Nyquist criterion for distortionless baseband transmission in the frequency domain
in absence of Noise :
(1) where Tb = 1/Rb
Ideal Nyquist Pulse
The simplest way of satisfying (1) is to specify the frequency function P( f ) to be
in the form of a rectangular function, as shown by
(2)
where rect( f ) stands for a rectangular function of unit amplitude and unit
support centered on f = 0 and the overall baseband system bandwidth W is
defined by
The signal that produce Zero ISI can be obtained by taking the IFT of p(f)
IFT of p(f) (3)
The special value of the bit rate Rb = 2W is called the Nyquist rate and W is itself called the
Nyquist bandwidth. Correspondingly, the baseband pulse p(t) for distortionless transmission
described in (3) is called the ideal Nyquist pulse, ideal in the sense that the bandwidth
requirement is one half the bit rate.
The function p(t) can be regarded as the
impulse response of an ideal low-pass
filter with passband magnitude response
1/2W and bandwidth W.
p(t) has its peak value at the origin and goes
through zero at integer multiples of the bit
duration Tb. If the received waveform y(t) is
sampled at the instants of time t = 0, 1Tb,
2Tb, .., then the pulses defined by ai p(t – iTb)
The normalized form of P( f ) is with amplitude ai and index i = 0, ±1, ± 2,..
plotted for positive and negative will not interfere with each other.
frequencies
Example: A series of sinc pulses
corresponding to the sequence Although the use of the ideal Nyquist pulse does indeed achieve
1011010. economy in bandwidth, in that it solves the problem of zero ISI with
the minimum bandwidth possible, there are two practical difficulties
that make it an undesirable objective for signal design:
1. It requires that the magnitude characteristic of P( f ) be flat from –
W to +W, and zero elsewhere. This is physically unrealizable
because of the abrupt transitions at the band edges ±W,
2. The pulse function p(t) decreases as 1/| t | for large | t |, resulting
in a slow rate of decay. This is also caused by the discontinuity of
P( f ) at ± W. Accordingly, there is practically no margin of error
in sampling times in the receiver.
To evaluate the effect of the timing error alluded to under point 2,
consider the sample of y(t) at t = ∆t, where ∆t is the timing error. To simplify the
exposition, we may put the correct sampling time ti equal to zero. In the absence
of noise, we thus have from the first line of (4):
(4)
(5)
(6) Since 2WTb = 1, by definition
(7)
The first term on the right-hand side of (7) defines the desired symbol, whereas the remaining series represents the
ISI caused by the timing error ∆t in sampling the receiver output y(t). Unfortunately, it is possible for this series to
diverge, thereby causing the receiver to make erroneous decisions that are undesirable.
Raised-Cosine Spectrum
(1) where Tb = 1/Rb
We may overcome the practical difficulties encountered with the ideal Nyquist pulse by
extending the bandwidth from the minimum value W = Rb/2 to an adjustable value between W
and 2W
P( f ) is designed to satisfy a condition more stringent than that for the ideal Nyquist pulse, in
that we retain three terms of the summation on the left-hand side of (1) and restrict the
frequency band of interest to [–W, W], as shown by
(2) where, on the right-hand side, we have set Rb = 1/2W
We may now devise several band-limited functions that satisfy (2). A particular form of P( f ) that embodies
many desirable features is provided by a raised-cosine (RC) spectrum. This frequency response consists of a flat
portion and a roll-off portion that has a sinusoidal form, as shown by:
(3)
In (3), we have introduced a new frequency f1 and a
dimensionless parameter α, which are related by
(4)
The parameter α is commonly called the roll-off factor; it
indicates the excess bandwidth over the ideal solution, W.
Specifically, the new transmission bandwidth is defined
by
(5)
The frequency response P( f ), normalized by multiplying it by the factor 2W, is plotted in Figure 1a for α = 0, 0.5,
and 1.
We see that for α = 0.5 or 1, the frequency response P( f ) rolls off gradually compared with the ideal Nyquist pulse
(i.e., α = 0) and it is therefore easier to implement in practice.
This roll-off is cosine-like in shape, hence the terminology “RC spectrum.”
Just as importantly, the P( f ) exhibits odd symmetry with respect to the Nyquist bandwidth W, which makes it
possible to satisfy the frequency-domain condition of (1).
Responses for different roll-off factors:
frequency response time response.
The time response p(t) is naturally the inverse Fourier transform of the frequency response P( f ). Hence, transforming
the P( f ) defined in (3) into the time domain, we obtain
(6)
Observations:
(i) For α = 0.5 and 1 , the p(f) changes gradually with respect to frequency, hence it is easier to realize this
practically
(ii) The time response has a sinc shape and all the sinc funtions pass through zero at t = ±Tb, ± 2Tb,..
(iii) The amplitude of side lobes increases with reduction in the value of α .
(iv) With α = 0, the bandwidth requirement is maximum equal to 2W.