Improving the scan statistic to design sensor detection systems

Fonseca, Benedito J. B.

doi:10.1109/sam.2016.7569690

Cited by 9 publications

(6 citation statements)

References 5 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Such a rule clearly ignores the spatial arrangement of the detections and thus increases false alarm probability. (ii) SST: SST [23][24][25] is based on applying the CRT inside a spatially sliding window and comparing the maximum with a threshold. Its value at the kth time frame is y k = max…”

Section: Non-statistical-based Fusion Rulesmentioning

confidence: 99%

“…The method considers only the spatial domain and its value at the

k th

time frame is defined as

y ()(, k) = \sum_{n = 1}^{N} u_{k, n},

which means that for each time frame k we determine the total number of detections of all sensors. Such a rule clearly ignores the spatial arrangement of the detections and thus increases false alarm probability. (ii) SST: SST [23–25] is based on applying the CRT inside a spatially sliding window and comparing the maximum with a threshold. Its value at the

k th

time frame is

y ()(, k) = max_{\overset{false~}{n} = 1, \dots, N - M + 1} \sum_{n = \overset{false~}{n}}^{\overset{false~}{n} + M - 1} u_{k, n},

where

1 \leq M \leq N

is the sliding window length.…”

Section: Existing Decentralised Detection Approachesmentioning

confidence: 99%

“…It is an appealing method since it requires no statistical knowledge; thus it can be applied without making any statistical assumptions. Scan Statistic Test (SST) is another fusion rule which requires no statistical knowledge [22][23][24][25]. It deals with the problem of local false alarms and spurious detections by applying a sliding window count test; thus it better employs the spatial relation between sensors' detections.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Detection of a recurrent static target using binary sensor networks

Levy¹,

Amar

2019

IET wirel. sens. syst.

View full text Add to dashboard Cite

The authors address the problem of decentralised detection of a recurrent event produced by a static uncooperative target in a wireless network of spatially distributed sensors. The presence or absence of the target is determined by applying fusion rules on local binary decisions of individual sensors. Optimal fusion rules require complete statistical knowledge about local detection and false alarm probabilities; hence, in practice they are infeasible. Previous works, such as the counting rule test, scan statistics and the generalised likelihood ratio test, have suggested sub-optimal fusion rules based on the total number of detecting sensors. Neither method takes full advantage of the spatial or temporal correlation between adjacent sensors. In practice, the decisions of adjacent sensors are dependent both in space (as the target is static) and in time (as the event reoccurs). Herein, the authors propose an approach that capitalises on the temporal and spatial arrangement of the local decisions. Simulation examples demonstrate that the detection probability of the proposed spatio-temporal rule is significantly higher than that of the common methods, and also achieves a lower probability of false alarm. Moreover, real data results show that the proposed approach is robust to strong environmental noises.

show abstract

Section: Non-statistical-based Fusion Rulesmentioning

confidence: 99%

“…The method considers only the spatial domain and its value at the

k th

time frame is defined as

y ()(, k) = \sum_{n = 1}^{N} u_{k, n},

k th

time frame is

y ()(, k) = max_{\overset{false~}{n} = 1, \dots, N - M + 1} \sum_{n = \overset{false~}{n}}^{\overset{false~}{n} + M - 1} u_{k, n},

where

1 \leq M \leq N

is the sliding window length.…”

Section: Existing Decentralised Detection Approachesmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Detection of a recurrent static target using binary sensor networks

Levy¹,

Amar

2019

IET wirel. sens. syst.

View full text Add to dashboard Cite

show abstract

“…In a distributed detection setting where sensors transmit a single bit to the fusion center, examples of such rules are the 'Or rule' [6], [7], which decides for H 1 if any sensor sends the bit '1'; or the 'Counting rule' [5], [7], [11], which decides for H 1 if multiple sensors send the bit '1'. In a setting where sensors transmit raw measurements or statistics, the 'Sum rule' [12], which decides for H 1 if the sum of measurements or statistics is above a threshold, also gives the same importance to all sensors. These rules however combine strong measurements from sensors near the emitter with weak measurements from sensors far from the emitter, reducing P D .…”

Section: Introductionmentioning

confidence: 99%

“…Although suboptimal, the scan statistic has been successfully used to detect anomalies in georeferenced data [9], [13]- [15] and many authors have recognized its value in distributed sensor detection systems [4], [12], [16]- [20]. Here, because of the signal attenuation between emitter and sensors, the change in the distribution of the measurements collected by sensors near the emitter can be considered an anomaly in the set of all measurements, motivating the use of the scan statistic.…”

Section: Introductionmentioning

confidence: 99%

A Clustering Algorithm to Improve the Scan Statistic in Sensor Detection Systems

Fonseca

2020

IEEE Access

Self Cite

View full text Add to dashboard Cite

The success of the scan statistic in detecting anomalies in georeferenced data has motivated its use in distributed sensor systems to detect an emitter at an unknown location. Sensors are grouped into clusters, cluster statistics are produced, and the scan statistic decides that the emitter is present if any cluster statistic is above a threshold. Although the scan statistic is not the optimal fusion rule, it avoids combining strong measurements from sensors near the emitter with weak measurements from sensors far from the emitter. The question that motivates this paper is: could a clustering algorithm improve the detection performance of the scan statistic? Previous studies on the scan statistics considered that the set of clusters is given or is the product of a scanning process; and previous studies on clustering algorithms for wireless sensor networks have not considered forming clusters specifically for the scan statistic. Our first goal is to highlight the opportunity of improving the scan statistic by carefully designing the cluster set. We discuss how the cluster set influences not only the detection performance, but also processing and communication in the system. Our second goal is to propose and study a new clustering algorithm to build the cluster set for the scan statistic. Although suboptimal, our algorithm produces cluster sets that reach similar or better detection performance than the usually considered cluster sets with a significantly lower number of clusters, which results in less processing and communication in the system.

show abstract

Designing Distributed Sensor Detection Systems Using the Scan Statistic

Fonseca

2024

Handbook of Scan Statistics

View full text Add to dashboard Cite

Improving the scan statistic to design sensor detection systems

Cited by 9 publications

References 5 publications

Detection of a recurrent static target using binary sensor networks

Detection of a recurrent static target using binary sensor networks

A Clustering Algorithm to Improve the Scan Statistic in Sensor Detection Systems

Designing Distributed Sensor Detection Systems Using the Scan Statistic

Contact Info

Product

Resources

About