Optimal distributed detection in the presence of Byzantines

Kailkhura, Bhavya; Brahma, Swastik; Han, Yunghsiang S.; Varshney, Pramod K.

doi:10.1109/icassp.2013.6638193

Cited by 39 publications

(33 citation statements)

References 13 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…As opposed to [10], due to the particular characteristics of the setup studied in [6], in some cases it is possible for the decision fusion center to make reliable decisions even when the number of Byzantines exceeds the number of honest nodes. The works that most closely resemble the present paper are [1], [11], which generalize the analysis carried out in [10]. In these works the authors consider the case of corrupted nodes, even if at the end the nodes act as in the case of corrupted reports, since the Byzantines do not take advantage of the knowledge of the observed sequences x i .…”

Section: Prior Work and Contributionmentioning

confidence: 86%

“…A first attempt to cast the decision fusion process under byzantine attacks into a game-theoretic framework can be found in [11]. In that paper, the FC is given the possibility of setting the local sensor threshold for the hypothesis testing problem at the nodes and the fusion rule, while the Byzantines can choose the flipping probability P mal .…”

Section: B the Decision Fusion Game: Definitionmentioning

confidence: 99%

“…With respect to [11], we study a more general version of the decision fusion game which includes the isolation scheme described in Section III. To this purpose, the FC is endowed with the possibility of setting the isolation threshold η, as well as the final fusion rule after removal of byzantine nodes.…”

Section: B the Decision Fusion Game: Definitionmentioning

confidence: 99%

“…Specifically, each node makes a local decision about the state of the observed system and forwards its one-bit decision to FC, which must decide between hypothesis H 0 (system is in state S 0 ) and hypothesis H 1 (system is in state S 1 ). We assume that a fraction α of the k nodes (or links, according to the definitions given in Figure 1) is under the control of byzantine attackers which, in order to make the fusion process fail, corrupt the reports by flipping the one-bit local decisions with probability P mal (as in [1], [11], we assume a symmetric attacking strategy). By referring to Figure 1, the above attack corresponds to the insertion of a binary symmetry channel with crossover probability P mal in the attacked links.…”

Section: Decision Fusion With Isolation Of Byzantine Nodes a Prmentioning

confidence: 99%

See 3 more Smart Citations

Decision fusion with corrupted reports in multi-sensor networks: A game-theoretic approach

Abrardo

Barni

Kallas

et al. 2014

53rd IEEE Conference on Decision and Control

View full text Add to dashboard Cite

Decision fusion in adversarial setting is receiving increasing attention due to its relevance in several applications including sensor networks, cognitive radio, social networks, distributed network monitoring. In most cases, a fusion center has to make a decision based on the reports provided by local agents, e.g. the nodes of a multi-sensor network. In this paper, we consider a setup in which the fusion center makes its decision on the status of an observed system by relying on the decisions made by a pool of local nodes and by taking into account the possibility that some nodes maliciously corrupt their reports to induce a decision error. We do so by casting the problem into a game-theoretic framework and looking for the existence of an equilibrium point defining the optimum strategies for the fusion center and the malicious nodes. We analyze two different strategies for the fusion center: a strategy recently introduced by Varshney et al. in a cognitive radio setup and a new approach based on soft identification of malicious nodes. The superior performance of the new decision scheme are demonstrated by resorting to the game-theoretic framework introduced previously.

show abstract

Section: Prior Work and Contributionmentioning

confidence: 86%

Section: B the Decision Fusion Game: Definitionmentioning

confidence: 99%

Section: B the Decision Fusion Game: Definitionmentioning

confidence: 99%

Section: Decision Fusion With Isolation Of Byzantine Nodes a Prmentioning

confidence: 99%

See 2 more Smart Citations

Decision fusion with corrupted reports in multi-sensor networks: A game-theoretic approach

Abrardo

Barni

Kallas

et al. 2014

53rd IEEE Conference on Decision and Control

View full text Add to dashboard Cite

show abstract

“…The robustness of distributed detection systems against attacks is of utmost importance for the functioning of distributed detection systems. The problem of optimal distributed detection with independent identical sensors in the presence of Byzantine attacks is considered in [15]. By considering an attacker to be strategic in nature, we address the issue of designing the optimal fusion rule and the local sensor thresholds that minimize the probability of error at the fusion center (FC).…”

Section: Distributed Inference In the Presence Of Byzantine (Unreliabmentioning

confidence: 99%

Multimodal Signal Processing for Personnel Detection and Activity Classification for Indoor Surveillance

Varshney¹

2013

Self Cite

View full text Add to dashboard Cite

This goal of this project was to develop novel schemes for the fusion of heterogeneous information. The target application was the detection and classification of personnel activity in both indoor and outdoor environments under dependent observations. We have identified features and designed a classifier that achieves up to 95% classification accuracy on classifying the occupancy with indoor footstep data. MDL-based copula selection strategies are investigated and a detector based on vines is designed that extends previous bivariate copula-based detectors to a multi-sensor application. Our copula-based detectors yield more than 40% improvement over the THIS PAGEThe public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. ABSTRACT Multimodal Signal Processing for Personnel Detection and Activity Classification for Indoor Surveillance Report TitleThis goal of this project was to develop novel schemes for the fusion of heterogeneous information. The target application was the detection and classification of personnel activity in both indoor and outdoor environments under dependent observations. We have identified features and designed a classifier that achieves up to 95% classification accuracy on classifying the occupancy with indoor footstep data. MDL-based copula selection strategies are investigated and a detector based on vines is designed that extends previous bivariate copula-based detectors to a multi-sensor application. Our copula-based detectors yield more than 40% improvement over the conventional data fusion methodologies. We extend our solution to quantized sensor information and demonstrate that injecting controlled noise can dramatically reduce computational complexity with insignificant performance loss. We propose and derive the Conditional Posterior Cramér-Rao Lower Bound (CPCRLB) for online tracking. We demonstrate that the PCRLB-based iterative approach converges quickly with significantly reduced computational cost as compared to a one-shot approach. Detector design in the presence of security threats, such as data falsification attacks to sensor networks, are also addressed. Error-control codes and decoding algorithms are used to reliably classify data in a network containing human agents.

show abstract