Open AccessMinimum mean squares error beamforming with cyclic sub‐vector optimisation
Author(s) -
Liu Tuanning,
Zhou Yuanping,
Ma Yaoting,
Miao Rongzhen
Publication year - 2020
Publication title -
iet signal processing
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.384
H-Index - 42
ISSN - 1751-9683
DOI - 10.1049/iet-spr.2019.0164
Subject(s) - algorithm , computational complexity theory , diagonal , beamforming , rate of convergence , conjugate gradient method , matrix (chemical analysis) , convergence (economics) , dimension (graph theory) , adaptive beamformer , mathematics , computer science , block matrix , coordinate descent , mathematical optimization , eigenvalues and eigenvectors , key (lock) , telecommunications , materials science , geometry , physics , quantum mechanics , pure mathematics , economics , composite material , economic growth , computer security
A cyclic sub‐vector optimisation (CSVO) beamforming approach is investigated. With the proposed algorithm, an array beamforming vector is partitioned into a number of sub‐vectors of small sizes, allowing reduced‐dimension processing. Then multiple optimisation cycles are carried out by the block coordinate descent method, which leads to an optimal beamforming vector. The proposed scheme still needs to compute the matrix inversion, but the size of the matrix can be flexibly chosen and the computational complexity is manageable. The proof of the convergence and complexity analysis is given. The simulation results demonstrate the effectiveness and fine features of the proposed algorithm. Although the convergence rate of the CSVO is slightly slower, the CSVO has lower computational complexity than that of the diagonal loading conjugate gradient applied to normal equations algorithm. In comparison with other sub‐vector approaches, the proposed algorithm gains a faster convergence rate and improved stability.