The nonlinear Fourier transform (NLFT) extends the classical Fourier transform by replacing addition with matrix multiplication. While the NLFT on
SU(1,1) has been widely studied, its
SU(2) variant has only recently attracted attention due to emerging applications in quantum signal processing (QSP) and quantum singular value transformation (QSVT). In this paper, we investigate the inverse NLFT on
SU(2) and establish the numerical stability of the layer stripping algorithm for the first time under suitable conditions. Furthermore, we develop a fast and numerically stable algorithm, called inverse nonlinear fast Fourier transform, for performing inverse NLFT with near-linear complexity. This algorithm is applicable to computing phase factors for both QSP and the generalized QSP (GQSP).
We define a nonlinear Fourier transform which maps sequences of contractive
n×n matrices to
SU(2n)-valued functions on the circle
T. We characterize the image of finitely supported sequences and square-summable sequences on the half-line, and construct an inverse for
SU(2n)-valued functions whose diagonal
n×n blocks are outer matrix functions. As an application, we relate this nonlinear Fourier transform with quantum signal processing over
U(2n) and multivariate quantum signal processing.
Quantum signal processing (QSP) and quantum singular value transformation (QSVT) are powerful techniques for the development of quantum procedures. They allow to derive circuits preparing desired polynomial transformations. Recent research [Alexis et al. 2024] showed that Non-Linear Fourier Analysis (NLFA) can be employed to numerically compute a QSP protocol, with provable stability. In this work we extend their result, showing that GQSP and the Non-Linear Fourier Transform over
SU(2) are the same object. This statement - proven by a simple argument - has a bunch of consequences: first, the Riemann-Hilbert-Weiss algorithm can be turned, with little modifications and no penalty in complexity, into a unified, provably stable algorithm for the computation of phase factors in any QSP variant, including GQSP. Secondly, we derive a uniqueness result for the existence of GQSP phase factors based on the bijectivity of the Non-Linear Fourier Transform. Furthermore, NLFA provides a complete theory of infinite generalized quantum signal processing, which characterizes the class of functions approximable by GQSP protocols.