EdgeRIC: Delivering Real-time Intelligence to Radio Access Networks

sample text

  • EdgeRIC is a platform for real-time AI-in-the-loop for decision and control in cellular networks. It is designed to access network and application-level information to execute AI-optimized and other policies in real-time (sub-millisecond) .

EdgeRIC Focus Areas

sample text

Demo Videos

Anti Jamming with BeamArmor

Controlling the MIMO weights in realtime to steer the beam null in the direction of jammer

System Performance Optimization with AI based scheduling

Controlling the scheduling decision with a Reinforcement Learning based policy that was trained to maximize the overall system throughput observed

Current Status

sample text
This release is based on srsRAN 4G, we will shortly release our 5G version. We suport:
  • Realtime KPI monitoring:

    • UL/ DL channel conditions

    • UL/ DL pending buffers for transmission

    • Per UE throughput

    • Baseband I/Q sample monitoring

  • Realtime Control actions:

    • MAC scheduling on the downlink, both AI training and inference capability

    • Control for MIMO weights

Current Publications

Empowering Real-time Intelligent Optimization and Control in NextG Cellular Networks Paper: EdgeRIC

Code: Github Respository

Website: https://wcsng.ucsd.edu/edgeric/

Funding

This work was funded primarily by NSF Grants CNS 2312978, CNS 2312979 and in part by CNS 1955696, ECCS 2030245, ARO grant W911NF- 19-1-0367.

Contact us


Ushasi Ghosh: PhD student, UCSD, ughosh@ucsd.edu
Woo Hyun Ko: Senior Research Engineer, TAMU, whko@tamu.edu
Ish Kumar Jain: Professor, RPI, ikjain@ucsd.edu
Dinesh Bharadia: Professor, UCSD, dineshb@ucsd.edu
Srinivas Shakkottai: Professor, TAMU, sshakkot@tamu.edu

Getting Started

Events

EdgeRIC Architecture

EdgeRIC tutorials

Dataset