Concept and theoretical performance analysis for decentralised digital synchronisation in distributed radar sensor networks

Abstract

This paper presents a decentralised technique for achieving frequency, time, and phase synchronisation of platforms cooperating in a distributed radar sensor network. The proposed method is advantageous for existing digital radar systems as it can be implemented entirely in the software without the use of additional RF hardware required by other techniques. The synchronisation signal model for signals transmitted and received in various clock domains is presented and an estimation model is subsequently derived for estimating and correcting the clock drifts. A modified version of a previously developed phase and clock bias correction procedure is outlined for correcting time and phase after frequency synchronisation. A comprehensive theoretical performance analysis of the technique is performed in which the expected maximum achievable performance is derived in terms of the Cramér–Rao lower bound for frequency, time, and phase measurements. Multiple Monte Carlo simulations show that the proposed technique approaches this performance limit. Finally, a simulated distributed transmit beamforming scenario is provided to show the application of the proposed technique in a practical system architecture. The results of this show that as the signal-to-noise ratio approaches moderate levels, the proposed synchronisation technique enables the beamforming network to achieve nearly optimal coherent energy gain at the beamforming destination.