Positive-instantaneous frequency and approximation

Abstract

Positive-instantaneous frequency representation for transient signals has always been a great concern due to its theoretical and practical importance, although the involved concept itself is paradoxical. The desire and practice of uniqueness of such frequency representation (decomposition) raise the related topics in approximation. During approximately the last two decades there has formulated a signal decomposition and reconstruction method rooted in harmonic and complex analysis giving rise to the desired signal representations. The method decomposes any signal into a few basic signals that possess positive instantaneous frequencies. The theory has profound relations to classical mathematics and can be generalized to signals defined in higher dimensional manifolds with vector and matrix values, and in particular, promotes kernel approximation for multi-variate functions. This article mainly serves as a survey. It also gives two important technical proofs of which one for a general convergence result (Theorem 3.4), and the other for necessity of multiple kernel (Lemma 3.7).

Expositorily, for a given real-valued signal f one can associate it with a Hardy space function F whose real part coincides with f. Such function F has the form F=f+ iHf, where H stands for the Hilbert transformation of the context. We develop fast converging expansions of F in orthogonal terms of the form

F= \sum _{k=1}^{\infty }{c}_k{B}_k

where B_k’s are also Hardy space functions but with the additional properties

B_k\left(t\right)={\rho }_k \left(t\right)e^{i{\theta }_k\left(t\right)},{\rho }_k\ge 0 ,{\theta }_k^{\text{'}} \left(t\right)\ge 0,a.. e

The original real-valued function f is accordingly expanded

f= \sum _{k=1}^{\infty }{\rho }_k\left(t\right)\cos {\theta }_k \left(t\right)

which, besides the properties of {\rho }_k and {\theta }_k given above, also satisfies

H( {\rho }_k\cos {\theta }_k\left(t\right)={\rho }_k \left(t\right)\sin {\theta }_k\left(t\right).

Real-valued functions f\left(t\right)=\rho \left(t\right)\cos \theta \left(t\right) that satisfy the condition

\rho \ge 0,\theta \text{'}\left(t\right)\ge 0,H\left(\rho \cos \theta \right) \left(t\right)= \rho \left(t\right)\sin \theta \left(t\right)

are called mono-components. If f is a mono-component, then the phase derivative \theta \text{'}\left(t\right)is defined to be instantaneous frequency of f. The above described positive-instantaneous frequency expansion is a generalization of the Fourier series expansion. Mono-components are crucial to understand the concept instantaneous frequency. We will present several most important mono-component function classes. Decompositions of signals into mono-components are called adaptive Fourier decompositions (AFDs). We note that some scopes of the studies on the 1D mono-components and AFDs can be extended to vector-valued or even matrix-valued signals defined on higher dimensional manifolds. We finally provide an account of related studies in pure and applied mathematics.

Keywords

Möbius transform / blaschke product / mono-component / hilbert transform / hardy space / inner and outer functions / adaptive fourier decomposition / rational orthogonal system / nevanlinna factorization / beurling-lax theorem / reproducing kernel hilbert space / several complex variables / clifford algebra / pre-orthogonal AFD

References

[1] AlpayD, Colombo F, QianT, SabadiniI. Adaptive orthonormal systems for matrix-valued functions. Proceedings of the American Mathematical Society, 2017, 145(5): 2089–2106

[2] AlpayD, Colombo F, QianT, SabadiniI. Adaptive Decomposition: The Case of the Drury-Arveson Space. Journal of Fourier Analysis and Applications, 2017, 23(6), 1426–1444

[3] AxelssonA, KouK I, QianT. Hilbert transforms and the Cauchy integral in Euclidean space. Studia Mathematica, 2009, 193(2): 161–187

[4] BaratchartL, Cardelli M, OliviM. Identification and rational L² approximation, a gradient algorithm. Automatica, 1991, 27: 413–418

[5] BaratchartL, DangP, QianT. Hardy-Hodge Decomposition of Vector Fields in Rⁿ. Transactions of the American Mathematical Society, 2018, 370: 2005–2022

