Hessam Mahdavifar
Northeastern University, MA·Engineering·Early-career
Every phone call, text message and stored file depends on codes, which add structured redundancy so that data survive noise and errors. Good code designs have historically been rare and have depended on individual insight.
- Fields
Engineering
Computer Science
Mathematics & Statistics
- Topics
coding theory
error-correcting codes
channel coding
- Grants
- 2 known grants, through May 2027
What they work on
- Generalize Plotkin Transform codes, the family containing Reed–Muller and polar codes, through nonlinear parameterizations, and search for new codes using machine learning, deep learning and gradient-descent optimization.
- Develop new decoding algorithms for the standardized Reed–Muller and polar codes, the latter used in 5G cellular networks, using machine learning tools, and release them as open-source code.
- Build a framework, including a metric space and an analog operator channel model, for non-coherent communication over wireless networks, and construct subspace codes for these channels with characterized performance.
- Characterize fundamental limits on compressing low-rank matrices, develop low-rank subspace recovery from constrained observations, and use the duality with subspace codes to design explicit compression mechanisms.
- Design subspace-coded distributed computation schemes that efficiently compute the outcomes of algorithms operating over matrices and subspaces while minimizing delay.
More about the research
Meanwhile, today's enormous data volumes and huge wireless networks strain classical approaches. Both projects seek new ways to encode, decode and compress information. One uses machine learning to discover codes, and the other uses linear-algebra objects called subspaces to carry and compress data.
Why it matters
Faster, data-driven code design could improve the error-correcting codes built into cellphones and cellular networks, including better decoders for the standardized polar and Reed–Muller codes. Subspace coding and compression could help future large wireless networks, storage systems and distributed computing platforms move and process massive datasets more efficiently and reliably.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Milad Marvian
University of New Mexico, NM·Physics·Early-career
Develop mathematical models describing how errors propagate through quantum circuits, and use them to construct resource-efficient fault-tolerant computing subroutines.
Zhiying Wang
University of California-Irvine, CA·Engineering·Early-career
Establish the fundamental limit on information density for composite DNA storage, where each position is a set mixture of A, C, G and T, and design coding schemes that reach it given symbol choice and storage-channel errors.
Hulya Seferoglu
University of Illinois Chicago, IL·Engineering·Early-career
Characterize the cost-benefit trade-offs of coded computation, which adds redundancy to tolerate slow or failed workers, in practical edge computing systems.
The school
Northeastern University: 27 labs on PI Pages, $347M a year on research, 251 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.