About Implementing a AGC
显示 更早的评论
I am trying to Implement AGC,After demodulator block i am using AGC Block.When i have given input to my AGC it is multiplying every sample with the same gain but agc should be choose the which sample has to be multiplied with what amount of gain.so what logic is usefull for me to wirte so that my AGC can choose the gain according to the sample...
5 个评论
Dimitris Kalogiros
2019-9-6
1) Are you working with matlb or simulink ?
2) do you prefer a feedback or a feedforward structure?
Keep in mind that AGCs aims to keep the average power of the incoming signal almost constant
Hari Ijjada
2019-9-6
shilpa kamble
2020-3-20
How to Implement AGC using simulink ?
Dimitris Kalogiros
2020-3-26
Dear Shilpa
Allow me to ask: Your problem is with simulink implementation or you feel that you don't have understood the functionality of an agc loop ?
shilpa kamble
2020-3-26
sir, I have problem with simulink implementation.Atcually I m tring to do implement the AGC to adjust gain of AGC to have an constant signal level output.
采纳的回答
Dimitris Kalogiros
2019-9-6
Dear Hari
Here you are a simple version of a "feedback" structured AGC
clearvars; clc;
%% time instances
dt=0.001
t=0:dt:20;
%% input signal
A=10; f=0.25;
xin=A*sin(2*pi*f.*t);
%% AGC parameters
% loop filter gain
L=round(10*f*(1/dt));
% Target power
powerTarget=0.2;
%% AGC process
yout=zeros(1,length(xin));
gainAGC=ones(1,length(xin));
for n=1:length(t)
% amplify current input sample
yout(n)=gainAGC(n)*xin(n);
% adjust agc gain, with feedback branch
gainAGC(n+1)=gainAGC(n)*(1 - (1/L)*(yout(n)^2-powerTarget) );
end
%% Plot input and output
figure;
subplot(2,1,1);
plot(t, xin, '-b'); hold on;
plot(t, yout, '-r'); zoom on; grid on;
xlabel('t in sec'); ylabel('amplitude');
legend('xin', 'yout'); title('AGC input and output');
subplot(2,1,2);
plot(t, gainAGC(1:end-1)); zoom on; grid on;
title('gain of AGC');
I have choosen a very simple version of gain adaptation.
Keep in mind that parameter L controls the speed of AGC lock. Large values of L leads to a fast but noisy convergence.
If you run this code, with the parameters I've choosen, you will get the following results:
2 个评论
Hari Ijjada
2019-9-7
Dimitris Kalogiros
2019-9-7
It depents on your sampling rate. For example, suppose you have voice signal , sampled at Fs. You can set L=100*Fs or L=1000*Fs (Fs is expressed in Hz)
By the way, you can experiment on the value of L.
更多回答（0 个）
类别
在 帮助中心 和 File Exchange 中查找有关 Modulation 的更多信息
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!
Translated by 
