Вы находитесь на странице: 1из 4

Basics of Signals and Systems

Jan 17th, 2007

Discrete Time signals


Continuous time domain signals are dicult to represent on computer. Converting them into discrete time and discrete amplitude signals enables to represent them on a computer and thereby carry out various signal processing tasks in a much easier and faster way. Discretizing in time is done by sampling the given signal by multiplying it with a train of impulses. Discretizing amplitude is done by quantizing them to a set of discrete values. This process leads to loss of information(distortion) but the advantages of digital representation covers for that loss.

x(t)

t 0

s(t)

t 0

y(t) = x(t) . s(t)

t 0

Figure 1: Continuous time to discrete time conversion

The Sampling Theorem


The sampling theorem states that Exact reconstruction of a continuous-time baseband signal from its samples is possible if the signal is bandlimited and the sampling frequency is greater than twice the signal bandwidth. A signal is bandlimited if it contains no energy at frequencies higher than some bandlimit or bandwidth fm . The samplingtheorem means that the uniformly spaced discrete samples are a complete representation of the signal if this bandwidth is less than half the sampling rate. The sampling rate is called Nyquist Rate fs = 2 fm . In practise sampling rate greater than fs is taken.

Pictorial eect of sampling


A signal x(t) and its FT X() is shown below. The FT the signal is restricted to a bandwidth fm .
x(t) X(w)

t
Original signal x(t)

fm

fm

FT of x(t)

Figure 2: Original signal and its FT

The sampling signal s(t) and its FT are shown below. fs = 2.fm
S(w) s(t)

t
Sampling signal s(t) with fs = 2fm

Figure 3: Sampling signal and its FT

The sampled signal at Nyquist rate and its FT are shown below. We can observe that the FT is the convolution of X() and S() which follows from the property that multiplication in time domain results in convolution in frequency domain.

Sampled signal y(t)

fs

fs

FT of y(t)

Figure 4: Original signal and its FT

Aliasing
When the signal is sampled at a rate less than Nyquist rate, the FT of the sampled signal contains overlapping of high frequency components which leads to distortion. This eect is called aliasing. Hence, in practise the sampling rate is always kept higher than the Nyquist rate.

y(t)

Y(w)

t fs < 2fm

fs

fs

Figure 5: Eect of aliasing

Signal reconstruction
Sampling theorem states that if a signal is sampled at a rate higher than the Nyquist rate then the original signal can be reconstructed exactly without any distortion. This can be visualized from the below gure. If the signal is sent through a lowpass lter (represented by a box in the gure) then the exact signal is recovered.

Y(w) y(t)

Figure 6: Reconstruction of the original signal from the sample signal

Вам также может понравиться