Research & Papers

CellSense turns 5G towers into radars with 94% detection accuracy

Sub-6 GHz ISAC system achieves 0.33m outdoor tracking error

Deep Dive

CellSense, developed by Bibhor Kumar, Ish Kumar Jain, and Vijay K Shah at NCSU, is a novel Integrated Sensing and Communication (ISAC) system natively integrated into the 5G cellular protocol stack. Unlike prior sub-6 GHz sensing work that focused on communication, CellSense uses existing cellular signals for passive target tracking without dedicated radar hardware. It leverages OFDM link-level simulations via Sionna and a USRP hardware prototype running OpenAirInterface (OAI) to validate performance in both indoor and outdoor environments.

Performance results are striking: in an indoor warehouse, CellSense achieves 74% detection probability with a 1.43m localization error; outdoors on the NCSU Centennial Oval, detection rises to 94% with sub-meter error of 0.33m. Hardware experiments in a highly cluttered laboratory confirm 1.28m accuracy and 76% detection. The system also analyzes the communication-sensing tradeoff by quantifying how pilot symbol density impacts throughput versus sensing accuracy, offering critical insights for future ISAC deployment in sub-6 GHz bands.

Key Points
  • Integrated into 5G protocol stack using OpenAirInterface and USRP hardware
  • Outdoor tracking achieves 94% detection probability with 0.33m error
  • Indoor warehouse performance: 74% detection, 1.43m error; lab prototype: 76% detection, 1.28m error

Why It Matters

Enables cost-effective, radar-like sensing using existing 5G infrastructure for logistics, smart cities, and industrial automation.

📬 Get the top 10 AI stories daily