SAMPLING THEOREM
It is one of useful theorem that applies to digital communication systems.
Sampling theorem states that “A band limited signal (𝒕) with 𝑿(ɷ) = 𝟎 for
|𝑚| ≥ ɷ𝒎 can be represented into and uniquely determined from its samples
(𝒏𝑻) if the sampling frequency 𝒇𝒔 ≥ 𝟐𝒇𝒎, where 𝒇𝒎 is the frequency
component present in it”.
(i.e) for signal recovery, the sampling frequency must be at least twice the
highest frequency present in the signal.
Analog signal (𝑡) is input signal as shown in Fig 4.1, 𝛿𝑇(𝑡) is the train of
impulse shown in Fig 4.2. Sampled signal 𝑥(𝑡) is the product of signal 𝑥(𝑡) and
impulse train 𝛿𝑇(𝑡) as shown in Fig 4.2
EC3354 SIGNALS AND SYSTEM
Applying Fourier transform on both sides
Spectrum of continuous time signal x(t) with maximum frequency 𝜔𝑚 is shown in
Fig 4.5
EC3354 SIGNALS AND SYSTEM
Fig 4.6 Spectrum of 𝑥𝑠 𝑡 when 𝑚𝒔 − 𝑚𝒎 > 𝑚𝒎
Fig 4.7 Spectrum of 𝑥𝑠 𝑡 when 𝑚𝒔 − 𝑚𝒎 = 𝑚𝒎
For 𝑚𝒔 > 2𝑚𝒎
The spectral replicates have a larger separation between them known as guard
band which makes process of filtering much easier and effective. Even a non-ideal
filter which does not have a sharp cut off can also be used.
For 𝑚𝒔 = 𝟐𝑚𝒎
There is no separation between the spectral replicates so no guard band exists and
(𝜔) can be obtained from 𝑋𝑠(𝜔) by using only an ideal low pass filter (LPF) with
sharp cutoff.
EC3354 SIGNALS AND SYSTEM
For 𝑚𝒔 < 2𝑚𝒎
The low frequency component in 𝑋𝑠 𝜔 overlap on high frequency components of 𝑋
𝜔 so that there is presence of distortion and 𝑋 𝜔 cannot be recovered from
𝑋𝑠 𝜔 by using any filter. This distortion is called aliasing.
So we can conclude that the frequency spectrum of 𝑋(𝜔) is not overlapped
for 𝑚𝒔 − 𝑚𝒎 ≥ 𝑚𝒎, therefore the Original signal can be recovered from the
sampled signal.
For 𝑚𝒔 − 𝑚𝒎 < 𝑚𝒎, the frequency spectrum will overlap and hence the
original signal cannot be recovered from the sampled signal.
∴ For signal recovery,
i.e., Aliasing can be avoided if 𝑓𝑠 ≥ 2𝑓𝑚
Aliasing effect (or) fold over effect
It is defined as the phenomenon in which a high frequency component in
the frequency spectrum of signal takes identity of a lower frequency component in
the spectrum of the sampled signal.
When 𝑓𝑠 < 2𝑓 , (i.e) when signal is under sampled, the individual terms in
equation
get overlap. This process of spectral overlap is called frequency folding effect.
EC3354 SIGNALS AND SYSTEM
Occurrence of aliasing
Aliasing Occurs if
i) The signal is not band-Limited to a finite range.
ii) The sampling rate is too low.
To Avoid Aliasing
i) (𝑡) should be strictly band limited.
It can be ensured by using anti-aliasing filter before the sampler.
ii) 𝑓𝑠 should be greater than 2𝑓 .
Nyquist Rate
It is the theoretical minimum sampling rate at which a signal can be sampled
and still be reconstructed from its samples without any distortion
𝑁𝑦𝑞𝑢𝑖𝑠𝑡 𝑟𝑎𝑡𝑒 𝑓𝑁 = 2𝑓 . 𝐻𝑧
The process of obtaining analog signal (𝑡) from the sampled signal 𝑥𝑠(𝑡) is
Data Reconstruction or Interpolation
called data reconstruction or interpolation.
The reconstruction filter, which is assumed to be linear and time invariant, has unit
impulse response h(𝑡).
The reconstruction filter, output (𝑡) is given by convolution of 𝑥𝑠(𝑡) and h(𝑡).
EC3354 SIGNALS AND SYSTEM
𝛿(𝑟 − 𝑛𝑇) 𝑒𝑥𝑖𝑠𝑡 𝑜𝑛𝑙𝑦 𝑎𝑡 𝑟 = 𝑛𝑇
(𝑟 − 𝑛𝑇) = 1 𝑎𝑡 𝑟 = 𝑛𝑇
∞
𝑦(𝑡) = ∑ 𝑥(𝑛𝑇)ℎ(𝑡 − 𝑛𝑇)
𝑛=−∞
Ideal Reconstruction filter
The sampled signal 𝑥(𝑡) is passed through an ideal LPF (Fig 4.9) with
bandwidth greater than 𝑓𝑚 and a pass band amplitude response of T, then the
filter output is 𝑥(𝑡).
Transfer function of ideal reconstruction filter is
The impulse response of ideal reconstruction filter is
𝑓𝑠
2
ℎ(𝑡) = ∫ 𝑇𝑒𝑗𝜔𝑡 𝑑𝑓
−𝑓
𝑠
2
EC3354 SIGNALS AND SYSTEM
Substitute equation 1 in above equation
EC3354 SIGNALS AND SYSTEM
