Skip to main content

WICON Research Openings

Please click here for Research Assistantships for Ph.D. Students

1. Postdoc Research Opening in Trustworthy and Physics-Informed Scientific AI

We are seeking a highly motivated Postdoctoral Researcher to play a leading role in the cutting-edge, Department of Energy (DOE) funded Genesis Mission project "PARSE: Physics-Informed Adversarial Robustness for Scientific AI-enabled Workflows."

Scientific Machine Learning (SciML) models are increasingly being used to accelerate discovery in computational fluid dynamics (CFD), climate forecasting, and more. However, these complex pipelines present unique vulnerabilities to sophisticated adversarial attacks, including data poisoning, evasion attacks, GenAI supply-chain corruption, and others. The overarching goal of the PARSE project is to develop AI-powered tools that leverage domain-specific properties, physical constraints, and system behaviors to detect and diminish the impact of adversarial attacks on scientific AI models without sacrificing macroscopic physical fidelity. The project is a collaboration between the University of Arizona and Lawrence Berkeley National Laboratory (LBNL). It aims to secure next-generation scientific AI workflows.

Key Responsibilities

As a Postdoctoral Researcher on this project, you will work at the intersection of AI, cybersecurity, and scientific computing. Your key responsibilities will include:

  • Adversarial Threat Modeling: Formulate and execute novel adversarial attacks (untargeted/targeted evasion, balanced/imbalanced training data poisoning, GenAI synthetic-data poisoning) against scientific AI workflows, targeting structured PDE trajectories, atmospheric fields, and CFD datasets.
  • Physics-Informed Defense & Sanitization: Develop automated data sanitization techniques that leverage hard constraints (conservation laws, boundary conditions, PDE residuals) to filter poisoned data both pre-ingestion and post-ingestion (in the embedding/latent space).
  • Model Hardening: Harden state-of-the-art architectures via physics-informed adversarial training, training-time influence monitoring, ensemble models, and the use of digital twins.
  • Benchmarking & Evaluation: Measure tradeoffs in robustness, predictive accuracy, and physical consistency across large-scale DOE datasets.
  • Collaboration & Publication: Work closely with cross-institutional teams (including domain scientists at DOE national labs), co-lead critical project milestones, write reports for the DOE, and publish findings in top-tier machine learning, computational science, and security venues.

Required Qualifications

  • Ph.D. in Computer Science, Applied Mathematics, Physics, Electrical/Computer Engineering, or a closely related field.
  • Strong foundational background in Machine Learning, Deep Learning, and their theoretical underpinnings.
  • Familiarity with Adversarial Machine Learning (AML), robust optimization, or AI safety/security.
  • Extensive hands-on programming experience with deep learning frameworks (e.g., PyTorch, TensorFlow, JAX) and Python.
  • Demonstrated ability to conduct independent, high-impact research, evidenced by a strong track record of peer-reviewed publications.
  • Excellent written and verbal communication skills for cross-disciplinary collaboration.

Preferred Qualifications

  • Experience with Scientific Machine Learning (SciML), such as Physics-Informed Neural Networks (PINNs), Fourier Neural Operators (FNO), or foundation models for science.
  • Domain knowledge in fluid dynamics, atmospheric modeling, or solving Partial Differential Equations (PDEs).
  • Experience processing large-scale scientific datasets (e.g., HDF5, NetCDF) and running workloads on High-Performance Computing (HPC) platforms.

What We Offer

  • The opportunity to work on high-impact, DOE-funded research that secures the future of AI in the physical sciences.
  • Close collaboration with world-class researchers at DOE National Laboratories (LBNL).
  • Access to state-of-the-art supercomputing and HPC resources (e.g., NERSC).
  • A vibrant, supportive research group dedicated to your professional growth and transition to a permanent academic or industry/national lab career.
  • Competitive salary and comprehensive benefits package.

How to Apply

Interested candidates should submit the following materials to krunz@arizona.edu with the subject line "PARSE Postdoc Application - [Your Last Name]":

  1. A cover letter detailing your research interests, relevant background, and how it aligns with the PARSE project.
  2. Detailed CV.
  3. Links to or PDFs of 2-3 relevant representative publications.
  4. Contact information for three professional references.

Review of applications will begin immediately and will continue until the position is filled. Target start date is October 2026.

2. Postdoc Opening in Machine Learning for Next-generation Wireless Communications

The Department of Electrical and Computer Engineering (ECE) at the University of Arizona invites candidates for a postdoctoral position in the area of machine learning for next-generation wireless communications. The selected candidate will be a part of and contribute to the WISPER Center (Center for Wireless Innovation towards Secure, Pervasive, Efficient, and Resilient Next G Networks) and the Wireless Communications and Networking (WICON) group at the University of Arizona. WISPER is a recently inaugurated NSF Industry University Research Cooperative Center (IUCRC), a partnership that includes several universities and industry members. 

Candidates must have a Ph.D. degree in EE/ECE/CS or a related discipline by the start date and must demonstrate experience in one or more of the following areas (as evidenced by research publications):

  • Applied AI and machine learning techniques, including experience with generative AI, reinforcement learning, transfer learning, federated learning, and/or ML-based signal intelligence. Proficiency with at least one high-level deep learning software library (TensorFlow/Keras, PyTorch, MXNet, etc.) is expected.
  • Communications protocols for wireless systems below 6 GHz (e.g., Wi-Fi, LTE, 5G) and/or millimeter-wave and sub-THz bands, including beamforming and precoding techniques.
  • Physical- and MAC-layer wireless security, including modeling of adversarial signals, LPI signal detection, characterization of mimics and rogue transmissions, obfuscation techniques, channel-based authentication, security and privacy for spectrum sharing and access, and next-generation cellular network security.

The initial appointment is for 1-2 years (depending on the candidate's preference and availability of funding), with possible extension for additional years and/or elevation to career-track assistant/associate research professor rank. The target starting date is January 15, 2025.

Interested individuals should submit by email a detailed CV, a research statement, and the names of at least three references to Prof. Marwan Krunz (krunz@arizona.edu). Only PDF attachments will be accepted. The subject field of the email message should indicate ‘Postdoctoral Position’. Applications will continue to be reviewed until the position is filled. Inquiries should be directed to krunz@arizona.edu. Only shortlisted candidates will be contacted.

Related Links:

UA ECE Dept.: https://www.ece.arizona.edu

WICON Lab: https://wireless.ece.arizona.edu