[6] BaratchartL, MaiW X, QianT. Greedy Algorithms and Rational Approximation in One and Several Variables. In: Bernstein S., Kaehler U., Sabadini I., Sommen F. (eds) Modern Trends in Hypercomplex Analysis. Trends in Mathematics, 2016: 19–33

[7] BellS. The Cauchy Transform, Potential theory and Conformal Mappings. CRC Press, Boca, Raton, 1992

[8] BoashashB. Estimating and interpreting the instantaneous frequency of a signal-Part 1: Fundamentals. Proceedings of The IEEE, 1992, 80(4): 520–538

[9] BeltránJ R, de León P. Instantaneous frequency estimation and representation of the audio signal through Complex Wavelet Additive Synthesis. International Journal of Wavelets, Multiresolution and Information Processing, 2014, 12(03): 1450030

[10] ChengM T, DengG T. Lecture notes on harmonic analysis. Peking University, 1979

[11] CoifmanR, Peyriére J. Phase unwinding, or invariant subspace decompositions of Hardy spaces. Journal of Fourier Analysis and Applications, 2019, 25: 684–695

[12] CoifmanR, Steinerberger S. Nonlinear phase unwinding of functions. J Fourier Anal Appl, 2017, 23: 778–809

[13] CoifmanR, Steinerberger S, WuH T. Carrier frequencies, holomorphy and unwinding. SIAM J. Math. Anal., 2017, 49(6): 4838–4864

[14] ChengQ S. Digital Signal Processing. Peking University Press, 2003, in Chinese

[15] ChenQ H, QianT, TanL H. Constructive Proof of Beurling-Lax Theorem, Chin. Ann. of Math., 2015, 36: 141–146

[16] CohenL. Time-Frequency Analysis: Theory and Applications. Prentice Hall, 1995

[17] ChenQ H, MaiW X, ZhangL M, Mi W. System identification by discrete rational atoms. Automatica, 2015, 56: 53–59

[18] ColomboF, Sabadini I, SommenF. The Fueter primitive of bi-axially monogenic functions. Communications on Pure and Applied Analysis, 2014, 13(2): 657–672

[19] ColomboF, Sabadini I, SommenF. The Fueter mapping theorem in integral form and the ℱ-functional calculus. Mathematical Methods in the Applied Sciences, 2010, 33(17): 2050–2066

[20] DangP, DengG T, QianT. A Sharper Uncertainty principle. Journal of Functional Analysis, 2013, 265(10): 2239–2266

[21] DangP, DengG T, QianT. A Tighter Uncertainty Principle For Linear Canonical Transform in Terms of Phase Derivative. IEEE Transactions on Signal Processing, 2013, 61(21): 5153–5164

[22] DangP, LiuH, QianT. Hilbert Transformation and Representation of ax + b Group. Canadian Mathematical Bulletin, 2018, 61(1): 70–84

[23] DangP, LiuH, QianT. Hilbert Transformation and rSpin(n)+Rⁿ Group. arXiv:1711. 04519v1[math.CV], 2017

[24] DavisG, MalletS, AvellanedaM. Adaptive Greedy Approximations, Constr. Approxi., 1997, 13: 57–98

[25] DangP, MaiW X, QianT. Fourier Spectrum Characterizations of Clifford H^p Spaces on R+n+1 for 1 ≤ p ≤ ∞. Journal of Mathematical Analysis and Applications, 2020, 483: 123598

[26] DangP, QianT. Analytic Phase Derivatives, All-Pass Filters and Signals of Minimum Phase. IEEE Transactions on Signal Processing, 2011, 59(10): 4708–4718

[27] DangP, QianT. Transient Time-Frequency Distribution based on Mono-component Decompositions. International Journal of Wavelets, Multiresolution and Information Processing, 2013, 11(3): 1350022

[28] DangP, QianT, ChenQ H. Uncertainty Principle and Phase Amplitude Analysis of Signals on the Unit Sphere. Advances in Applied Clifford Algebras, 2017, 27(4): 2985–3013

