PCM Transmission Bandwidth: Digital Communications

University of Diyala
Digital Communications
College of Engineering
Lec (3)
Time: (4 hrs)
Date: Oct. 31st , 2013
Dept. of Communications
PCM transmission bandwidth

As we discussed in the previous lecture, each quantization level gets a unique
binary code which contains n binary digits. Recalling the relationship between the
number of quantization levels (L) and the number of binary digits required to
represent each quantization level uniquely,
L=
or
n = log2 L
(1)
Now, since our message signal is band limited signal to B Hz, we require a
minimum of 2B samples per second as a sampling rate, according to the sampling
theorem we discussed before. Therefore, the total bit per second required is 2nB
bits per second (bps).
What is the relationship between bit rate and the transmission bandwidth?
One of the most fundamental relationships in communications is that we can

transmit maximum of 2B pieces of information per second error free over a
noiseless channel of a bandwidth
is given by,
(2)
Example: A signal m(t), which is band limited to 3 KHz, is sampled at a rate
33.33 % higher than the Nyquist rate. The maximum quantization error is 0.5% of
the peak amplitude
. (1) If the quantization levels are binary coded, find the
minimum transmission bandwidth required to transmit the encoded binary signal.
(2) If we time multiplex 24 such signals, what would be the total transmission
bandwidth required to transmit the time multiplexed signal?
Solution: First, we should calculate the Nyquist rate,
Since the signal m(t) is sampled at a rate 33.33% higher than the Nyquist rate, the
actual sampling rate is
Next is the maximum quantization error which can be calculated as follows:

Max error =
where
is the quantization step size
Then,
L= 100/0.5= 200 levels
Note: L must be a power of 2; for example, 4, 8, 16, 32

The next higher L values which satisfies the condition above is 256.
(1) The number of bits required to represent each quantization level uniquely is
n = log2 L = log2 256 = 8 bits per sample.
Now, the data rate as it is given in the example is 8000 samples/s, therefore, we
need to transmit total of 8000*8 = 64000 bits/s.
Because we may only transmit 2 bits/s per one hertz, we need
64000/2= 32000 = 32 KHz of BW.
(2) What would be the transmission bandwidth if we time multiplex 24 such
signals?
If we do time multiplexing, the total bit/s is
24*64000 = 1.536 M bits/s and that amount needs
1.536 M /2 = 0.768 MHz of BW.
Historical Note:
The first mathematician who worked on binary representation of data is
Goltfried Wilhem Leibnitz (1646- 1716). He believed that 1 represents unity
which is a symbol of God and 0 represents the nothingness. Therefore,
representing data using 0 and 1 proves that God has created the universe out of
nothing!
Output SNR:
During the quantization process of PCM, the range (zones which are equally spaced by where,
) is divided into L
(3)
Each sample is approximated to the closest quantization level and finally each
quantization level is binary coded. These codes are sent as binary pulses to their
desired destination. At that destination, some of those samples are detected
incorrectly. Therefore, two sources of errors, quantization errors and pulse code
detection errors. Pulse detection error is so small compared to the quantization
error and for that reason, we can ignore it. Then,
(4)
And,
where,
(5)
represents the message signal.

represents the reconstructed signal.
represents the kth sample of the message signal.
Now, we should calculate the distortion component q(t) which is
(6)
(7)
Where,
q(t) = Unwanted signal (i.e. noise signal) caused by the quantizer. Hence, it is
called quantization noise.
To get the power of the quantization noise, we apply:
(8)
=
(9)
Note:
Since (m = n) in our case, then,

(10)
Since 2B represents the sampling rate and T represents the total interval, the right
side of equation (10) represents the average of the square of the quantization error.
The quantization step size is
then, the mean square quantization error is given by
= 1/
/12 =
(11)
Finally, the reconstructed signal at the receiver side will be

+
(12)
For the purpose of S/N calculations, we can use the following equations:
which is the power of the message signal
which is the power of the quantization noise
Therefore,
(13)
Exponential increase of the output SNR (SNR is directly proportional to BW)

From equation (1),
(14)
After we substitute (14) in (13), we get
where,
(15)
4
from
, n=
using this value of n in equation (15),

