Performance of Zadoff-Chu Sequences for Time Synchronization

Distributed coherent beamforming requires close alignment of transceivers in frequency, time, and phase. Common methods for time synchronization rely on knowledge of transmission and reception times of a known signal. The arrival time is estimated by computing the correlation of the received signal with the known signal. An interpolation method, such as quadratic or higher-order polynomial interpolation, is used to estimate the location of the correlation peak with sub-sample precision. The choice of signal, the signal-to-noise ratio (SNR), and the interpolation method all affect the accuracy of the synchronization. The auto-correlation properties of Zadoff-Chu sequences make them good candidates for synchronization signals. The accuracy of the arrival-time estimation may be improved by increasing the length of the sequence or increasing the signal-to-noise ratio. At high SNRs, the accuracy may also be improved by using higher-order polynomial interpolation.