write one pages outline from the following essay

Running Head: Fourier Transform: Time-Frequency Analysis. 1

Fourier Transform: Time-Frequency Analysis. 13

Fourier Transform: Time-Frequency Analysis.

Student’s Name

University Affiliation

Fourier Transform: Time-Frequency Analysis.

Fourier transform articulates a function of time in terms of the amplitude and phase of every of the frequencies that build it up. This is just like the approach in which a musical chord can be expressed because the amplitude (or loudness) of the notes that build it up. The ensuing function, a (complex) amplitude that depends on frequency, is termed the frequency domain illustration of the natural phenomenon modelled by the initial function. The term Fourier transform refers each to the operation that associates to a function its frequency domain illustration, and to the frequency domain illustration itself.

For many functions of sensible interest, there's an inverse Fourier transform, thus it's attainable to recover the initial function of time from its Fourier transform. The quality case of this is often the Gaussian perform, of considerable importance in applied math and statistics likewise as within the study of physical phenomena exhibiting distribution (e.g., diffusion). With applicable normalizations, the Gaussian goes to itself below the Fourier remodel. Joseph Fourier introduced the remodel in his study of heat transfer, wherever Gaussian functions seem as solutions of the heat equation.

When functions are recoverable from their Fourier transforms, linear operations performed in one domain (time or frequency) have corresponding operations within the different domain, which are generally easier to perform. The operation of differentiation within the time domain corresponds to multiplication by the frequency, thus some differential equations are easier to research within the frequency domain. Also, convolution within the time domain corresponds to normal multiplication within the frequency domain. Concretely, this implies that any linear time-invariant system, like associate electronic filter applied to a signal, may be expressed comparatively merely as an operation on frequencies. thus vital simplification is usually achieved by remodeling time functions to the frequency domain, playacting the specified operations, and remodeling the result back to time. Fourier analysis is the systematic study of the connection between the frequency and time domains, as well as the types of functions or operations that are "simpler" in one or the other, and has deep connections to the majority areas of recent arithmetic.

The Fourier transform may be formally outlined as an (improper) Riemann integral, creating it an integral remodel, though that definition isn't appropriate for several applications requiring a a lot of subtle integration theory.[note 4] It may also be generalized to functions on Euclidean space, causing a function of area (a scalar field) to a function of momentum. this concept makes the abstraction Fourier transform terribly natural within the study of waves, furthermore as in quantum physics, wherever it's necessary to be able to represent wave solutions either as functions of space or as functions of momentum (or both, as seems within the description of wave fronts). Still additional generalization is feasible to functions on groups that notably includes the distinct Fourier transform and circular Fourier remodel (that is, Fourier series).

The Fourier transform is key in mathematics, physical sciences and engineering.  Its discrete equivalent, the Discrete Fourier Transform (DFT) that is typically computed by means of the Fast Fourier Transform (FFT), has transformed the modern society, as it is everywhere in signal processing and digital electronics.  Radio astronomers are principally keen users of the Fourier transforms for the reason that Fourier transforms are crucial constituents in data processing, for instance, periodicity searches, and instruments, e.g., receivers, antennas, spectrometers, and they are the corner stores of interferometry and aperture synthesis.

The Fourier transform is a rescindable, linear transform with numerous essential properties.  For any function f(x), the Fourier transform may be denoted as F(s), where the outcome of x and s is non-dimensional.  Often x is a function of time t (that is, the time-domain signal) and s relates to the inverse time/frequency (i.e., frequency-domain signal).  

The Fourier transform is defined by the equation   which is typically referred to as the forward transform, and 

which is the inverse transform. 

In the two cases, i −1.

Different definitions of the Fourier transform are founded on angular frequency (=2 ), have different standardizations, or the reverse sign convention in the multifaceted exponential.  As the Fourier transformation is rescindable, the symmetric symbol, , is regularly used to indicate "is the Fourier transform of"; e.g., F(s) f(x).

The complex exponential is the core of the whole transform.  It is simply a complex number where mutually the real and imaginary portions are sinusoids.  The precise relation is known as Euler's formula 

which results in the famous identity ei +1=0 that relates 5 of the most necessary numbers in arithmetic. Complex exponentials are abundant easier to govern than trigonometric functions, and that they give a compact notation for managing sinusoids of arbitrary phase, that form the idea of the Fourier transform.