(16)
To calculate S/N in dB,
=
= 10
= (Alpha + 6n) dB
(17)
Where, Alpha =10

Example: A signal m(t) whose bandwidth is 4 KHz is transmitted using binary PCM.
Compare the case of L=64 with the case of L=256 from their transmission bandwidth and
the output S/N point of views. (Alpha=-8.51 dB)
Solution:
a) for L=64, n=6, then the transmission bandwidth= nB= 24 KHz
Therefore,
(Alpha + 6n) dB = -8.51+36 = 27.49 dB
b) for L=256, n=8, then the transmission bandwidth= nB= 32 KHz
Therefore,
(Alpha + 6n) dB = -8.51+48 = 39.49 dB
The difference is 12 dB !! while the second one needs only 33% more BW than
the first one does.
To summarize the BW of the PCM, we should write it down as:

where B represents the bandwidth of the original signal.
For example, when n=3, the bandwidth of the generated PCM signal will be at
least 3 times wider than the original signal.
H.W. Looking at equation (17), how much is the improvement in S/N when
increasing n by one bit only? Can you write down a simple rule for that?
Limitation of PCM:
For the purpose of audio and video transmission, PCM has been the most favored
modulation scheme ever used. Some of the advantages of PCM can be
summarized as follows:
1. Less vulnerable to get affected by noise and interferences.
2. Since it involves the transmission of binary pulses, it is easy to regenerate these
pulses during the transmission path.
3. The flexibility of getting SNR improved when the B.W. is slightly increased.
4. TDM may be used to time multiplex PCM signal from different sources.
5. Encryption and decryption can be involved to secure data being transmitted
using PCM scheme.
On the other hand, PCM has come up with mainly two disadvantages,
implementation complexity and the expansion of transmission B.W. The first
one is no longer a big issue due to the availability of very large scale integrated
(VLSI) circuits. These chips are available to implement PCM in a simple way.
The second disadvantage is the transmission bandwidth of PCM. Obviously,
the B.W. of PCM is much wider than the original signal B.W.
Delta Modulation
The idea of oversampling in PCM, as we mentioned in the previous lectures,
is exploited to generate delta modulated signal. That means that the sampling rate
in delta modulation is much higher than the Nyquist rate. The purpose of doing
oversampling is to increase the correlation between the adjacent samples.
Therefore, what delta modulation generates is a staircase approximation to the
over sampled signal and then the difference between the input and the staircase
approximation is quantized into only two levels + and - for the positive and
negative differences. If the approximated version falls below the input signal, the
approximated version is increased by . On the other hand, if the approximation
lies above the input signal, the approximated version is decreased by .
If m(t) is the input signal,

(t) is the staircase approximation, then,
m[n] = m (n
Where, n=0, +1, -1, +2, -2, +3, -3,

m[n ] is the sampled signal of the analog message (t) taken at n
The following set of equations show the steps of DM
e[n] = m[n]=
[n-1]
(18)
* sgn (e[n])
(n) =
(19)
[n]
(20)
Equation (18) computes the error signal e[n]. which is the difference between the
present sample (n) and the previous sample (n-1). Then, this value is quantized to
generate
. The output of equation (19) is either - or + and that value is
added to the previous value to generate the present value as it is shown in

equation (20). Finally, the output is coded into 1 bit only. Figure (1) shows
the procedure of delta modulation.
Figure (1) Shows the procedure of DM

7
The main advantage of DM is the implementation simplicity since it represents

each level using only one bit. The block diagram of DM is shown in figure (2).
DM has three main parts or procedures, comparator, quantizer, and accumulator.
The comparator computes the difference between the two inputs, m[n] and
. The quantizer has an output which is scaled version of signum
function which produces either - or + as we explained. Finally, the quantizer
output is applied to the accumulator to find out the value of the present value
based on the previous values.
Figure (2) DM modulator and demodulator

PCM Transmission Bandwidth: Digital Communications

Uploaded by

Copyright:

Available Formats

PCM Transmission Bandwidth: Digital Communications

Uploaded by

Document Information

Original Description:

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

PCM Transmission Bandwidth: Digital Communications

Uploaded by

Copyright:

Available Formats

University of Diyala

PCM transmission bandwidth

One of the most fundamental relationships in communications is that we can

Next is the maximum quantization error which can be calculated as follows:

is the quantization step size

L= 100/0.5= 200 levels

Note: L must be a power of 2; for example, 4, 8, 16, 32

represents the message signal.

Since (m = n) in our case, then,

Finally, the reconstructed signal at the receiver side will be

Exponential increase of the output SNR (SNR is directly proportional to BW)

using this value of n in equation (15),

To calculate S/N in dB,

Where, Alpha =10

To summarize the BW of the PCM, we should write it down as:

If m(t) is the input signal,

Where, n=0, +1, -1, +2, -2, +3, -3,

added to the previous value to generate the present value as it is shown in

Figure (1) Shows the procedure of DM

The main advantage of DM is the implementation simplicity since it represents

Figure (2) DM modulator and demodulator

You might also like