Performance bounds for the rate-constrained universal decentralized estimators

Jin Jun Xiao; Zhi Quan Luo; Georgios B. Giannakis

doi:10.1109/LSP.2006.881522

Performance bounds for the rate-constrained universal decentralized estimators

Jin Jun Xiao, Zhi Quan Luo, Georgios B. Giannakis

Electrical and Computer Engineering

Research output: Contribution to journal › Article › peer-review

21 Scopus citations

Abstract

We consider decentralized estimation of a noise-corrupted deterministic parameter using a bandwidth-constrained sensor network with a fusion center (FC). Each sensor's noise is additive, zero mean, and independent across sensors. A decentralized estimator is said to be universal if the local sensor quantization rules and the final fusion rule at the FC are independent of sensor noise pdf. Assuming that information rate from each sensor to the FC is constrained to one bit per sample, we derive a Cramér-Rao lower bound (CRLB) on the mean-squared error (MSE) performance of a class of rate-constrained universal decentralized estimators. Our results show that if sensor observation noise has finite range in [-U,U], then the minimum MSE performance of any one-bit rate-constrained universal decentralized estimator is at least U²/(4K), where K is the total number sensors. This bound implies that the recently proposed universal decentralized estimators are optimal up to a constant factor of 4.

Original language	English (US)
Pages (from-to)	47-50
Number of pages	4
Journal	IEEE Signal Processing Letters
Volume	14
Issue number	1
DOIs	https://doi.org/10.1109/LSP.2006.881522
State	Published - Jan 2007

Bibliographical note

Funding Information:
Manuscript received January 22, 2006; revised June 9, 2006. This work was supported in part by the National Science Foundation under Grant DMS-0312416, in part by the USDOD ARMY under Grant W911NF-05-1-0567, and in part by Army Research Office under Grant 49453-MA. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Hongbin Li.

Keywords

Cramér-Rao bound
Distributed estimation
Quantization
Sensor network

Access

10.1109/LSP.2006.881522

OpenUrl availability

Full text

Cite this

@article{dbf2a82b5e5e42d0bc89e9cd653c2be2,

title = "Performance bounds for the rate-constrained universal decentralized estimators",

abstract = "We consider decentralized estimation of a noise-corrupted deterministic parameter using a bandwidth-constrained sensor network with a fusion center (FC). Each sensor's noise is additive, zero mean, and independent across sensors. A decentralized estimator is said to be universal if the local sensor quantization rules and the final fusion rule at the FC are independent of sensor noise pdf. Assuming that information rate from each sensor to the FC is constrained to one bit per sample, we derive a Cram{\'e}r-Rao lower bound (CRLB) on the mean-squared error (MSE) performance of a class of rate-constrained universal decentralized estimators. Our results show that if sensor observation noise has finite range in [-U,U], then the minimum MSE performance of any one-bit rate-constrained universal decentralized estimator is at least U2/(4K), where K is the total number sensors. This bound implies that the recently proposed universal decentralized estimators are optimal up to a constant factor of 4.",

keywords = "Cram{\'e}r-Rao bound, Distributed estimation, Quantization, Sensor network",

author = "Xiao, {Jin Jun} and Luo, {Zhi Quan} and Giannakis, {Georgios B.}",

note = "Funding Information: Manuscript received January 22, 2006; revised June 9, 2006. This work was supported in part by the National Science Foundation under Grant DMS-0312416, in part by the USDOD ARMY under Grant W911NF-05-1-0567, and in part by Army Research Office under Grant 49453-MA. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Hongbin Li.",

year = "2007",

month = jan,

doi = "10.1109/LSP.2006.881522",

language = "English (US)",

volume = "14",

pages = "47--50",

journal = "IEEE Signal Processing Letters",

issn = "1070-9908",

publisher = "Institute of Electrical and Electronics Engineers Inc.",

number = "1",

}

TY - JOUR

T1 - Performance bounds for the rate-constrained universal decentralized estimators

AU - Xiao, Jin Jun

AU - Luo, Zhi Quan

AU - Giannakis, Georgios B.

N1 - Funding Information: Manuscript received January 22, 2006; revised June 9, 2006. This work was supported in part by the National Science Foundation under Grant DMS-0312416, in part by the USDOD ARMY under Grant W911NF-05-1-0567, and in part by Army Research Office under Grant 49453-MA. The associate editor coordinating the review of this manuscript and approving it for publication was Dr. Hongbin Li.

PY - 2007/1

Y1 - 2007/1

N2 - We consider decentralized estimation of a noise-corrupted deterministic parameter using a bandwidth-constrained sensor network with a fusion center (FC). Each sensor's noise is additive, zero mean, and independent across sensors. A decentralized estimator is said to be universal if the local sensor quantization rules and the final fusion rule at the FC are independent of sensor noise pdf. Assuming that information rate from each sensor to the FC is constrained to one bit per sample, we derive a Cramér-Rao lower bound (CRLB) on the mean-squared error (MSE) performance of a class of rate-constrained universal decentralized estimators. Our results show that if sensor observation noise has finite range in [-U,U], then the minimum MSE performance of any one-bit rate-constrained universal decentralized estimator is at least U2/(4K), where K is the total number sensors. This bound implies that the recently proposed universal decentralized estimators are optimal up to a constant factor of 4.

AB - We consider decentralized estimation of a noise-corrupted deterministic parameter using a bandwidth-constrained sensor network with a fusion center (FC). Each sensor's noise is additive, zero mean, and independent across sensors. A decentralized estimator is said to be universal if the local sensor quantization rules and the final fusion rule at the FC are independent of sensor noise pdf. Assuming that information rate from each sensor to the FC is constrained to one bit per sample, we derive a Cramér-Rao lower bound (CRLB) on the mean-squared error (MSE) performance of a class of rate-constrained universal decentralized estimators. Our results show that if sensor observation noise has finite range in [-U,U], then the minimum MSE performance of any one-bit rate-constrained universal decentralized estimator is at least U2/(4K), where K is the total number sensors. This bound implies that the recently proposed universal decentralized estimators are optimal up to a constant factor of 4.

KW - Cramér-Rao bound

KW - Distributed estimation

KW - Quantization

KW - Sensor network

UR - http://www.scopus.com/inward/record.url?scp=33947278360&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=33947278360&partnerID=8YFLogxK

U2 - 10.1109/LSP.2006.881522

DO - 10.1109/LSP.2006.881522

M3 - Article

AN - SCOPUS:33947278360

SN - 1070-9908

VL - 14

SP - 47

EP - 50

JO - IEEE Signal Processing Letters

JF - IEEE Signal Processing Letters

IS - 1

ER -

Performance bounds for the rate-constrained universal decentralized estimators

Abstract

Bibliographical note

Keywords

Access

OpenUrl availability

Other files and links

Fingerprint

Cite this