Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis

Daan van der Valk; Marina Krcek; Stjepan Picek; Shivam Bhasin

doi:10.1109/DAC18072.2020.9218520

Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis

Daan van der Valk, Marina Krcek, Stjepan Picek, Shivam Bhasin

Cyber Security

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

4 Citations (Scopus)

Abstract

Recently, deep learning has emerged as a powerful technique for side-channel attacks, capable of even breaking common countermeasures. Still, trained models are generally large, and thus, performing evaluation becomes resource-intensive. The resource requirements increase in realistic settings where traces can be noisy, and countermeasures are active. In this work, we exploit mimicking to compress the learned models. We demonstrate up to 300 times compression of a state-of-the-art CNN. The mimic shallow network can also achieve much better accuracy as compared to when trained on original data and even reach the performance of a deeper network.

Original language	English
Title of host publication	2020 57th ACM/IEEE Design Automation Conference, DAC 2020
Place of Publication	Piscataway
Publisher	IEEE
Pages	1-6
Number of pages	6
ISBN (Electronic)	9781450367257
ISBN (Print)	978-1-7281-5802-0
DOIs	https://doi.org/10.1109/DAC18072.2020.9218520
Publication status	Published - 2020
Event	DAC 2020: 57th ACM/IEEE Design Automation Conference - San Francisco, United States Duration: 20 Jul 2020 → 24 Jul 2020 Conference number: 57th

Publication series

Name	Proceedings - Design Automation Conference
Volume	2020-July
ISSN (Print)	0738-100x

Conference

Conference	DAC 2020
Country/Territory	United States
City	San Francisco
Period	20/07/20 → 24/07/20

Access to Document

10.1109/DAC18072.2020.9218520

Cite this

@inproceedings{13af9346b26546518dd8b7ef7526eaba,

title = "Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis",

abstract = "Recently, deep learning has emerged as a powerful technique for side-channel attacks, capable of even breaking common countermeasures. Still, trained models are generally large, and thus, performing evaluation becomes resource-intensive. The resource requirements increase in realistic settings where traces can be noisy, and countermeasures are active. In this work, we exploit mimicking to compress the learned models. We demonstrate up to 300 times compression of a state-of-the-art CNN. The mimic shallow network can also achieve much better accuracy as compared to when trained on original data and even reach the performance of a deeper network.",

author = "{van der Valk}, Daan and Marina Krcek and Stjepan Picek and Shivam Bhasin",

year = "2020",

doi = "10.1109/DAC18072.2020.9218520",

language = "English",

isbn = "978-1-7281-5802-0",

series = "Proceedings - Design Automation Conference",

publisher = "IEEE",

pages = "1--6",

booktitle = "2020 57th ACM/IEEE Design Automation Conference, DAC 2020",

address = "United States",

note = "DAC 2020 : 57th ACM/IEEE Design Automation Conference ; Conference date: 20-07-2020 Through 24-07-2020",

}

van der Valk, D, Krcek, M , Picek, S & Bhasin, S 2020, Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis. in 2020 57th ACM/IEEE Design Automation Conference, DAC 2020., 9218520, Proceedings - Design Automation Conference, vol. 2020-July, IEEE, Piscataway, pp. 1-6, DAC 2020, San Francisco, United States, 20/07/20. https://doi.org/10.1109/DAC18072.2020.9218520

Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis. / van der Valk, Daan; Krcek, Marina ; Picek, Stjepan et al.
2020 57th ACM/IEEE Design Automation Conference, DAC 2020. Piscataway: IEEE, 2020. p. 1-6 9218520 (Proceedings - Design Automation Conference; Vol. 2020-July).

Research output: Chapter in Book/Conference proceedings/Edited volume › Conference contribution › Scientific › peer-review

TY - GEN

T1 - Learning From A Big Brother

T2 - DAC 2020

AU - van der Valk, Daan

AU - Krcek, Marina

AU - Picek, Stjepan

AU - Bhasin, Shivam

N1 - Conference code: 57th

PY - 2020

Y1 - 2020

N2 - Recently, deep learning has emerged as a powerful technique for side-channel attacks, capable of even breaking common countermeasures. Still, trained models are generally large, and thus, performing evaluation becomes resource-intensive. The resource requirements increase in realistic settings where traces can be noisy, and countermeasures are active. In this work, we exploit mimicking to compress the learned models. We demonstrate up to 300 times compression of a state-of-the-art CNN. The mimic shallow network can also achieve much better accuracy as compared to when trained on original data and even reach the performance of a deeper network.

AB - Recently, deep learning has emerged as a powerful technique for side-channel attacks, capable of even breaking common countermeasures. Still, trained models are generally large, and thus, performing evaluation becomes resource-intensive. The resource requirements increase in realistic settings where traces can be noisy, and countermeasures are active. In this work, we exploit mimicking to compress the learned models. We demonstrate up to 300 times compression of a state-of-the-art CNN. The mimic shallow network can also achieve much better accuracy as compared to when trained on original data and even reach the performance of a deeper network.

UR - http://www.scopus.com/inward/record.url?scp=85093947106&partnerID=8YFLogxK

U2 - 10.1109/DAC18072.2020.9218520

DO - 10.1109/DAC18072.2020.9218520

M3 - Conference contribution

SN - 978-1-7281-5802-0

T3 - Proceedings - Design Automation Conference

SP - 1

EP - 6

BT - 2020 57th ACM/IEEE Design Automation Conference, DAC 2020

PB - IEEE

CY - Piscataway

Y2 - 20 July 2020 through 24 July 2020

ER -

Learning From A Big Brother: Mimicking Neural Networks in Profiled Side-channel Analysis

Abstract

Publication series

Conference

Access to Document

Other files and links

Fingerprint

Cite this