors to sample from among a sensor deployment of n > k sensors, in order to predict aggregate functions over all the sensor values. The sensor data being measured are assumed to be spatially correlated, in the sense that the values at two sensors can differ by at most a monotonically increasing, concave function of their distance. The goal is then to select a subset of sensors so as to minimize the prediction error, assuming that the actual values at unsampled sensors are worst-case subject to the constraints imposed by their distances from sampled sensors.Even selecting sensors for the optimal prediction of the mean, maximum or minimum is NP-hard; we present approximation algorithms to select near-optimal subsets of k sensors that minimize the worst-case prediction error. In general, we show that for any aggregate function satisfying certain concavity, symmetry and monotonicity conditions, the sensor selection problem can be modeled as a k-median clustering problem, and solved using efficient approximation algorithms designed for k-median clustering.Our theoretical results are complemented by experiments on tworeal-world sensor data sets; our experiments confirm that ouralgorithms lead to prediction errors that are usually less thanthe (normalized) standard deviation of the test data, using only around 10% of the sensors.
ISBN:9780769531571
0769531571
DOI:10.1109/IPSN.2008.40