Reply by Tim Wescott November 5, 20072007-11-05
On Mon, 05 Nov 2007 07:51:25 -0600, Vladimir Vassilevsky wrote:

> "uvbaz" <uvbaz@stud.uni-karlsruhe.de> wrote in message > news:1194260617.582181.288990@o80g2000hse.googlegroups.com... >> Hi, >> >> I thought the code gain of one channel code is independent upon >> modulation, but the simulation in MATLAB/Simulink with different >> modulation shown different codegain. > > Of course it does. The code and the modulation are always considered > altogether. It has to do with the nonlinearity and the way the code mapped > to the signal. > >> >> Channelcode: Convolutional, coderate 1/2, memory depth 6, traceback >> 32, 3bits-softdecision >> Bitrate after decoder: 20kbit/s >> >> 1. FM, with 10khz deviation, 40khz passband before FM-Demodulate, >> 25khz baseband after FM-Demodulate, Codegain 1.5dB. >> >> 2. FM, with 10khz deviation, 40khz passband before FM-Demodulate, >> 40khz baseband after FM-Demodulate, Codegain 2.6dB. >> >> 3. FM, with 40khz passband before PM-Demodulate, 25khz baseband after >> PM-Demodulate, Codegain 3.5dB. >> >> How should I explain it? > > The gain of the code depends on the input SNR. The higher the SNR the better > is the gain. > Your demodulator is unoptimal and highly nonlinear, therefore output SNR != > input SNR. >
You're forgetting the noise characteristics -- atmospheric noise at 300kHz has essentially infinite variance, so any coding scheme that will recover data across a lightning crash has "infinite" gain. This doesn't mean that any coding scheme is infinitely good, it just means that you need to go find another metric... -- Tim Wescott Control systems and communications consulting http://www.wescottdesign.com Need to learn how to apply control theory in your embedded system? "Applied Control Theory for Embedded Systems" by Tim Wescott Elsevier/Newnes, http://www.wescottdesign.com/actfes/actfes.html
Reply by Vladimir Vassilevsky November 5, 20072007-11-05
"uvbaz" <uvbaz@stud.uni-karlsruhe.de> wrote in message
news:1194260617.582181.288990@o80g2000hse.googlegroups.com...
> Hi, > > I thought the code gain of one channel code is independent upon > modulation, but the simulation in MATLAB/Simulink with different > modulation shown different codegain.
Of course it does. The code and the modulation are always considered altogether. It has to do with the nonlinearity and the way the code mapped to the signal.
> > Channelcode: Convolutional, coderate 1/2, memory depth 6, traceback > 32, 3bits-softdecision > Bitrate after decoder: 20kbit/s > > 1. FM, with 10khz deviation, 40khz passband before FM-Demodulate, > 25khz baseband after FM-Demodulate, Codegain 1.5dB. > > 2. FM, with 10khz deviation, 40khz passband before FM-Demodulate, > 40khz baseband after FM-Demodulate, Codegain 2.6dB. > > 3. FM, with 40khz passband before PM-Demodulate, 25khz baseband after > PM-Demodulate, Codegain 3.5dB. > > How should I explain it?
The gain of the code depends on the input SNR. The higher the SNR the better is the gain. Your demodulator is unoptimal and highly nonlinear, therefore output SNR != input SNR. Vladimir Vassilevsky DSP and Mixed Signal Consultant www.abvolt.com
Reply by uvbaz November 5, 20072007-11-05
Hi,

I thought the code gain of one channel code is independent upon
modulation, but the simulation in MATLAB/Simulink with different
modulation shown different codegain.

Channelcode: Convolutional, coderate 1/2, memory depth 6, traceback
32, 3bits-softdecision
Bitrate after decoder: 20kbit/s

1. FM, with 10khz deviation, 40khz passband before FM-Demodulate,
25khz baseband after FM-Demodulate, Codegain 1.5dB.

2. FM, with 10khz deviation, 40khz passband before FM-Demodulate,
40khz baseband after FM-Demodulate, Codegain 2.6dB.

3. FM, with 40khz passband before PM-Demodulate, 25khz baseband after
PM-Demodulate, Codegain 3.5dB.

How should I explain it?

Thanks!
uvbaz