Blind Detection of Drones using OFDM-Based Zadoff-Chu Sequences with Field Tests

In recent years, drones, or unmanned aerial vehicles (UAVs), have become widely used across various applications, from aerial photography and videography to the delivery of packages and medical supplies. However, their increasing presence has raised concerns about physical safety and privacy, highlighting the need for effective drone detection and monitoring solutions. To address this, we utilize the fact that most commercial drones use the Zadoff-Chu (ZC) sequence as the synchronization sequence in their communications, making it a useful feature for detection. Yet, detecting the ZC sequence blindly is challenging, as the transmitter's frequency is unknown to the receiver. While existing studies on ZC sequence detection with different frequency offsets focus largely on Long Term Evolution (LTE) scenarios, the ZC sequence structure and length used by drones differ, leading to unique detection challenges. In this paper, we analyze the autocorrelation properties of the specific ZC sequence used by drones under various center frequency offsets. We further propose a blind detection and identification algorithm that can detect and identify multiple drones utilizing ZC sequences in their video transmission protocols and autocorrelation properties. We study the performance of the proposed algorithm with extensive simulations and field tests. Even in low signal-to-noise ratio (SNR) conditions, with an SNR as low as -14 dB, our algorithm achieves a detection rate exceeding 99 %.