Fourier Series/Transform

Hafiz Saad Khaliq(11TC03)1, Mohammad Hunain Khan(11TC41)2, Syed Zulfiqar Haider Bukhari(11TC51)3, Syed Noor-ul-Hassan Bukhari(11TC63)4, Ahmad Faheem Alam(11TC10)5,Javed Iqbal(11TC20)6
Department of Telecommunication Engineering, University College of Engineering & Technology, The Islamia University of Bahawalpur, Pakistan.
engr.saadkhaliq@gmail.com1, engr.hunain@gmail.com2 , engr.szhb@gmail.com3, syednoorulhassanbukhari@gmail.com4, Faheemalam2012@gmail.com5, engr.javediqbal4u@gmail.com6

Abstract This paper Fourier Series / Transform looks at one of the jewels in science and engineering. Fourier Series / Transform have lot of applications in new technology. The Fourier Series, the founding principle behind the field of Fourier Analysis, is an infinite expansion of a periodic function/Signals into the sum of simple oscillating functions, namely sine and cosines (or complex exponentials). In physics and engineering, decomposition of functions in terms of sine and cosine is useful because it allows to manipulate functions from one to more easily that are, for example, discontinuous or simply difficult to represent analytically. In particular, the Fourier Series heavily used in the field of electronics, quantum mechnics, and electrodynamics. Additionally, many methods based on the Fourier Series, such as the FFT (Fast Fourier Transform a form of a Descrete Fourier Transform [DFT]), are heavily used in the fields of Digital Signal Processing (DSP) and Spectral Analysis. In this paper, we will briefly surveys Fourier Series/Transform applications and looks at possible ways of enhancing the performance of Fast Fourier Transform. Keywords: Descrete Fourier Transform (DFT), Orthogonal , Fast Fourier Transform (FFT), Schwarz Reflection Principle (SRP), Conformal Map.

Fourier transform is a relationship between a signal in the time domain and its representation in the frequency domain. Actually, Fourier transform based on Fourier series. Study of Fourier series provides motivation for the Fourier transform. Fourier transform is a mathematical operation that breaks a signal into its constituent frequencies. The original signal that changed over time is called the time domain representation and the Fourier transform is called the frequency domain representation of a signal since it depends on frequency. The frequency domain representation and the process used to transform a signal into the frequency domain are referred to as the Fourier transform. Fourier transform has some basic properties such as translation, linearity, modulation, conjugation, duality, scaling and convolution. Fourier transform is closely related to Laplace transformation because it is used in solving differential equations. Fourier transform have applications in nuclear magnetic resonance (NMR) and in other kind of spectroscopy.[3] II. MATHEMATICAL FOUNDATIONS The usual Fourier series involving the sums of sines and cosine is obtained by taking and Since these functions form a complete orthogonal system over , the Fourier series of a function f(x) is given by

I. INTRODUCTION he Fourier series / transforms are the most important discoveries in mathematics. It fulfills all aspects of science and engineering, from the most environs of modern physics such as symmetry / quantum mechanics to applied areas such as crystallography, holography, imaging, in signal processing areas covering compression, modulation, multiplexing etc. Variants of the same are used in audio and video signals such as JPEG, MPEG etc[1]. Fourier series is an expansion of a periodic function into a sum of sines and cosines with different amplitude and frequencies. Fourier series is a branch of Fourier analysis and it was introduced by Joseph Fourier[2]. Fourier series was initially developed for the solvation of heat equations but later it was found out that it can be used to solve large set of mathematical problems specially the problems that involve linear differential equations with constant coefficients. Today, Fourier series has applications in many fields including vibration analysis, optics, signal processing, electrical engineering, image processing, econometrics, acoustics and quantum mechanics[3].

Where

Chebyshev, Bessel etc. Transforming differential equations to algebraic equations thus enabling algorithmic implementations is one of the major applications of Fourier transform. The topic of harmonic functions with applications and deep roots in complex analysis has its beginnings in Fourier Analysis. A direct application of the above in physics is in the theory of Diffraction where the far field diffraction pattern is the spatial Fourier transform of the unit cell. Another direction of generalization has been through the idea of reciprocal space / phase space etc in quantum mechanics and modern physics where a joint time / space-frequency space is used for analysis. This idea leads to the Heisenberg Uncertainty Principle. The idea of periodicity is extended to more general symmetry groups as well as more generalized higher dimensional spaces in the theory of automorphic functions, representation theory, modular forms, harmonic groups etc where one want to quantify the symmetry operations to which an object is invariant. Other application is in the direction of adaptive data driven basis functions extracted through algorithms such as Singular Value Decomposition etc. The links between modulation and convolution as transform pairs is a key out of the major properties of the Fourier transform[1]. III. DISCRETE FOURIER TRANSFORM The continuous Fourier transform is defined as For a periodic function in , these becomes

Figure 1. Fourier Series Examples

and n = 1,2,3,. Note that the coefficients of the constant term ao has been written in a special form compared to the general form for a generalized Fourier series in order to preserve symmetry with the definitions of an and bn[4]. The series is defined in imaginary exponential form as:

where An is given by the expression

Now consider generalization in the case of a discrete function, by letting , where , with
, ..., transform as: . Writing this out gives the discrete Fourier

