Cell Sensing: Traffic detection

This work presents a passive sensing system for traffic monitoring using ambient Long Term Evolution (LTE) signals as a non-intrusive and scalable alternative to traditional surveillance methods. The approach employs a dual-receiver architecture analyzing Channel State Information (CSI) to isolate differential Doppler shifts induced by moving targets, effectively mitigating hardware-induced phase impairments. Implemented with a Software Defined Radio (SDR) platform and srsRAN software, the system demonstrated over 90% detection accuracy for speeds above 6000 mm/min in controlled indoor tests, and provided reliable speed estimations for pedestrians and vehicles in outdoor evaluations. Despite challenges at low speeds, directional ambiguity, and multipath fading in urban settings, the results validate LTE-based passive sensing as a feasible traffic monitoring method, identifying critical areas for future research such as angle-of-arrival (AoA) integration, machine learning, and real-time embedded system development.