[29] DangP, QianT, YouZ. Hardy-Sobolev spaces decomposition and applications in signal analysis. J. Fourier Anal. Appl., 2011. 17(1): 36–64

[30] DangP, QianT, YangY. Extra-strong uncertainty principles in relation to phase derivative for signals in Euclidean spaces. Journal of Mathematical Analysis and Applications, 2016, 437(2): 912–940

[31] DengG T, QianT. Rational approximation of Functions in Hardy Spaces. Complex Analysis and Operator Theory, 2016, 10(5): 903–920

[32] EisnerT, PapM. Discrete orthogonality of the Malmquist Takenaka system of the upper half plane and rational interpolation. Journal of Fourier Analysis and Applications, 2014, 20(1): 1–16

[33] FalcãoM I, Cruz J F, MalonekH R. Remarks on the generation of monogenic functions. International Conference on the Applications of Computer Science and Mathematics in Architecture and Civil Engineering, 17, Weimar, 2006

[34] FulcheriP, OliviP. Matrix rational H² approximation: a gradient algorithm based on schur analysis. SIAM I. Control Optim., 1998, 36(6): 2103–2127

[35] GaborD. Theory of communication. J. IEE., 1946, 93: 429–457

[36] GaudryG I, LongR, QianT. A Martingale proof of L²-boundednessof Clifford-Valued Singular Integrals. Annali di Mathematica Pura Ed Applicata, 1993, 165: 369–394

[37] GaudryG, QianT, WangS L, Boundedness of singular integrals with holomorphic kernels on star-shaped closed Lipschitz curves. Colloquium Mathematicum, 1996: 133–150

[38] GarnettJ B. Bounded Analyic Functions. Academic Press, 1981

[39] GomesN R. Compressive sensing in Clifford analysis. Doctoral Dissertation, Universidade de Aveiro (Portugal), 2015

[40] GongS. Private comminication, 2002

[41] GorusinG M. Geometrical Theory of Functions of One Complex Variable. translated by Jian-Gong Chen, 1956

[42] GantaP, ManuG, Anil SooramS. New Perspective for Health Monitoring System. International Journal of Ethics in Engineering and Management Education, 2016

[43] HummelJ A. Multivalent starlike function. J. d’ analyse Math., 1967, 18: 133–160

[44] HuangN E. The empirical mode decomposition and the Hilbert spectrum for nonlinear and non-stationary time series analysis. Proc. R. Soc. Lndon, 1998: 903–995

[45] KirkbasA, Kizilkaya A, BogarE. Optimal basis pursuit based on jaya optimization for adaptive fourier decomposition. Telecommunications and Signal Processing, 2017 40th International Conference on IEEE: 538–543

[46] KraussharR S, RyanJ. Clifford and harmonic analysis on cylinders and tori. Revista Matematica Iberoamericana, 2005, 21(1): 87–110

[47] LiH C, DengG T, QianT. Fourier Spectrum Characterizations of H^p Spaces on Tubes Over Cones for 1 ≦ p ≦ ∞. Complex Analysis and Operator Theory, 2018, 12: 1193–1218

[48] LiH C, DengG T, QianT. Hardy space decomposition of on the unit circle: 0¡p¡1. Complex Variables and Elliptic Equations: An International Journal, 2016, 61(4): 510–523

[49] LeiY, FangY, ZhangL M. Iterative learning control for discrete linear system with wireless transmission based on adaptive fourier decomposition. Control Conference (CCC), 2017 36th Chinese IEEE

[50] LiangY, JiaL M, CaiG. A new approach to diagnose rolling bearing faults based on AFD. Proceedings of the 2013 International Conference on Electrical and Information Technologies for Rail Transportation-Volume II, Springer

[51] LiC, McIntosh A, QianT. Clifford algebras, Fourier transforms, and singular Convolution operators on Lipschitz surfaces. Revista Matematica Iberoamericana, 1994, 10(3): 665–695

[52] LiC, McIntosh A, SemmesS. Convolution Singular Integrals on Lipschitz Surfaces. Journal of the American Mathematical Society, 1992: 455–481