Complex exponentials (or sines and cosines) are periodic functions, and therefore the set of complex exponentials is complete and orthogonal. So the Fourier transform will represent any piecewise continuous can and minimizes the least-square error between the operate and its representation. There exist different complete and orthogonal sets of periodic functions; for instance,Walsh functions (square waves) are helpful for digital electronics. Why do we perpetually encounter complex exponentials once solving physical problems? Why are monochromatic waves curving, and not periodic trains of square waves or triangular waves? The rationale is that the derivatives of complex exponentials are simply rescaled complex exponentials. In alternative words, the complex exponentials are the Eigen functions of the differential operator. Most physical systems conform linear differential equations. so an analog electronic filter can convert a sine wave into another sine wave having constant frequency (but not essentially an equivalent amplitude and phase), while a filtered square wave won't be a square wave. This property of advanced exponentials makes the Fourier transform uniquely helpful in fields starting from radio propagation to quantum physics.

Time-Frequency Analysis.

In signal processing, time–frequency analysis includes those techniques that study a signal in both the time and frequency domains at the same time, using varied time–frequency representations. instead of viewing a 1-dimensional signal (a operate, real or complex-valued, whose domain is the real line) and a few transform (another function whose domain is the real line, obtained from the first via some transform), time–frequency analysis studies a two-dimensional signal – a function whose domain is the two-dimensional real plane, obtained from the signal via a time–frequency remodel.

The mathematical motivation for this study is that functions and their remodel representation are typically tightly connected, and that they can be understood better by learning them collectively, as a two-dimensional object, instead of individually. an easy example is that the four-fold cyclicity of the Fourier transform – and also the indisputable fact that two-fold Fourier rework reverses direction – may be taken by considering the Fourier rework as a 90° rotation in the associated time–frequency plane: 4 such rotations yield the identity, and a pair of such rotation merely reverse direction (reflection through the origin).

The practical motivation for time–frequency analysis is that classical harmonic analysis assumes that signals ar infinite in time or periodic, whereas several signals in practice are of short period, and alter substantially over their period. As an example, ancient musical instruments don't produce infinite period sinusoids, however instead begin with an attack, then bit by bit decay. This is often poorly delineated by ancient ways that motivates time–frequency analysis.

One of the most basic types of time–frequency analysis is the short-time Fourier transform (STFT), but a lot of subtle techniques have been developed, notably wavelets.

Principles of the Time-Frequency Analysis

There are many alternative ways to formulate a sound time–frequency distribution function, leading to many well-known time–frequency distributions, such as:

i. Short-time Fourier transform (including the Dennis Gabor transform),

ii. Wavelet transform,

iii. Bilinear time–frequency distribution function (Wigner distribution function),

iv. Modified wigner distribution function, Gabor–Wigner distribution operate, and so on (see Gabor–Wigner transform).

A time–frequency distribution operate ideally has the following properties:

a) High clarity to create it easier to be analyzed and understood.

b) No cross-term to avoid confusing real elements from artifacts or noise.

c) A list of fascinating mathematical properties to ensure such strategies profit real-life application.

d) Lower computational complexness to confirm the time required to represent and process an indication on a time–frequency plane permits real-time implementations.

History of the Time-Frequency Analysis

Early work in time–frequency analysis may be seen in the Haar wavelets (1909) of Alfréd Haar, although these weren't considerably applied to signal processing. a lot of substantial work was undertaken by Gabor, like Dennis Gabor atoms (1947), an early type of wavelets, and the Dennis Gabor remodel, a changed short-time Fourier transform. The Wigner–Ville distribution (Ville 1948, during a signal processing context) was another foundational step.

Particularly within the Nineteen Thirties and Nineteen Forties, early time–frequency analysis developed mutually with quantum physics (Wigner developed the Wigner–Ville distribution in 1932 in quantum physics, and was influenced by quantum physics); this is often mirrored within the shared arithmetic of the position-momentum plane and also the time–frequency plane – as within the Werner Karl Heisenberg scientific theory (quantum mechanics) and also the Gabor limit (time–frequency analysis), ultimately each reflective a symplectic structure.

An early sensible motivation for time–frequency analysis was the event of radio detection and ranging.

