Data bits

FSK wave form

is the modulation used in GSM cell phones.

ISI (Inter symbol Interference) : Symbol when prolonged gets mixed (or interfered) with the next symbol.
When symbol's time duration is very small => number of frequencies become more as bandwidth is inversely proportional to the
symbol's time duration.
Symbol with very small duration
Channel which can pass only little ==> symbol with frequencies that are
or Symbol with more frequencies ==>
number of frequencies f1, f2 allowed by the channel alone. (only f1, f2 fq.es alone)
(f1, f2, --- ; fn)

At the o/p of the Channel; symbol has lesser number of frequencies => symbol duration gets prolonged in time domain.
Before passing through channel, 1st symbol duration = 0 to T1; 2nd symbol duration = T1 to 2 T1. In the channel o/p; when symbol1 is
prolonged (i.e. from 0 to 3 T1/2) ; but symbol2 gets started from T1 onwards, then 1st symbol gets mixed with 2nd symbol. 2nd symbol
when prolonged gets mixed with 3rd symbol.
Sidebands increase the no. of fq.es in the symbol. Before passing the symbol through the
.
channel itself, side bands of the symbol are removed by passing the symbol through LPF (or Gaussian filter). Ex: Speech signal has max. strength
at around 3.4 KHz . Also, speech signal has min. strength on fq.es 4K to 20 KHz. So, filter removes the fq.es 4KHz to 20KHz, so that signal
can be passed
(or symbol)
through the
channel as it is
with out getting affected.

T1

O/P of channel

Fq.es in the o/p of channel are :

f1 to f1 + delta.f .
Also, f1 + delta.f < f2 .

(I/p signal has fq.es 0 to f2. Bit, signal in the o/p of

3 T1
channel has lesser fq.es. ==> channel o/p gets
2
extended in its time duration. Data bit1 when
extended gets mixed with data bit2. This is ISI.

Gaussian
symbol with

I/p Symbol the removal of

side bands

Gaussian filter in the time domain

does not prolong the symbol duration
(more than the required duration ).
Gaussian filter in fq. domain
does not produce side bands.

Sidebands in the symbol gets

removed in the o/p of gaussian
filter.

(I/P Symbols) Filter has gaussian

impulse response.
This filter passes
only some fq.es as
F.T. for gaussian
impulse
response is
gaussian in fq.
domain.
 Minimum shift Keying

(ISI is removed because of Equalizer)

Equalizer

X(f) =
Receiver gets X(f) H(f) Equalizer O/p
I/P symbol in the
Disturbances in channel Receiver
Equalizer o/p =
( ISI is created because of
imperfections in channel) X(f).H(f). 1
= X(f) = Input. ==> Disturbances in the
H(f) channel get nullified if we
know the
channel H(f) perfectly
at Receiver.
(Code division multiple Access)

Base station assigns different codes to different cell phones.

Only the base station and the cell phone knows about the assigned code.
i.e. cell phone2 knows only about the code that is assigned to cell phone 2 alone. and
cellphone 2 does not know about codes assigned to other cell phones.
 bit1.code1 (or )

bit1 is transmitted from cell phone1

bit2 is transmitted
from cellphone2

bit2.code2

====>

=> Base station receives this

signal.

Base station multiplies the received

signal with code1 to get the bits
transmitted from cell phone1.
Similarly, to get the transmitted signal
from cell phone2, Base station multiplies
the received signal with code2.

= data that is transmitted from cell phone1

Related to the cell phone or user eqipment (UE) .

UE transmit (or cell phone is transmitting to the base station)
SDR architecture
(for wide band RF
Front-end)

Base station is sending some data to cell phone

UE receive

in wideband-CDMA (or W-CDMA).

WCDMA
( or wideband-
CDMA) in 3G
cellular system.
 (200KHZ is the bandwidth in
GSM cellular systems).

cos(--2 pi fc t) = cos(2 pi fc t)
===> In-Phase component

modulated
Related to the wave
SDR architecture
(for wide band RF
Front-end -- CDMA (In-Phase) = QPSK * cos( 2 pi fc t )
Base station )

= In-phase component of the transmitted signal

passed through
Low pass filter
.
Sin(-- 2 pi fc t ) = -- sin ( 2 pi fc t ) ==> Quadrature-phase component.

= QPSK * sin( 2 Pi fc t )

Quadrature - phase c
= component of transmitted
signal
 (Please see in the text book)

Tx
 Design of any communication system is affected by the following ::

a)

( Flow of electrons in random directions in any analog

electronic device creates this Thermal noise).
(Ex.: Bike starting creates sparks. These
sparks create interference. Also, when a
person is moving with more speeds, then that
person's cell phone is affected by doppler
shifts. )

(When it is raining, lightning and thunders happen. These create interference to the required signals.)

(In CDMA, all cell phones use the same fq. channel. If some number of cell phones operate under the same Base station, then there
is no problem in interpreting the cell phone signal at Base station because of orthogonal codes. But, if huge no. of cell phones are
communicating with Base station, then Base station may not exactly identify the cell phone signal. This is because, all cell phones
are using the same fq. channel. ) Ex : In a class room, if 10 persons are speaking, then we can
understand each person's speech. But, if 100 persons are
speaking at the same time, then no one can understand anything.)
Cascaded Analog System ==>:

1st Analog
system
produces noise.
Amplifier1 amplifies
When this noise both of the i/p signal
is passed through & i/p noise. along
the 2nd analog system, with the amplification of i/p noise,
it generates its own noise as well.
then this 2nd analog system
produces some other noise
along with the 1st system generated noise.

In order to remove the noise in analog devices, we have to design the gains and noise -figures perfectly in the Receiver.

Digital Systems are affected by Quantization noise ==>

Same
uantization
Original analog voltage is : 6 Volt.
But, Quantization levels that are that was
supported in ADC are : 0V, 5V & generated in ADC.
10V. Then, ADC takes the Noise generated in ADC.
i/p analog voltage as 5V.
This 5V when transmitted gets received as 5V Digital systems also suffer from Overflow errors.
only. That implies that, we can't retrieve 6V in the
Receiver as well. ==> Digital systems have to
face the problem of Quantization.

For Ex :: C -- language supports 16 bits.

Then, maximum quantity that can be used in the C-language is :

16 16
2 --- 1 . . Then, we can't use the value 2 .
 in W/Hz/K

in Kelvin

IN Hz.

Noise affects the capacity ( or data rate ) of the transmission system.

(or) number of bits that can be received in a second with bearable number of errors in the
Receiver.

received in the Rx

generated in the Rx
(or noise received by the Rx)

Ratio