PIPages
Labs

Dinesh Bharadia

University of California-San Diego, CA·Engineering

Wireless networks are getting crowded and harder to manage. New uses such as augmented reality, self-driving cars and remote healthcare need connections that are fast, reliable and nearly instant.

Write an email
Fields

Engineering

Computer Science

Topics

wireless networking

cellular networks

NextG

Grants
2 known grants, through Sep 2027

What they work on

  1. Develop real-time reinforcement learning algorithms that schedule cellular radio resources by application priority, using the structure of the optimal policy to learn quickly and at scale.
  2. Create robust, fast-adapting and hierarchical learning policies, drawing on meta and federated reinforcement learning, that work across network layers and sites as conditions change.
  3. Build and test EdgeRIC, an open platform for sub-second intelligent control of NextG cellular networks, implementing learning algorithms with the OpenAI Gym toolkit.
  4. Build a low-cost SweepSpace sensor from off-the-shelf parts using a software-defined radio and a sweeping phased-array antenna to scan frequencies and directions.
  5. Reconstruct full spatial and time maps of radio activity from sparse sensor samples, and use them to find spectrum gaps where new transmissions can be safely added.

More about the research

Meanwhile the radio spectrum, the band of frequencies these devices share, is limited. This lab works on making wireless systems smarter and more efficient. One collaborative project uses machine learning to control cellular networks in real time. The other builds low-cost sensors that track who is using which frequencies, and in which direction, so more devices can share the spectrum safely.

Why it matters

Real-time learning-based control could help future cellular networks reliably support augmented reality, autonomous vehicles and remote healthcare. Low-cost directional spectrum sensors could let regulators and operators fit more wireless communication and sensing into the same crowded spectrum. That would reduce interference without raising spectrum costs.

Known grants

From NIH and NSF research grants. Labs often have other funding we can't see.

Collaborative Research: NeTS: Medium: EdgeRIC: Empowering Real-time Intelligent Control and Optimization for NextG Cellular Radio Access Networks

NSF · Oct 2023 – Sep 2027

Collaborative Research: SII-NRDZ: SweepSpace: Enabling Autonomous Fine-Grained Spatial Spectrum Sensing and Sharing

NSF · Oct 2022 – Sep 2026

Closest labs

Alireza Vahid

Rochester Institute of Technology, NY·Engineering

Develop algorithms that recover sets of similar gene sequences from unordered DNA sequencing reads, to detect antimicrobial resistance genes in microbial communities.

Vuk Marojevic

Mississippi State University, MS·Engineering

Build a passive microwave radiometer flown on a drone on the AERPAW testbed to collect the first large active/passive radio coexistence dataset with ground-truth information.

Aly Sabri Abdalla

Mississippi State University, MS·Engineering

Characterize O-RAN attacks caused by out-of-sync state changes, timing-induced protocol deviations, malicious user devices and unstable control-app decisions.

The school

University of California-San Diego: 84 labs on PI Pages, $1.9B a year on research, 637 research doctorates a year.

More about University of California-San Diego

Reach out

A start for your email. Fill in the brackets.