Applications of Time-Frequency Analysis.

i. Signal process applications

The following applications needn't solely the time–frequency distribution functions however additionally some operations to the signal. The Linear canonical remodel (LCT) is absolutely useful. By LCTs, type} and placement on the time–frequency plane of a proof is within the absolute form that we wish it to be. as an example, the LCTs will shift the time–frequency distribution to any location, dilate it within the horizontal and vertical direction while not ever-changing its space on the plane, shear (or twist) it, and rotate it (Fractional Fourier transform). This powerful operation, LCT, create it a lot of versatile to research and apply the time–frequency distributions.

· Instantaneous frequency estimation

The definition of fast frequency is that the time rate of amendment of section

where    is that the fast section of a proof. we will apprehend the fast frequency from the time–frequency plane directly if the image is obvious enough. as a result of the high clarity is crucial, we frequently use WDF to research it.

· TF filtering and signal decomposition[edit]

The goal of filter style is to get rid of the unwanted element of a proof. Conventionally, we will simply filter within the time domain or within the frequency domain severally as shown below.

The filtering strategies mentioned on top of can’t work well for each signal which can overlap within the time domain or within the frequency domain. By exploitation the time–frequency distribution operate, we will filter within the euclidean time–frequency domain or within the fragmental domain by using the fragmental Fourier remodel.

Filter style in time–frequency analysis continually deals with signals composed of multiple parts, thus one cannot use WDF as a result of cross-term. The physicist remodel, Gabor-Wigner distribution operate, or Cohen's category distribution operate is also higher decisions.

The idea of signal decomposition relates to the {requirement} to separate one element from the others in an exceedingly signal; this could be achieved through a filtering operation that require a filter style stage. Such filtering is historically wiped out the time domain or within the frequency domain; but, may} not be potential within the case of non-stationary signals that square measure multicomponent intrinsically parts could overlap in each the time domain and additionally within the frequency domain; as a consequence, the sole potential thanks to win element separation and so a proof decomposition is to implement a time–frequency filter.

ii. Optics, acoustics, and biomedicine

Light may be a radiation, therefore we tend to apply the time–frequency analysis to optics within the same manner on radiation propagation. Within the same manner, a characteristic of acoustic signals is that, often, its frequency varies extremely severely with time. As a result of the acoustic signals typically contain lots of information, it's appropriate to use less complicated TFDs like the physicist remodel to investigate the acoustic signals because of the lower process complexness. If speed isn't a problem, then a close comparison with well outlined criteria ought to be created before choosing a selected TFD. Another approach is to outline a symbol dependent TFD that's tailored to the info. In biomedicine, one will use time–frequency distribution to investigate the Electroencephalography (EEG), electromyography (EMG), Electrocardiogram (ECG) or otoacoustic emissions (OAEs).

iii. Economic data Analysis with the Gabor Spectgrogram

Economic data is historically considered an erratic and clamorous statistic. The orthodox theory asserts that economic movements are random walks and primarily no regularity exists. Therefore, the predominant analysis tools employed in the economic community have been applied math analysis, like the motor vehicle regressive model (AR). rather than attempting to explore the regularities, economists historically use the empirical data to suit the substitute linear models. For a decent work, the real data has got to be “prewhitened” (in alternative words, we'd like to create the information even additional noisy). However, the stochastic models normally are solely capable of conniving “mean” and “variance,” that aren't adequate for many business applications.

On the opposite hand, expertise from our way of life tells us that there are sure business cycles, although their length could change in time. as an example, most economists believe that since world war II, the trade cycle within the U.S. has been about four to seven years. sadly, neither random analysis nor ancient analysis support their hypothesis. Therefore, we tend to are motivated to hunt new strategies to better perceive the economic movements.

Recently, we tend to applied the joint time-frequency analysis (JTFA) techniques, developed by National Instruments, with encouraging results. in contrast to the traditional Fourier analysis, the joint time-frequency analysis strategies characterize the time series within the time and frequency domains at the same time. Applying JTFA, we all know not solely what quite cycles exist, however conjointly after they occur and the way long they last. whereas the purposeful cycle (the cycles that last for an exact amount of time) is targeted within the joint time-frequency domain, the random noise tends to equally unfold into the whole time-frequency plane. Consequently, the JTFA possesses higher ratio.

References.

http://en.wikipedia.org/wiki/Fourier_transform

http://en.wikipedia.org/wiki/Time%E2%80%93frequency_analysis