A GNU Radio SDR Framework for Bistatic Sensing Using 5G NR Zadoff-Chu and SSB Waveforms

Abstract

We present a real-time bistatic sensing chain built entirely in GNU Radio that uses 5G New Radio (NR) synchronization waveforms as the sensing signal. The approach exploits two properties of the 5G synchronization signal block (SSB): (i) its frame structure is fully specified, so a receiver can reconstruct the transmitted waveform locally as a matched-filter reference, and (ii) its synchronization sequences are short and fully deterministic. In 5G NR the PSS is a length-127 binary m-sequence rather than a Zadoff-Chu (ZC) sequence, so for the sensing reference itself we use the ZC family, which 5G NR employs for PRACH preambles and whose constant-amplitude zero-autocorrelation (CAZAC) property yields a range response free of correlation sidelobes. Because the standard SSB is transmitted at a low duty cycle, a few percent at most in the configuration considered here, we transmit the synchronization pilot continuously and use an extended ZC reference, trading standards compliance for sensing energy. The processing chain comprises adaptive least-mean-squares (LMS) direct-path and near-range clutter cancellation, FFT-based matched filtering, and an exponential-average moving-target indicator (MTI). It is implemented as streaming GNU Radio out-of-tree blocks and runs in real time at 30.72 MS/s on NI USRP-2901 hardware. MATLAB simulations of a drone target at four bistatic excess ranges (19.5–195.3 m) give target-to-clutter ratios above 35 dB with correct range and Doppler recovery. Over-the-air experiments in a 38 m indoor corridor track a walking human continuously over 4.5 - 40.4 m with a range RMSE of 3.53 m and a radial-velocity RMSE of 0.26 m/s using the extended ZC waveform, and we compare this directly against the 5G NR PSS radiated as a continuous pilot.