Scribd
Upload
32 views

Description of The Input

The document describes an experiment using a correlator receiver to transmit and receive binary sequences using non-return-to-zero (NRZ) baseband transmission with and without additive white Gaussian noise (AWGN). It generates bit error rate (BER) versus signal-to-noise ratio (SNR) curves from the simulation and compares it to the theoretical curve calculated using the Q-function. Key results shown include error-free reception for NRZ without noise, detection of errors for NRZ with noise, and a plot matching simulated and theoretical BER versus SNR curves.

Uploaded by

ASHIRWAD RAY
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
32 views

Description of The Input

The document describes an experiment using a correlator receiver to transmit and receive binary sequences using non-return-to-zero (NRZ) baseband transmission with and without additive white Gaussian noise (AWGN). It generates bit error rate (BER) versus signal-to-noise ratio (SNR) curves from the simulation and compares it to the theoretical curve calculated using the Q-function. Key results shown include error-free reception for NRZ without noise, detection of errors for NRZ with noise, and a plot matching simulated and theoretical BER versus SNR curves.

Uploaded by

ASHIRWAD RAY
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

EXPERIMENT-2

Baseband transmission (NRZ) using a correlator receiver with and without


noise. Also generate a Pe vs SNR curve using the code and compare it with the
curve generated using the Q-function.

1. Description of the input

A random 20 binary sequence is generated in MATLAB. It is then mapped to


+1 and -1 amplitude level for NRZ polar Line coded signal. 1000000 bits are
generated while plotting Pe vs SNR curve.

2. Description of the output

For NRZ without noise, all the bits that are transmitted are successfully received
indicating number of errors (noe variable) in MATLAB to be zero and
consequently the bit error rate is found to zero (when there is no noise in
channel).

For NRZ with AWGN noise, there will be some significant changes in the bits
received resulting in error. The number of errors and bit error rate are found out
and is NOT equals to zero for this case.

For plotting of simulated bit error rate and theoretical bit error rate, both plots
are in same figure indicated by legends in the right most corner of the figure.

3. All required waveforms or numerical results


seq indicates the detected output bits at receiver and x is the original bit stream
and noe = number of error and ber = bit error rate (NRZ without noise o/p):

The above waveform is for NRZ pulse in presence of AWGN noise. The output
displayed det = detected output bits and x = original bit stream :
The above result is obtained for SNR (in dB) = -10dB. The number of errors
found = 12 out of 20 bits leading to a bit error rate = 0.600 (as shown in the
image in the previous page).

The above image indicates the plot comparison of Pe vs SNR for both simulated
and theoretical results.

4. How the waveforms or numerical results were obtained briefly

For NRZ pulse in absence of noise

 Randomly ‘N’ number of binary sequences are generated and mapped


accordingly to +1 and -1 amplitude levels for NRZ line coding.
 Basically the correlator receiver constitutes two block steps: multiplying
the message signals with the basis signals and then integrating over the
interval Tb.
 So, a basis signal is generated : a unit step of duration 1 unit is generated
in MATLAB for this case.

Then a code is written d1 = length(y); %Total samples in NRZ pulse


h = 1;
ant = [ ];
j=1;
for i=1: d1
if(t(i)<=h)

r1(i) = y(i).*x1(j); % each message of Tb =1 is multiplied with basis signal and accumulated
ant =[ant r1(i)];
ds(h) = sum(ant);
j=j+1;
else
h = h+1;
ant=[];
ds(h) = 0;
j=1;
end
end

 After reception the threshold value is selected as 0 as per nearest


neighboring principle. The value that are above 0 are considered as bit-
‘1’ else it is ‘0’.

For NRZ pulse in presence of noise

 An Gaussian random noise can be generated based on the variance of the


noise that is calculated by the user as:

 So, by using the ‘randn’ built in command , a Gaussian random noise is


generated and added with the message signal as
r=x+sigma.*randn(1,N); Here x is the NRZ signal.
 Then the same product modulator and integrator block is followed as
described for NRZ pulse in absence of noise.
 After reception, it is again compared with the threshold value = 0 ,
anything above 0 is considered as bit-‘1’ is transmitted or else bit ‘0’ is
transmitted.
 Accordingly , the original message signal and the detected output
sequence is compared to find out the numbers in the bit streams.
 BER = (Total number of bit errors )/(Total input binary numbers)
 NOTE: In MATLAB code, SNR is user defined. The user can type for
any value to check the number of errors occurring in ‘N’ number of bits.
For Plotting Pe vs SNR Values:

 A for loop is initiated starting from -35dB to 20 dB and simulated ber


is calculated and stored in a empty matrix.
 For theoretical bit error rate, the formula is

 In matlab , it is written as qfunc(sqrt(2*E bNo))


 Here EbNo inside the Q function is in linear scale. EbNo indicates the
SNR values.
 It is plotted using ‘semilogy’ command in MATLAB which indicates
to plot the values in logarithmic form in y axis (Pe) and SNR (in dB)
along x axis.

You might also like