A Statistics-Based Data Fusion for Ad-Hoc Sensor Networks

Fang WANG; Zhe WEI

doi:10.1587/transfun.E97.A.2675

A Statistics-Based Data Fusion for Ad-Hoc Sensor Networks

Fang WANG, Zhe WEI

Full Text Views

0

Share
Cite this

Summary :

Misbehaving nodes intrinsic to the physical vulnerabilities of ad-hoc sensor networks pose a challenging constraint on the designing of data fusion. To address this issue, a statistics-based reputation method for reliable data fusion is proposed in this study. Different from traditional reputation methods that only compute the general reputation of a node, the proposed method modeled by negative binomial reputation consists of two separated reputation metrics: fusion reputation and sensing reputation. Fusion reputation aims to select data fusion points and sensing reputation is used to weigh the data reported by sensor nodes to the fusion point. So, this method can prevent a compromised node from covering its misbehavior in the process of sensing or fusion by behaving well in the fusion or sensing. To tackle the unexpected facts such as packet loss, a discounting factor is introduced into the proposed method. Additionally, Local Outlier Factor (LOF) based outlier detection is applied to evaluate the behavior result of sensor nodes. Simulations show that the proposed method can enhance the reliability of data fusion and is more accurate than the general reputation method when applied in reputation evaluation.

Publication: IEICE TRANSACTIONS on Fundamentals Vol.E97-A No.12 pp.2675-2679

Publication Date: 2014/12/01

Publicized

Online ISSN: 1745-1337

DOI: 10.1587/transfun.E97.A.2675

Type of Manuscript: LETTER

Category: Mobile Information Network and Personal Communications

Authors

Fang WANG
Civil Aviation Flight University of China
Zhe WEI
Sichuan Normal University

Keyword

ad-hoc sensor networks, data fusion, misbehaving node, statistics based

Cite this

Copy

Fang WANG, Zhe WEI, "A Statistics-Based Data Fusion for Ad-Hoc Sensor Networks" in IEICE TRANSACTIONS on Fundamentals, vol. E97-A, no. 12, pp. 2675-2679, December 2014, doi: 10.1587/transfun.E97.A.2675.
Abstract: Misbehaving nodes intrinsic to the physical vulnerabilities of ad-hoc sensor networks pose a challenging constraint on the designing of data fusion. To address this issue, a statistics-based reputation method for reliable data fusion is proposed in this study. Different from traditional reputation methods that only compute the general reputation of a node, the proposed method modeled by negative binomial reputation consists of two separated reputation metrics: fusion reputation and sensing reputation. Fusion reputation aims to select data fusion points and sensing reputation is used to weigh the data reported by sensor nodes to the fusion point. So, this method can prevent a compromised node from covering its misbehavior in the process of sensing or fusion by behaving well in the fusion or sensing. To tackle the unexpected facts such as packet loss, a discounting factor is introduced into the proposed method. Additionally, Local Outlier Factor (LOF) based outlier detection is applied to evaluate the behavior result of sensor nodes. Simulations show that the proposed method can enhance the reliability of data fusion and is more accurate than the general reputation method when applied in reputation evaluation.
URL: https://globals.ieice.org/en_transactions/fundamentals/10.1587/transfun.E97.A.2675/_p

Copy

@ARTICLE{e97-a_12_2675,
author={Fang WANG, Zhe WEI, },
journal={IEICE TRANSACTIONS on Fundamentals},
title={A Statistics-Based Data Fusion for Ad-Hoc Sensor Networks},
year={2014},
volume={E97-A},
number={12},
pages={2675-2679},
abstract={Misbehaving nodes intrinsic to the physical vulnerabilities of ad-hoc sensor networks pose a challenging constraint on the designing of data fusion. To address this issue, a statistics-based reputation method for reliable data fusion is proposed in this study. Different from traditional reputation methods that only compute the general reputation of a node, the proposed method modeled by negative binomial reputation consists of two separated reputation metrics: fusion reputation and sensing reputation. Fusion reputation aims to select data fusion points and sensing reputation is used to weigh the data reported by sensor nodes to the fusion point. So, this method can prevent a compromised node from covering its misbehavior in the process of sensing or fusion by behaving well in the fusion or sensing. To tackle the unexpected facts such as packet loss, a discounting factor is introduced into the proposed method. Additionally, Local Outlier Factor (LOF) based outlier detection is applied to evaluate the behavior result of sensor nodes. Simulations show that the proposed method can enhance the reliability of data fusion and is more accurate than the general reputation method when applied in reputation evaluation.},
keywords={},
doi={10.1587/transfun.E97.A.2675},
ISSN={1745-1337},
month={December},}

Copy

TY - JOUR
TI - A Statistics-Based Data Fusion for Ad-Hoc Sensor Networks
T2 - IEICE TRANSACTIONS on Fundamentals
SP - 2675
EP - 2679
AU - Fang WANG
AU - Zhe WEI
PY - 2014
DO - 10.1587/transfun.E97.A.2675
JO - IEICE TRANSACTIONS on Fundamentals
SN - 1745-1337
VL - E97-A
IS - 12
JA - IEICE TRANSACTIONS on Fundamentals
Y1 - December 2014
AB - Misbehaving nodes intrinsic to the physical vulnerabilities of ad-hoc sensor networks pose a challenging constraint on the designing of data fusion. To address this issue, a statistics-based reputation method for reliable data fusion is proposed in this study. Different from traditional reputation methods that only compute the general reputation of a node, the proposed method modeled by negative binomial reputation consists of two separated reputation metrics: fusion reputation and sensing reputation. Fusion reputation aims to select data fusion points and sensing reputation is used to weigh the data reported by sensor nodes to the fusion point. So, this method can prevent a compromised node from covering its misbehavior in the process of sensing or fusion by behaving well in the fusion or sensing. To tackle the unexpected facts such as packet loss, a discounting factor is introduced into the proposed method. Additionally, Local Outlier Factor (LOF) based outlier detection is applied to evaluate the behavior result of sensor nodes. Simulations show that the proposed method can enhance the reliability of data fusion and is more accurate than the general reputation method when applied in reputation evaluation.
ER -