Matches in SemOpenAlex for { <https://semopenalex.org/work/W4313118910> ?p ?o ?g. }
Showing items 1 to 79 of
79
with 100 items per page.
- W4313118910 endingPage "1833" @default.
- W4313118910 startingPage "1820" @default.
- W4313118910 abstract "Signal classification is a universal problem in adversarial wireless scenarios, especially when an eavesdropping radio receiver attempts to glean information about a target transmitter’s patterns, attributes, and contents over a wireless channel. In recent years, research surrounding the idea of Machine Learning (ML)-based signal classification has focused on modulation classification, with the downstream objective of demodulation. However, while the computer vision data domain has made significant progress in ensuring robust classification of images despite crafted perturbations, this success has not been translated to secure modulation classification. In this work, we perform the first-ever physical test of an eavesdropping ML-based modulation classifier radio, which we trained offline using a ensemble of i.i.d. models. Each model is trained with a weighted mixture of data perturbed by iterative“, least likely” white box attacks and non-attacked data. We then tested the ensemble online using coaxial-connected Software Defined Radios (SDRs). We conducted a case study comparing our results to the state-of-the-art computer vision approaches to investigate the presence of “label leaking”, model capacity sensitivity, understand the viability of parallel and sequential variations on perturbation training, and assess the effectiveness of iterative attack training. Our results show that perturbations can result in guessing-level classification performance from eavesdroppers, and that varying levels of robustness can be achieved against all presented attacks. These findings confirm that any receiver presents a new attack vector by utilizing ML techniques for classification tasks, and can be vulnerable to evasion attacks at little-to-no cost to transmitters. Consequently, we argue for the use of our training scheme in all ML-based classifying radios where security is a concern." @default.
- W4313118910 created "2023-01-06" @default.
- W4313118910 creator A5001445693 @default.
- W4313118910 creator A5011856267 @default.
- W4313118910 creator A5016717133 @default.
- W4313118910 creator A5030088093 @default.
- W4313118910 creator A5060359183 @default.
- W4313118910 date "2022-01-01" @default.
- W4313118910 modified "2023-10-06" @default.
- W4313118910 title "Countering Physical Eavesdropper Evasion with Adversarial Training" @default.
- W4313118910 doi "https://doi.org/10.1109/ojcoms.2022.3213371" @default.
- W4313118910 hasPublicationYear "2022" @default.
- W4313118910 type Work @default.
- W4313118910 citedByCount "1" @default.
- W4313118910 countsByYear W43131189102023 @default.
- W4313118910 crossrefType "journal-article" @default.
- W4313118910 hasAuthorship W4313118910A5001445693 @default.
- W4313118910 hasAuthorship W4313118910A5011856267 @default.
- W4313118910 hasAuthorship W4313118910A5016717133 @default.
- W4313118910 hasAuthorship W4313118910A5030088093 @default.
- W4313118910 hasAuthorship W4313118910A5060359183 @default.
- W4313118910 hasBestOaLocation W43131189101 @default.
- W4313118910 hasConcept C104317684 @default.
- W4313118910 hasConcept C119857082 @default.
- W4313118910 hasConcept C12267149 @default.
- W4313118910 hasConcept C124101348 @default.
- W4313118910 hasConcept C127162648 @default.
- W4313118910 hasConcept C154945302 @default.
- W4313118910 hasConcept C171115542 @default.
- W4313118910 hasConcept C185592680 @default.
- W4313118910 hasConcept C192220659 @default.
- W4313118910 hasConcept C2776788033 @default.
- W4313118910 hasConcept C31258907 @default.
- W4313118910 hasConcept C38652104 @default.
- W4313118910 hasConcept C41008148 @default.
- W4313118910 hasConcept C47798520 @default.
- W4313118910 hasConcept C55493867 @default.
- W4313118910 hasConcept C555944384 @default.
- W4313118910 hasConcept C63479239 @default.
- W4313118910 hasConcept C76155785 @default.
- W4313118910 hasConcept C95623464 @default.
- W4313118910 hasConceptScore W4313118910C104317684 @default.
- W4313118910 hasConceptScore W4313118910C119857082 @default.
- W4313118910 hasConceptScore W4313118910C12267149 @default.
- W4313118910 hasConceptScore W4313118910C124101348 @default.
- W4313118910 hasConceptScore W4313118910C127162648 @default.
- W4313118910 hasConceptScore W4313118910C154945302 @default.
- W4313118910 hasConceptScore W4313118910C171115542 @default.
- W4313118910 hasConceptScore W4313118910C185592680 @default.
- W4313118910 hasConceptScore W4313118910C192220659 @default.
- W4313118910 hasConceptScore W4313118910C2776788033 @default.
- W4313118910 hasConceptScore W4313118910C31258907 @default.
- W4313118910 hasConceptScore W4313118910C38652104 @default.
- W4313118910 hasConceptScore W4313118910C41008148 @default.
- W4313118910 hasConceptScore W4313118910C47798520 @default.
- W4313118910 hasConceptScore W4313118910C55493867 @default.
- W4313118910 hasConceptScore W4313118910C555944384 @default.
- W4313118910 hasConceptScore W4313118910C63479239 @default.
- W4313118910 hasConceptScore W4313118910C76155785 @default.
- W4313118910 hasConceptScore W4313118910C95623464 @default.
- W4313118910 hasLocation W43131189101 @default.
- W4313118910 hasOpenAccess W4313118910 @default.
- W4313118910 hasPrimaryLocation W43131189101 @default.
- W4313118910 hasRelatedWork W2024497893 @default.
- W4313118910 hasRelatedWork W2135829874 @default.
- W4313118910 hasRelatedWork W2149802390 @default.
- W4313118910 hasRelatedWork W2599313184 @default.
- W4313118910 hasRelatedWork W2944980800 @default.
- W4313118910 hasRelatedWork W2972206846 @default.
- W4313118910 hasRelatedWork W3213067907 @default.
- W4313118910 hasRelatedWork W4292969940 @default.
- W4313118910 hasRelatedWork W4379211539 @default.
- W4313118910 hasRelatedWork W4385392551 @default.
- W4313118910 hasVolume "3" @default.
- W4313118910 isParatext "false" @default.
- W4313118910 isRetracted "false" @default.
- W4313118910 workType "article" @default.