An Adaptive Hierarchical Search Algorithm for Millimetre-Wave Beam Alignment

We present an adaptive hierarchical search (AHS) algorithm for millimetre-wave beam-alignment. AHS owes its adaptivity to the use of Generalised Likelihood Ratio Test that iteratively tests if the Signal to Noise Ratio (SNR) of the accumulated training signal meets a target threshold before stopping beam searching. With the adaptivity, AHS mitigates the problem of error propagation of non-adaptive hierarchical search by automatically spending longer search time at layers with wider beams and in low SNRs. A new method for refining the search interval is also developed to reduce the search time and improve the beam search accuracy. Simulation results verify the effectiveness of the proposed algorithm.