---
type: Scholarly Article
title: "P2W: From Power Traces to Weights Matrix - An Unconventional Transfer Learning Approach"
description: Turns power traces captured from an embedded SoC into an approximate weight matrix that initialises a new model when the original model and its training data are out of reach.
resource: https://doi.org/10.1145/3748522.3779939
tags:
  - embedded machine learning
  - power side channels
  - transfer learning
  - model extraction
  - system-on-chip
  - power side-channel analysis
generated:
  by: process:roozbeh.nl/astro-build
  at: 2026-09-06T00:00:00Z
status: stable
sources:
  - id: crossref
    resource: https://api.crossref.org/works/10.1145/3748522.3779939
    title: Crossref metadata record
  - id: publisher
    resource: https://doi.org/10.1145/3748522.3779939
    title: Publisher record (ACM)
  - id: pdf
    resource: https://roozbeh.nl/papers/P2W.pdf
    title: arXiv preprint, Feb 2025 (PDF)
  - id: preprint
    resource: https://arxiv.org/abs/2502.14968
    title: arXiv:2502.14968
doi: 10.1145/3748522.3779939
authors:
  - name: Roozbeh Siyadatzadeh
    orcid: 0000-0002-2258-4110
  - name: Fatemeh Mehrafrooz
    orcid: 0000-0001-5844-3793
  - name: Nele Mentens
    orcid: 0000-0001-8753-7895
  - name: Todor Stefanov
    orcid: 0000-0001-6006-9366
year: 2026
published: 2026-03-23
venue: Proceedings of the 41st ACM/SIGAPP Symposium on Applied Computing (SAC 2026)
venue_type: conference
publisher: ACM
pages: 653-661
web_page: https://roozbeh.nl/publications/p2w-power-traces-to-weights/
pdf: https://roozbeh.nl/papers/P2W.pdf
pdf_version: preprint
bibtex: https://roozbeh.nl/publications/p2w-power-traces-to-weights.bib
license: https://creativecommons.org/licenses/by/4.0/
funding:
  - Horizon Europe project NeuroSoC (European Union, grant 101070634)
citation_count: 0
citation_count_source: Semantic Scholar Graph API, 2026-09-02
---

# P2W: From Power Traces to Weights Matrix - An Unconventional Transfer Learning Approach

Roozbeh Siyadatzadeh, Fatemeh Mehrafrooz, Nele Mentens, Todor Stefanov. Proceedings of the 41st ACM/SIGAPP Symposium on Applied Computing (SAC 2026), 2026, pp. 653-661. DOI: [10.1145/3748522.3779939](https://doi.org/10.1145/3748522.3779939).

# Abstract

The rapid growth of deploying machine learning (ML) models within embedded systems on a chip (SoCs) has led to transformative shifts in fields like healthcare and autonomous vehicles. One of the primary challenges for training such embedded ML models is the lack of publicly available high-quality training data. Transfer learning approaches address this challenge by utilizing the knowledge encapsulated in an existing ML model as a starting point for training a new ML model. However, existing transfer learning approaches require direct access to the existing model which is not always feasible, especially for ML models deployed on embedded SoCs. Therefore, in this paper, we introduce a novel unconventional transfer learning approach to train a new ML model by extracting and using weights from an existing ML model running on an embedded SoC without having access to the model within the SoC. Our approach captures power consumption measurements from the SoC while it is executing the ML model and translates them to an approximated weights matrix used to initialize the new ML model. This improves the learning efficiency and predictive performance of the new model, especially in scenarios with limited data available to train the model. Our novel approach can effectively increase the accuracy of the new ML model up to 3 times compared to classical training methods using the same amount of limited training data.

# Key result

Up to 3 times higher accuracy than classical training on the same limited dataset, without any access to the model inside the SoC.

# Cite

R. Siyadatzadeh, F. Mehrafrooz, N. Mentens and T. Stefanov, "P2W: From Power Traces to Weights Matrix - An Unconventional Transfer Learning Approach," in Proceedings of the 41st ACM/SIGAPP Symposium on Applied Computing (SAC 2026), Thessaloniki, Greece, 2026, pp. 653-661, doi: 10.1145/3748522.3779939.

```bibtex
@inproceedings{siyadatzadeh2026p2w,
  author    = {Siyadatzadeh, Roozbeh and Mehrafrooz, Fatemeh and Mentens, Nele and Stefanov, Todor},
  title     = {{P2W}: From Power Traces to Weights Matrix - An Unconventional Transfer Learning Approach},
  booktitle = {Proceedings of the 41st ACM/SIGAPP Symposium on Applied Computing (SAC 2026)},
  year      = {2026},
  pages     = {653--661},
  publisher = {ACM},
  address   = {Thessaloniki, Greece},
  doi       = {10.1145/3748522.3779939},
  url       = {https://doi.org/10.1145/3748522.3779939},
}
```

# Files

* [Web page](https://roozbeh.nl/publications/p2w-power-traces-to-weights/)
* [arXiv preprint, Feb 2025 (PDF)](https://roozbeh.nl/papers/P2W.pdf)
* [arXiv:2502.14968](https://arxiv.org/abs/2502.14968)
* [AI-generated audio overview (NotebookLM), 5 min](https://roozbeh.nl/audio/P2W.m4a)
* [AI-generated video overview (NotebookLM), 7 min](https://roozbeh.nl/videos/P2W.mp4)

# Related

* [NeuroSoC: security assessment of an in-memory computing SoC](../projects/neurosoc.md)
* [Security Challenges in Analog In-Memory Computing](../notes/imc-security.md)
* [From Power Traces to Model Weights: P2W and TraceFormer](../notes/weight-extraction.md)

Author record: [Roozbeh Siyadatzadeh](../person/roozbeh-siyadatzadeh.md).