[53] LiS, QianT, MaiW X. Sparse Reconstruction of Hardy Signal And Applications to Time-Frequency Distribution. International Journal of Wavelets, Multiresolution and Information Processing, 2013

[54] LyzzaikA. On a conjecture of M.S. Robertson. Proc. Am. Math. Soc., 1984, 91: 108–210

[55] McIntoshA, QianT. Convolution singular integrals on Lipschitz curves. Springer-Verlag, Lecture Notes in Maths, 1991, 1494: 142–162

[56] McIntoshA, QianT. L^p Fourier multipliers along Lipschitz curves. Transactions of The American Mathematical Society, 1992, 333(1): 157–176

[57] MashreghiJ, Fricain E. Blaschke products and their applications. Springer, 2013

[58] MaiW X, QianT, SaitohS. Adaptive Decomposition of Functions with Reproducing Kernels. in preparation

[59] MiW, QianT. Frequency Domain Identification: An Algorithm Based On Adaptive Rational Orthogonal System. Automatica, 2012, 48(6): 1154–1162

[60] MoY, QianT, MiW. Sparse Representation in Szego Kernels through Reproducing Kernel Hilbert Space Theory with Applications. International Journal of Wavelet. Multiresolution and Information Processing, 2015, 13(4): 1550030

[61] MiW, QianT, WanF. A Fast Adaptive Model Reduction Method Based on Takenaka-Malmquist Systems. Systems and Control Letters, 2012, 61(1): 223–230

[62] MozesF E, SzalaiJ. Computing the instantaneous frequency for an ECG signal. Scientific Bulletin of the “Petru Maior” University of Targu Mures, 2012, 9(2): 28

[63] NahonM. Phase Evaluation and Segmentation. Ph.D. Thesis, Yale University, 2000

[64] PerottiA. Directional quaternionic Hilbert operators. Hypercomplex analysis, Birkhüser Basel, 2008: 235–258

[65] PicinbonoB. On instantaneous amplitude and phase of signals. IEEE Transactions on Signal Processing, 1997, 45(3): 552–560

[66] QianT. Singular integrals with holomorphic kernels and Fourier multipliers on starshape Lipschitz curves. Studia Mathematica, 1997, 123(3): 195–216

[67] QianT. Characterization of boundary values of functions in Hardy spaces with applications in signal analysis. Journal of Integral Equations and Applications, 2005, 17(2): 159–198

[68] QianT. Analytic Signals and Harmonic Measures. Journal of Mathematical Analysis and Applications, 2006, 314(2): 526–536

[69] QianT. Mono-components for decomposition of signals. Mathematical Methods in the Applied Sciences, 2006, 29(10): 1187–1198

[70] QianT. Boundary Derivatives of the Phases of Inner and Outer Functions and Applications. Mathematical Methods in the Applied Sciences, 2009, 32: 253–263

[71] QianT. Intrinsic mono-component decomposition of functions: An advance of Fourier theory. Mathematical Methods in Applied Sciences, 2010, 33: 880–891

[72] QianT. Two-Dimensional Adaptive Fourier Decomposition. Mathematical Methods in the Applied Sciences, 2016, 39(10): 2431–2448

[73] QianT. Adaptive Fourier Decomposition: A Mathematical Method Through Complex Analysis. Harmonic Analysis and Signal Analysis, Science Press (in Chinese), 2015

[74] QianT, LiP T. Singular Integrals and Fourier Theory. Science Press (in Chinese), 2017

[75] QianT. Fourier analysis on starlike Lipschitz surfaces. Journal of Functional Analysis, 2001, 183: 370–412

[76] QianT. Cyclic AFD Algorithm for Best Approximation by Rational Functions of Given Order. Mathematical Methods in the Applied Sciences, 2014, 37(6): 846–859

[77] QianT, ChenQ H, TanL H. Rational Orthogonal Systems are Schauder Bases. Complex Variables and Elliptic Equations, 2014, 59(6): 841–846

[78] QianT, ChenQ H, LiL Q. Analytic unit quadrature signals with non-linear phase. Physica D: Nonlinear Phenomena, 2005, 303: 80–87

