01.05.2017 Views

563489578934

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Sec. 2–8 Discrete Fourier Transform 97<br />

least N numbers must be stored, and furthermore, the average sampling rate must be at least<br />

the Nyquist rate. That is, † f s Ú 2B<br />

(2–175)<br />

Thus, in this first type of application, the dimensionality theorem is used to calculate the number<br />

of storage locations (amount of memory) required to represent a waveform.<br />

The second type of application is the inverse type of problem. Here the dimensionality<br />

theorem is used to estimate the bandwidth of waveforms. This application is discussed in detail<br />

in Chapter 3, Sec. 3–4, where the dimensionality theorem is used to give a lower bound<br />

for the bandwidth of digital signals.<br />

2–8 DISCRETE FOURIER TRANSFORM ‡<br />

With the convenience of personal computers and the availability of digital signal-processing<br />

integrated circuits, the spectrum of a waveform can be easily approximated by using the discrete<br />

Fourier transform (DFT). Here we show how the DFT can be used to compute samples<br />

of the continuous Fourier transform (CFT), Eq. (2–26), and values for the complex Fourier series<br />

coefficients of Eq. (2–94).<br />

DEFINITION.<br />

The discrete Fourier transform (DFT) is defined by<br />

X(n) =<br />

k=N-1<br />

a x(k)e -j(2p/N)nk<br />

k=0<br />

(2–176)<br />

where n = 0, 1, 2, Á , N - 1, and the inverse discrete Fourier transform (IDFT) is defined<br />

by<br />

n=N-1<br />

X(n)e j(2p/N)nk<br />

x(k) = 1 (2–177)<br />

N a n=0<br />

where k = 0, 1, 2, Á , N - 1.<br />

Time and frequency do not appear explicitly, because Eqs. (2–176) and (2–177) are just definitions<br />

implemented on a digital computer to compute N values for the DFT and IDFT, respectively. One<br />

should be aware that other authors may use different (equally valid) definitions. For example, a<br />

1/1N factor could be used on the right side of Eq. (2–176) if the 1/N factor of Eq. (2–177) is<br />

replaced by 1/1N. This produces a different scale factor when the DFT is related to the CFT.<br />

Also, the signs of the exponents in Eq. (2–176) and (2–177) could be exchanged. This would<br />

reverse the spectral samples along the frequency axis. The fast Fourier transform (FFT) is a fast<br />

algorithm for evaluating the DFT [Ziemer, Tranter, and Fannin, 1998].<br />

† If the spectrum of the waveform being sampled has a line at f = +B, there is some ambiguity as to whether<br />

the line is included within bandwidth B. The line is included by letting f s 7 2B (i.e., dropping the equality sign).<br />

‡ The fast Fourier transform (FFT) algorithms are fast ways of computing the DFT. The number of complex<br />

multiplications required for the DFT is N 2 , whereas the FFT (with N selected to be a power of 2) requires only (N/2)<br />

log 2 N complex multiplications. Thus, the FFT provides an improvement factor of 2N/(log 2 N) compared with the<br />

DFT, which gives an improvement of 113.8 for an N = 512 point FFT.

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!