If the frequency components of the sine and cosine terms are continuous, the concept of the Fourier Transform must be introduced. The imaginary exponential form of Fourier Transform is defined as:

Then the inverse Fourier transform is

Discrete Fourier transforms are very useful because they reveal periodicities in input data as well as the relative strengths of any periodic components[5]. If the coefficients can be thought of as representing a vector, then polynomial multiplication can be looked at as a convolution. Since modulation / multiplication in the original time / space domain corresponds to convolution in the Fourier domain, if the transform and the inverse can be computed quickly, then we could simplify multiplication substantially by

The Fourier series is one of the primary examples of orthogonal expansions / decompositions which has been generalized / extended to other families such as Legendre,

multiplying component wise. The real advantage become important when many such polynomials have to be multiplied together, for example involving matrix of functions in case of computations etc. A really interesting possibility is that of inverting the above to factorize polynomials, i.e. , given a polynomial, we compute its transform and then factorize each of the components and then try to invert the transform. One problem is that the factors seem to get mixed up but a hypothesis that would be interesting to investigate is that these could possibly correspond to symmetries possessed by the polynomial and / or possibly unique prime factors possibly ensuring reconstructability. The other approach of course is to find the norm and height of the original polynomial and incorporate these in the factoring so that it becomes a constrained optimization problem. One problem is that the coefficients could be complex, it is necessary to use the group theory and connecting it to such discrete structures such as Gaussian primes etc. An alternative approach is to expand it in terms of other family of orthogonal functions such as Legendre Polynomials etc[1]. IV. GOING A STEP FURTHER, SCHWARZ REFLECTION PRINCIPLE AND APPLICATIONS The Schwarz reflection principle in Complex Analysis is an attempt to extend the function analytically. One of the ways is through reflection across the real axis[6],

definition of an analytic region is divided into domains / regions which are defined by singularities. Since 1/f(z) type of inversion are typically related to Fourier transforms / reciprocal space, they could find applications in the theory of diffraction. It may also be useful in the case of evaluation / asymptotic approximation of very complex functions such as the partition function[1]. A related idea is the link between an infinite series and its associated limit integral (passing from Fourier series to transform) by evaluating the function at roots of unity related to incommensurate multiples of the arc-length[8]. V. CONCLUSION In future, there is a possibility of implementing the Fourier transform in the field of quantum / optical computing etc[10]. talks about the possibility of applying Fourier methods to synthesis of mechanisms which act as function / path generators[9]. There is a possibility of extending these methods to navigation in abstract higher dimensional spaces including mathematical logic, information compression, molecular logical circuits, molecular synthesis etc[10]. ACKNOWLEDGMENT We wish to thank our respected teacher Engr. Abdul Rehman Chishti and friends for their excellent contribution and support for the completion of this paper. Specially we are very much thankful to Dr. Mohammad Mukhtar the Vice Chancellor of the university, and Professor Jan Mohammad Keerio the Principal of UCET, for their efforts to facilitate and provide us most suitable environment for research and development. REFERENCES
[1] Sundaram Ramchandaran, Fourier Series / transform ,Old wine, New flavors , International Journal of Innovative Research in Computer and Communication Engineering , (www.ijircce.com) , Vol. 1, Issue 2 , April 2013 . http://en.wikipedia.org/wiki/Fourier_series http://www.differencebetween.com/difference-between-fourier-seriesand-vs-fourier-transform/ Weisstein, Eric W. "Fourier Series." From MathWorld--A Wolfram Web Resource http://mathworld.wolfram.com/FourierSeries.html Weisstein, Eric W. "Discrete Fourier Transform." From MathWorld A Wolfram Web Resource. http://mathworld.wolfram.com/DiscreteFourierTransform.html Lars Ahlfors , 'Complex Analysis'. Rowland, Todd. "Schwarz Reflection Principle." From MathWorld--A Wolfram Web Resource created by Eric W. Weisstein http://mathworld.wolfram.com/SchwarzReflectionPrinciple.html Walter Rudin, 'Principles of Mathematical Analysis', Third Edition, 1976, pp :200. Sundaram Ramchandran ,'Philosophical Ramblings on Some Mathematical Dichotomies and the Concept of Scale' , International Journal of Scientific Engineering and Technology , (www.ijset.com), Volume No.1, Issue No.4, pp : 28-34 ,Oct 2012. Xichun Nie, Venkat Krovi, 'Fourier Methods for Kinematic Synthesis of Coupled Serial Chain Mechanisms', Transactions of the ASME,232 , Vol. 127, MARCH 2005.

and the values for z reflected across the real axis are the reflection of across the real axis.

[2] [3] [4] Figure 2. Schwarz Reflection Principle

This is called the Schwarz reflection principle, and is sometimes also known as Schwarzs symmetric principle[7]. The above figure shows the reflection principle applied to a function defined for the upper half-disk (left figure; red) and its image (right figure; red). The function is real on the real axis, so it is possible to extend the function to the reflected domain (left and right figures; pink). One could analytically continue the function along the nth roots of unity forming the vertices of a n-sided regular polygon till one reaches a singularity of the function. This is linked to the idea / hypothesis that the function behaves like a Conformal Map. This is similar to the idea that the domain of

[5]

[6] [7]

[8] [9]

[10]