TY - GEN
T1 - Multi-Unimodular Waveform Design With Low Peak Sidelobe Level Via Direct Phase Optimizations
AU - Zhao, Xiaohan
AU - Li, Yongzhe
AU - Tao, Ran
N1 - Publisher Copyright:
© 2023 European Signal Processing Conference, EUSIPCO. All rights reserved.
PY - 2023
Y1 - 2023
N2 - In this paper, we propose an efficient algorithm for designing multi-unimodular waveforms with low peak sidelobe level (PSL) toward any time lags of interest, which differs from existing approaches. The generic PSL metric defined on the TLOI is formulated for minimization, and the phase values of waveforms are directly optimized. To be specific, we first convert the PSL-minimization based waveform design into an lp-norm based minimization problem. Then, we reformulate it into an unconstrained optimization problem with respect to the phase values of waveform elements, wherein the inherent waveform property of constant envelopes and a discrete Fourier transform matrix are both used. Our remaining contributions lie in deriving the complex gradient of the unconstrained objective and also elaborating its majorant via a Lipschitz-constant related quantity which is properly designed. A closed-form solution that updates the phase values through iterations is obtained by the majorization-minimization framework, and it boils down to a gradient descent regime. The fast implementation of our algorithm is also provided, whose performance improvements are verified.
AB - In this paper, we propose an efficient algorithm for designing multi-unimodular waveforms with low peak sidelobe level (PSL) toward any time lags of interest, which differs from existing approaches. The generic PSL metric defined on the TLOI is formulated for minimization, and the phase values of waveforms are directly optimized. To be specific, we first convert the PSL-minimization based waveform design into an lp-norm based minimization problem. Then, we reformulate it into an unconstrained optimization problem with respect to the phase values of waveform elements, wherein the inherent waveform property of constant envelopes and a discrete Fourier transform matrix are both used. Our remaining contributions lie in deriving the complex gradient of the unconstrained objective and also elaborating its majorant via a Lipschitz-constant related quantity which is properly designed. A closed-form solution that updates the phase values through iterations is obtained by the majorization-minimization framework, and it boils down to a gradient descent regime. The fast implementation of our algorithm is also provided, whose performance improvements are verified.
KW - Multi-waveform design
KW - peak sidelobe level (PSL)
KW - phase optimizations
KW - time lags of interest (TLOI)
UR - https://www.scopus.com/pages/publications/85178373445
U2 - 10.23919/EUSIPCO58844.2023.10289833
DO - 10.23919/EUSIPCO58844.2023.10289833
M3 - Conference contribution
AN - SCOPUS:85178373445
T3 - European Signal Processing Conference
SP - 1579
EP - 1583
BT - 31st European Signal Processing Conference, EUSIPCO 2023 - Proceedings
PB - European Signal Processing Conference, EUSIPCO
T2 - 31st European Signal Processing Conference, EUSIPCO 2023
Y2 - 4 September 2023 through 8 September 2023
ER -