Robert Heath
University of California-San Diego, CA·Engineering
Faster cellular networks increasingly depend on cell towers and devices that use many antennas at once. This approach is called multiple-input multiple-output (MIMO) communication.
- Fields
Engineering
Computer Science
- Topics
wireless communications
multiple-input multiple-output
extreme MIMO
- Grants
- 3 known grants, through Sep 2029
What they work on
- 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.
- Develop algorithms, including machine-learning methods, for over-the-air beam calibration and multi-user beam training in tri-hybrid MIMO, which combines digital beamforming, analog beamforming and antenna reconfiguration.
- Characterize the tradeoff between capacity and energy efficiency in extreme MIMO tri-hybrid arrays with hundreds of antennas, using simulation, prototype testbeds and digital twins (virtual replicas of the system).
- Create online reconfiguration algorithms and AI-native channel state information feedback, then demonstrate them on a manufactured, lab-deployed tri-hybrid array serving mobile users.
- Design distributed learning methods that let devices with different antenna hardware jointly predict MIMO beams and optimize codebooks (sets of preset beam patterns), including ways to represent hardware differences for machine learning.
More about the research
More antennas can carry more data to more users, but they also draw more power. Phones, cars, drones and robots also differ widely in their antenna hardware, which makes them hard to configure well. Across three projects, the research asks how to build and control large antenna arrays that are both high-performing and energy-efficient. It combines communication theory, electromagnetics, circuit models and machine learning to prepare for 5G-and-beyond and 6G cellular systems.
Why it matters
Large antenna arrays that use less power could lower the operating costs and energy use of cellular networks. They could also improve data rates and reliability for phone users. Better beam control for diverse, fast-moving devices could support wearables, autonomous vehicles, and emergency and critical communications in future 6G systems. The work may also guide how distributed machine learning is run over wireless networks with varied hardware.
Known grants
From NIH and NSF research grants. Labs often have other funding we can't see.
Collaborative Research: VINES: Track 1: Tri-Hybrid Optimization and Reconfiguration (THOR) for Extreme MIMO Systems
NSF · Oct 2026 – Sep 2029
Tri-hybrid MIMO communication with metasurface antenna
NSF · Aug 2024 – Jul 2027
NSF-AoF: SOLID: System-wide Operation via Learning In-device Dissimilarities
NSF · Jan 2024 – Sep 2026
Closest labs
Taejoon Kim
Arizona State University Campus Immersion, AZ·Engineering
Develop Zero Trust Chain security software that lets military squads share situational awareness over untrusted 5G networks, using O-RAN Intelligent Controllers on the OAIC platform to detect and block threats in near-real time.
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.
Hualiang Zhang
University of Massachusetts-Lowell, MA·Engineering
Design and fabricate photonically driven THz reflectarray unit cells built from hybrid gold–germanium mesa-array semiconductor structures whose reflection phase changes under light.
The school
University of California-San Diego: 84 labs on PI Pages, $1.9B a year on research, 637 research doctorates a year.
Reach out
A start for your email. Fill in the brackets.