Matches in SemOpenAlex for { <https://semopenalex.org/work/W2990187689> ?p ?o ?g. }
- W2990187689 abstract "Differential Power Analysis (DPA) has been an active area of research for the past two decades to study the attacks for extracting secret information from cryptographic implementations through power measurements and their defenses. Unfortunately, the research on power side-channels have so far predominantly focused on analyzing implementations of ciphers such as AES, DES, RSA, and recently post-quantum cryptography primitives (e.g., lattices). Meanwhile, machine-learning, and in particular deep-learning applications are becoming ubiquitous with several scenarios where the Machine Learning Models are Intellectual Properties requiring confidentiality. Expanding side-channel analysis to Machine Learning Model extraction, however, is largely unexplored. This paper expands the DPA framework to neural-network classifiers. First, it shows DPA attacks during inference to extract the secret model parameters such as weights and biases of a neural network. Second, it proposes the $textit{first countermeasures}$ against these attacks by augmenting $textit{masking}$. The resulting design uses novel masked components such as masked adder trees for fully-connected layers and masked Rectifier Linear Units for activation functions. On a SAKURA-X FPGA board, experiments show that the first-order DPA attacks on the unprotected implementation can succeed with only 200 traces and our protection respectively increases the latency and area-cost by 2.8x and 2.3x." @default.
- W2990187689 created "2019-12-05" @default.
- W2990187689 creator A5008761074 @default.
- W2990187689 creator A5062027964 @default.
- W2990187689 creator A5085883266 @default.
- W2990187689 date "2019-10-29" @default.
- W2990187689 modified "2023-09-27" @default.
- W2990187689 title "MaskedNet: The First Hardware Inference Engine Aiming Power Side-Channel Protection" @default.
- W2990187689 cites W1490557050 @default.
- W2990187689 cites W1503689022 @default.
- W2990187689 cites W1562542037 @default.
- W2990187689 cites W1564536482 @default.
- W2990187689 cites W1862426464 @default.
- W2990187689 cites W1882350379 @default.
- W2990187689 cites W1992291252 @default.
- W2990187689 cites W2005006301 @default.
- W2990187689 cites W2021705882 @default.
- W2990187689 cites W2037844954 @default.
- W2990187689 cites W2053832511 @default.
- W2990187689 cites W2127059139 @default.
- W2990187689 cites W2144630005 @default.
- W2990187689 cites W2154909745 @default.
- W2990187689 cites W2180612164 @default.
- W2990187689 cites W2260663238 @default.
- W2990187689 cites W2284081067 @default.
- W2990187689 cites W2293768274 @default.
- W2990187689 cites W2294868896 @default.
- W2990187689 cites W2296452361 @default.
- W2990187689 cites W2300242332 @default.
- W2990187689 cites W2319920447 @default.
- W2990187689 cites W2466176639 @default.
- W2990187689 cites W2484027757 @default.
- W2990187689 cites W2525440592 @default.
- W2990187689 cites W2535690855 @default.
- W2990187689 cites W2535955463 @default.
- W2990187689 cites W2556867355 @default.
- W2990187689 cites W2586046457 @default.
- W2990187689 cites W2591922920 @default.
- W2990187689 cites W2603766943 @default.
- W2990187689 cites W2724189024 @default.
- W2990187689 cites W2739889860 @default.
- W2990187689 cites W2789993878 @default.
- W2990187689 cites W2794898833 @default.
- W2990187689 cites W2798781612 @default.
- W2990187689 cites W2802314446 @default.
- W2990187689 cites W2809523935 @default.
- W2990187689 cites W2811276992 @default.
- W2990187689 cites W2891810898 @default.
- W2990187689 cites W2892344663 @default.
- W2990187689 cites W2892525958 @default.
- W2990187689 cites W2902207657 @default.
- W2990187689 cites W2902280009 @default.
- W2990187689 cites W2914572864 @default.
- W2990187689 cites W2928083746 @default.
- W2990187689 cites W2942216431 @default.
- W2990187689 cites W2963844355 @default.
- W2990187689 cites W2964318098 @default.
- W2990187689 cites W2965565691 @default.
- W2990187689 cites W3049152512 @default.
- W2990187689 cites W3104216513 @default.
- W2990187689 cites W2785910692 @default.
- W2990187689 cites W2927948414 @default.
- W2990187689 cites W2947331147 @default.
- W2990187689 hasPublicationYear "2019" @default.
- W2990187689 type Work @default.
- W2990187689 sameAs 2990187689 @default.
- W2990187689 citedByCount "3" @default.
- W2990187689 countsByYear W29901876892020 @default.
- W2990187689 countsByYear W29901876892021 @default.
- W2990187689 crossrefType "posted-content" @default.
- W2990187689 hasAuthorship W2990187689A5008761074 @default.
- W2990187689 hasAuthorship W2990187689A5062027964 @default.
- W2990187689 hasAuthorship W2990187689A5085883266 @default.
- W2990187689 hasConcept C108583219 @default.
- W2990187689 hasConcept C113775141 @default.
- W2990187689 hasConcept C11413529 @default.
- W2990187689 hasConcept C142362112 @default.
- W2990187689 hasConcept C149635348 @default.
- W2990187689 hasConcept C153349607 @default.
- W2990187689 hasConcept C154945302 @default.
- W2990187689 hasConcept C178489894 @default.
- W2990187689 hasConcept C2776214188 @default.
- W2990187689 hasConcept C2777402240 @default.
- W2990187689 hasConcept C41008148 @default.
- W2990187689 hasConcept C42935608 @default.
- W2990187689 hasConcept C49289754 @default.
- W2990187689 hasConcept C50644808 @default.
- W2990187689 hasConcept C71743495 @default.
- W2990187689 hasConceptScore W2990187689C108583219 @default.
- W2990187689 hasConceptScore W2990187689C113775141 @default.
- W2990187689 hasConceptScore W2990187689C11413529 @default.
- W2990187689 hasConceptScore W2990187689C142362112 @default.
- W2990187689 hasConceptScore W2990187689C149635348 @default.
- W2990187689 hasConceptScore W2990187689C153349607 @default.
- W2990187689 hasConceptScore W2990187689C154945302 @default.
- W2990187689 hasConceptScore W2990187689C178489894 @default.
- W2990187689 hasConceptScore W2990187689C2776214188 @default.
- W2990187689 hasConceptScore W2990187689C2777402240 @default.
- W2990187689 hasConceptScore W2990187689C41008148 @default.
- W2990187689 hasConceptScore W2990187689C42935608 @default.