Asymptotic Results for Decentralized Detection in Power Constrained Wireless Sensor Networks

Chamberland, Jean-François; Veeravalli, Venugopal V.

doi:10.1109/jsac.2004.830894

Cited by 248 publications

(170 citation statements)

References 19 publications

Supporting

Mentioning

168

Contrasting

Order By: Relevance

“…The network aims to minimize the probability of error or some other cost function at the fusion center, by choosing optimal transmission functions and fusion rules. Various properties and variants of the decentralized detection problem in a parallel configuration have been extensively studied over the last twenty-five years; examples include the following: [4][5][6][7][8] study the properties of optimal fusion rules and quantizers at sensor nodes; [9] shows the existence of optimal strategies, and proves that likelihood ratio quantizers are optimal for a large class of problems including the decentralized detection problem; and [10][11][12][13][14] consider constrained decentralized detection. The reader is referred to [15,16] for a survey of the work done in this area.…”

Section: Introductionmentioning

confidence: 99%

Error Exponents for Decentralized Detection in Tree Networks

Tay

Tsitsiklis

2008

Networked Sensing Information and Control

View full text Add to dashboard Cite

Summary. We consider the problem of decentralized detection in a sensor network consisting of nodes arranged as a tree of bounded height. We characterize the optimal error exponent under a Neyman-Pearson formulation, and show that the Type II error probability decays exponentially with the number of nodes. Surprisingly, the optimal error exponent is often the same as that corresponding to a parallel configuration. We provide sufficient, as well as necessary, conditions for this to happen. We also consider the impact of failure-prone sensors or unreliable communications between sensors on the detection performance. Simple strategies that nearly achieve the asymptotically optimal performance in these cases are also developed.

show abstract

Section: Introductionmentioning

confidence: 99%

Error Exponents for Decentralized Detection in Tree Networks

Tay

Tsitsiklis

2008

Networked Sensing Information and Control

View full text Add to dashboard Cite

show abstract

“…One step further to a more realistic scenario for sensornet with large amount of sensors is to allow orthogonal channel allocation such that signals sent by each sensor go through independent and orthogonal wireless links (e.g. TDMA, FDMA), where distortion and interference are possibly introduced, to the fusion center [1], [2]. However, we can foresee that coordination and resulting overheads are overwhelming in order to achieve exact orthogonalization in large scale sensor networks.…”

Section: Introductionmentioning

confidence: 99%

Spreading Sequence-Based Non-coherent Sensor Fusion and its Resulting Large Deviation Exponents

Wei

2007

2007 IEEE International Conference on Acoustics, Speech and Signal Processing - ICASSP '07

View full text Add to dashboard Cite

Abstract-To address the coordination issue of sensors communicating with a fusion center, we propose a spreading sequence based non-coherent detection scheme for sensor networks to reduce the coordination between sensors to the largest extent. In this scheme, sensors employ independent spreading sequences to transmit their measurements. Non-coherent detection is conducted at the fusion center where only statistics regarding channel gains and sensor measurement uncertainties are needed. To evaluate the detector's performance, we first derive the large deviation exponents of detection error probabilities and then compare them with the approaches assuming orthogonal channel allocation (e.g.TDMA/FDMA). Numerical and simulation results demonstrate the dependence of large deviation exponent on the asymptotic number of sensors per chip (defined as c), as well as the better performance of our proposed scheme than the one using non-coherent detection with orthogonal link, for some c. Keywords: Spreading Sequence, Large Deviation Exponents and Non-coherent Detection.I. INTRODUCTIONIn this paper, we are mainly concerned with how to effectively send data from a set of sensor nodes to a fusion center via a one-hop network. In particular, we are interested in how to reduce the coordination among sensor nodes to the minimum and how fusion performance in terms of binary detection error probability varies with respect to the size of sensornet as a consequence of such reduced coordinations.In a one-hop sensor network for data fusion, the conventional wisdom was the measured data from each sensor is immediately available at the fusion center. This holds when sensors are traditional radars and a wired link exists between each front end radar and the fusion center. One step further to a more realistic scenario for sensornet with large amount of sensors is to allow orthogonal channel allocation such that signals sent by each sensor go through independent and orthogonal wireless links (e.g. TDMA, FDMA), where distortion and interference are possibly introduced, to the fusion center[1], [2]. However, we can foresee that coordination and resulting overheads are overwhelming in order to achieve exact orthogonalization in large scale sensor networks.Recently, A. Anandkumar and L. Tong based transmission schemes, where each element of possible quantized outputs is assigned an orthogonal spreading code. At the fusion center, what matters to detector's performance is the superposed signals corresponding to each orthogonal sequence. However, they have shown that when channel experiences zero mean fading, the detection error no longer scales exponentially with respect to the number of sensors.When we use sensor net to facilitate a binary decision at a fusion center, the ultimate objective is to distinguish between two hypothesis corresponding to the presence or absence of a target. Therefore, there is no necessity to faithfully reconstruct the measurements taken at each sensor node as long as the fusion/detection performance satisfies our need...

show abstract

“…Detection on wireless sensor networks [4,5] assumed orthogonal schemes like TDMA, FDMA or CDMA. TBMA was proposed as a multi-access scheme by Mergen and Tong [6] and by Liu and Sayeed [2], independently.…”

Section: Related Workmentioning

confidence: 99%

“…LDP characterizes the probability of large excursions of Y from its "mean" behavior by quantifying the so-called rate function I(·) [12]. In essence for Y satisfying LDP, 4 Pr{Y ∈ B (y)} = e −λ(I(y)+O( ))+o(λ) . = e −λI(y) .…”

Section: Minimum Rate Detectormentioning

confidence: 99%

See 1 more Smart Citation

A Large Deviation Analysis of Detection Over Multi-Access Channels with Random Number of Sensors

Anandkumar

Tong

2006 IEEE International Conference on Acoustics Speed and Signal Processing Proceedings

View full text Add to dashboard Cite

We consider the problem of distributed detection over a multiaccess channel. Assuming a random number of sensors transmitting their observations using Type Based Multiple Access, we derive the detection performance using Large Deviations Principle as the mean number of sensors goes to infinity. We characterize the performance in terms of error exponents. We provide comparison with the case when the number of sensors is deterministic. We generalize this scheme to multiple collections, propose a Minimum Sum-Rate detector and characterize its error exponents.

show abstract

Asymptotic Results for Decentralized Detection in Power Constrained Wireless Sensor Networks

Cited by 248 publications

References 19 publications

Error Exponents for Decentralized Detection in Tree Networks

Error Exponents for Decentralized Detection in Tree Networks

Spreading Sequence-Based Non-coherent Sensor Fusion and its Resulting Large Deviation Exponents

A Large Deviation Analysis of Detection Over Multi-Access Channels with Random Number of Sensors

Contact Info

Product

Resources

About