[79] QianT, HuangJ S. AFD on the n-Torus. in preparation

[80] QianT, HoI T, LeongI T, Wang Y B. Adaptive decomposition of functions into pieces of non-negative instantaneous frequencies. International Journal of Wavelets, Multiresolution and Information Processing, 2010, 8(5): 813–833

[81] QianT, LiH, StessinM. Comparison of Adaptive Mono-component Decompositions. Nonlinear Analysis: Real World Applications, 2013, 14(2): 1055–1074

[82] QianT, TanL H. Characterizations of Mono-components: the Blaschke and Starlike types. Complex Analysis and Operator Theory, 2015: 1–17

[83] QianT, TanL H. Backward shift invariant subspaces with applications to band preserving and phase retrieval problems. Mathematical Methods in the Applied Sciences, 2016, 39(6): 1591–1598

[84] QianT, WangY B. Adaptive Fourier Series-A Variation of Greedy Algorithm. Advances in Computational Mathematics, 2011, 34(3): 279–293

[85] QianT, WegertE. Optimal Approximation by Blaschke Forms. Complex Variables and Elliptic Equations, 2013, 58(1): 123–133

[86] QianT, SproessigW,Wang J X. Adaptive Fourier decomposition of functions in quaternionic Hardy spaces. Mathematical Methods in the Applied Sciences, 2012, 35(1): 43–64

[87] QianT, TanL H, WangY B. Adaptive Decomposition by Weighted Inner Functions: A Generalization of Fourier Serie. J. Fourier Anal. Appl., 2011, 17(2): 175–190

[88] QianT, WangJ X. Adaptive Decomposition of Functions by Higher Order Szegö Kernels I: A Method for Mono-component Decomposition. submitted to Acta Applicanda Mathematicae

[89] QianT, WangJ Z. Gradient Descent Method for Best Blaschke-Form Approximation of Function in Hardy Space. http://arxiv.org/abs/1803.08422

[90] QianT, SproessigW,Wang J X. Adaptive Fourier decomposition of functions in quaternionic Hardy spaces. Mathematical Methods in the Applied Sciences, 2012, 35: 43–64

[91] QianT, WangJ X, YangY. Matching Pursuits among Shifted Cauchy Kernels in Higher-Dimensional Spaces. Acta Mathematica Scientia, 2014, 34(3): 660–672

[92] QianT, WangR, XuY S, Zhang H Z. Orthonormal Bases with Nonlinear Phase. Advances in Computational Mathematics, 2010, 33: 75–95

[93] QianT, XuY S, YanD Y, Yan L X, YuB. Fourier Spectrum Characterization of Hardy Spaces and Applications. Proceedings of the American Mathematical Society, 2009, 137(3): 971–980

[94] QianT, YangY. Hilbert Transforms on the Sphere With the Clifford Algebra Setting. Journal of Fourier Analysis and Applications, 2019, 15: 753–774

[95] QianT. ZhangL M, LiZ X. Algorithm of Adaptive Fourier Decomposition. IEEE Transaction on Signal Processing, Dec., 2011, 59(12): 5899–5902

[96] QuW, DangP. Rational Approximation in a Class of Weighted Hardy Spaces. Complex Analysis and Operator Theory volume, 2019, 13: 1827–1852

[97] SalomonL. Analyse de l’anisotropie dans des images texturées, 2016

[98] SakaguchiF, Hayashi M. General theory for integer-type algorithm for higher order differential equations. Numerical Functional Analysis and Optimization, 2011, 32(5): 541–582

[99] SakaguchiF, Hayashi M. Differentiability of eigenfunctions of the closures of differential operators with rational coefficient functions, arXiv:0903.4852, 2009

[100] SakaguchiF, Hayashi M. Practical implementation and error bound of integer-type algorithm for higher-order differential equations. Numerical Functional Analysis and Optimization, 2011, 32(12): 1316–1364

[101] SakaguchiF, Hayashi M. Integer-type algorithm for higher order differential equations by smooth wavepackets. arXiv:0903.4848, 2009

