Mingjie Lin
University of Central Florida, FL·Engineering
Wireless networks are running short of radio spectrum, especially below 6 GHz. Massive MIMO (multi-input multi-output) helps by sending signals through large, dense arrays of antennas.
- Fields
Engineering
Computer Science
- Topics
massive MIMO
mMIMO
wireless communications
- Grants
- 1 known grant, through Aug 2027
What they work on
- Design a dual-input hybrid load modulated balanced amplifier (DI-HLMBA), a reconfigurable radio power amplifier, that delivers high efficiency, bandwidth and linearity in massive MIMO transmitters.
- Formulate real-time tuning of the amplifier against antenna mismatch and temperature rise as a reinforcement learning problem based on nonstationary Markov decision processes.
- Implement the learning-based controller on reconfigurable FPGA hardware using a meta-stability-based strategy for low-latency, real-time optimization.
- Extend AI-assisted, multi-band operation from single transmitters to full massive MIMO arrays using a wideband fractal-shaped antenna array design.
More about the research
But packing many high-power radio transmitters together causes problems. Neighboring antennas disturb each other's electrical load, causing an impedance mismatch, so the amplifiers waste energy and heat up sharply. Today's transmitters are designed as if load and temperature stay fixed. This project aims to make them adapt in real time using data and artificial intelligence.
Why it matters
Next-generation wireless networks need to carry more data within limited spectrum while using less power. Transmitters that adapt in real time to antenna interference and heat could raise both the energy and spectrum efficiency of base stations. The same AI-based tuning framework could also apply to phased array radars, wireless imaging and sensing, and wireless power transfer.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Closest labs
Robert Heath
University of California-San Diego, CA·Engineering
Build MIMO input-output models of dynamic metasurface antennas (low-power panels of tunable elements) that capture waveguide loss, coupling between elements and reconfiguration effects, using existing circuit and antenna models.
Subhanshu Gupta
Washington State University, WA·Engineering
Use the frequency-dependent beams of true-time-delay arrays to cut beam-training overhead, and design a low-latency protocol that trains beams and estimates interference at the same time in dense millimeter-wave networks.
Yasaman Ghasempour
Princeton University, NJ·Engineering·Early-career
Steer narrow terahertz beams electronically using a new multi-slit antenna, in which the beam's angle depends on the signal frequency and the slit geometry, to overcome the high signal loss at frequencies above 100 GHz.
The school
University of Central Florida: 15 labs on PI Pages, $285M a year on research, 290 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.