BEGIN:VCALENDAR
VERSION:2.0
PRODID:Linklings LLC
BEGIN:VTIMEZONE
TZID:America/Los_Angeles
X-LIC-LOCATION:America/Los_Angeles
BEGIN:DAYLIGHT
TZOFFSETFROM:-0800
TZOFFSETTO:-0700
TZNAME:PDT
DTSTART:19700308T020000
RRULE:FREQ=YEARLY;BYMONTH=3;BYDAY=2SU
END:DAYLIGHT
BEGIN:STANDARD
TZOFFSETFROM:-0700
TZOFFSETTO:-0800
TZNAME:PST
DTSTART:19701101T020000
RRULE:FREQ=YEARLY;BYMONTH=11;BYDAY=1SU
END:STANDARD
END:VTIMEZONE
BEGIN:VEVENT
DTSTAMP:20260730T152637Z
LOCATION:
DTSTART;TZID=America/Los_Angeles:20260728T124100
DTEND;TZID=America/Los_Angeles:20260728T124200
UID:dac_DAC 2026_sess315_RESEARCH2648@linklings.com
SUMMARY:PD-Net: Learning Device-Invariant Representations for Heterogeneou
 s Cross-Device Side-Channel Attacks
DESCRIPTION:Dalin He, Wei Cheng, Yuejun Liu, Jingdian Ming, and Yongbin Zh
 ou (Nanjing University of Science and Technology)\n\nHeterogeneous cross-d
 evice side-channel attacks remain a critical yet underexplored challenge, 
 as models trained on one device often fail to generalize across architectu
 res. This paper presents PD-Net, a domain generalization framework that le
 arns device-invariant features by disentangling algorithmic content from d
 evice-specific style and aligning feature distributions using prototypical
  and Maximum Mean Discrepancy (MMD) losses. \nPD-Net is trained on nine he
 terogeneous source domains spanning ARM/AVR/FPGA and power/electromagnetic
  leakage modalities, including 32-bit ARM Cortex-M0/M1/M3/M4, 8-bit AVR AT
 mega (three series), and 128-bit Xilinx Virtex-5 FPGA, and evaluated in a 
 zero-shot setting without target-specific adaptation. \nExperimental resul
 ts demonstrate robust zero-shot cross-architecture transfers between 8-bit
  and 32-bit devices, with consistent gains over existing generalization an
 d transfer-learning approaches. In particular, PD-Net delivers 29 successf
 ul attacks with only 10 divergences across 70 settings, markedly outperfor
 ming the state of the art, which succeeds in only 4 cases and diverges 19 
 times. \nTo the best of our knowledge, this is the first domain generaliza
 tion (DG)-based deep learning framework to systematically demonstrate prac
 tical zero-shot heterogeneous cross-device side-channel attacks.\n\nTopics
 : Quantum\n\nTrack: DES6. Quantum Computing, Systems\n\nSession Chairs: Ho
 keun Kim (Arizona State University); BaekGyu Kim (Daegu Gyeongbuk Institut
 e of Science and Technology); Xiang Chen (Peking University); Sangyeob Kim
  (Yonsei University); Robert M. Radway (University of Pennsylvania); Prash
 anth Mohan (University of Pittsburgh); Xiaoxuan Yang (University of Virgin
 ia, Stanford University); Ramtin Zand (University of South Carolina); Hadi
  Kamali (University of Central Florida); Jiafeng Xie (Villanova University
 ); Prabuddha Chakraborty (University of Maine); Junichiro Kadomoto (Univer
 sity of Tokyo); and Yosuke Ueno (RIKEN)\n\n
END:VEVENT
END:VCALENDAR