[102] SchepperD, QianT, SommenF, Wang J X. Holomorphic Approximation of L₂-functions on the Unit Sphere in R3. Journal of Mathematical Analysis and Applications, 2014, 416(2): 659–671

[103] SharpleyR C, Vatchev V. Analysis of intrinsic mode functions. Constructive Approximation, 2006, 24: 17–47

[104] SteinE M, WeissG. Introduction to Fourirer Analysis on Euclidean Spaces. Princeton University Press, Princeton, New Jersey, 1971

[105] TanL H, ShenL X, YangL H. Rational orthogonal bases satisfying the Bedrosian Identity. Advances in Computational Mathematics, 2010, 33: 285–303

[106] TanL H, YangL H, HuangD R. The structure of instantaneous frequencies of periodic analytic signals. Sci. China Math., 2010, 53(2): 347–355

[107] TanL H, QianT. Backward Shift Invariant Subspaces With Applications to Band Preserving and Phase Retrieval Problems

[108] TanL H, QianT. Extracting Outer Function Part from Hardy Space Function. Science China Mathematics, 2017, 60(11): 2321–2336

[109] TanL H, QianT, ChenQ H. New aspects of Beurling Lax shift invariant subspaces. Applied Mathematics and Computation, 2015: 257–266

[110] VatchevV. A class of intrinsic trigonometric mode polynomials. International Conference Approximation Theory. Springer, Cham, 2016: 361–373

[111] VlietD V. Analytic signals with non-negative instantaneous frequency. Journal of Integral Equations and Applications, 2009, 21: 95–111

[112] WalshJ L. Interpolation and Approximation by Rational Functions in the Complex Plane. American Mathematical Society: Providence, RI, 1969

[113] WangS L. Simple Proofs of the Bedrosian Equality for the Hilbert Transform. Science in China, Series A: Mathematics, 2009, 52(3): 507–510

[114] WangJ X, QianT. Approximation of monogenic functions by higher order Szegö kernels on the unit ball and the upper half space. Sciences in China: Mathematics, 2014, 57(9): 1785–1797

[115] WeissG, WeissM. A derivation of the main results of the theory of Hp-spaces. Rev. Un. Mat. Argentina, 1962, 20: 63–71

[116] WuW. Applications in Digital Image Processing of Octonions Analysis and the Qian Method. South China Normal University, 2014

[117] WangZ, da Cruz J N, WanF. Adaptive Fourier decomposition approach for lung-heart sound separation. Computational Intelligence and Virtual Environments for Measurement Systems and Applications (CIVEMSA). 2015 IEEE International Conference on. IEEE, 2015

[118] WuM Z, WangY, LiX M. Fast Algorithm of The Qian Method in Digital Watermarking. Computer Engineering and Desining, 2016

[119] WuM Z, WangY, LiX M. Improvement of 2D Qian Method and its Application in Image Denoising. South China Normal University, 2016

[120] XuY S. Private comminication, 2005

[121] YangY, QianT, SommenF. Phase Derivative of Monogenic Signals in Higher Dimensional Spaces. Complex Analysis and Operator Theory, 2012, 6(5): 987–1010

[122] YuB, ZhangH Z. The Bedrosian Identity and Homogeneous Semi-convolution Equations. Journal of Integral Equations and Applications, 2008, 20: 527–568

[123] ZhangL. A New Time-Frequency Speech Analysis Approach Based On Adaptive Fourier Decomposition. World Academy of Science, Engineering and Technology, International Journal of Electrical, Computer, Energetic, Electronic and Communication Engineering, 2013

[124] ZhangL M, LiuN, YuP. A novel instantaneous frequency algorithm and its application in stock index movement prediction. IEEE Journal of Selected Topics in Signal Processing, 2012, 6(4): 311–318

[125] ZhangL M, QianT, MaiW X, Dang P. Adaptive Fourier decomposition-based Dirac type time-frequency distribution. Mathematical Methods in the Applied Sciences, 2017, 40(8): 2815–2833

RIGHTS & PERMISSIONS

2022 Higher Education Press