Multi-Period-Clustering Algorithm for Carrier Doppler-Shift Acquisition in Space Communications

Introducing the clustering idea in machine learning, a multi-period-clustering (MPC) algorithm is proposed to address the energy dispersion problem for the carrier Doppler-shift acquisition in space communications. The MPC algorithm first selects some large elements, then conducts the density-clustering on them, finally searches the largest cluster as the acquisition result. By clustering multiple signal elements into one larger signal cluster, the MPC algorithm re-accumulates the dispersed energy. Simulation results demonstrate that the proposed algorithm achieves a higher acquisition probability than the FFT, STFT and SRC algorithms and consumes a lower computational complexity than the ES and FRFT algorithms.