Maker Pro
Maker Pro

DVB-S TV Signal symbol rate, baud rate and bit rate

Lawrenciumbc

Nov 2, 2015
41
Joined
Nov 2, 2015
Messages
41
Context:
I am trying to build a simulation to represent an optical LNB transmitting down a fibre optic distribution network (using Optsim). I am using a bit generator and modulating it with a laser. To match the real network I am working with, it needs to be QPSK with a 22 Msps symbol rate. The bit generator parameters are as bit rate (as Gbit/s) and baud rate. Obviously, the baud rate being half the bit rate for QPSK as its 2 bits per symbol.

Question:
To convert my known symbol rate to bit rate for my simulation, is it as simple as; the symbol rate IS the baud rate and therefore the bit rate is just double the symbol rate?

therefore:
22 Mega symbols/sec = 44 Megabit/sec

or does symbol rate not translate like this?
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
The relation of bit rate to symbol rate basically looks o.k. as stated in your post.

However, I wouldn't think that sending a data stream with a bit rate twice the symbol rate is equivalent to sending a QPSK signal. The modulation scheme is an important factor in your simulation. A single symbol error will result in 2 bit errors, but depending on the coding used some errors may be correctable, others not.
 

Lawrenciumbc

Nov 2, 2015
41
Joined
Nov 2, 2015
Messages
41
ok, so the coding rate also plays a part?

would the bit rate therefore be stated as the full number of bits sent or the number of successful bits sent (with respect to TV)? I would assume the former as data is data regardless of whether is it useful or not.

if my coding rate is 2/3 for a QPSK 22Ms/s signal, would it therefore be 44 Mbits/s with around 29 Mbit/s actually being used?

the simulation is more of an ideal scenario so that I can play with parameters such as cable lengths and other settings in the network, I just wondered what my bit rate would be to start with (or at least a typical realistic value).
 

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
ok, so the coding rate also plays a part?
I think so.

would the bit rate therefore be stated as the full number of bits sent or the number of successful bits sent (with respect to TV)? I would assume the former as data is data regardless of whether is it useful or not.
The transmission channel doesn't know about the contents of the data stream. The channel only sees the raw data and therefore you'll have to deal with the raw data (icl. all synchronization and forward error correction data that may be contained in the data stream) when analyzing the transmission channel.

the simulation is more of an ideal scenario so that I can play with parameters such as cable lengths and other settings in the network, I just wondered what my bit rate would be to start with (or at least a typical realistic value).
Replacing a QPSK signal of 22 Msym/sec by a 44 Mbit/sec binary code is not equivalent. While mathematically the number of bits per second is the same, the physical behavior on the transmission channel is different.
 

Lawrenciumbc

Nov 2, 2015
41
Joined
Nov 2, 2015
Messages
41
Replacing a QPSK signal of 22 Msym/sec by a 44 Mbit/sec binary code is not equivalent. While mathematically the number of bits per second is the same, the physical behavior on the transmission channel is different.

Ok that's fine and partially answers my query, in that case, what might I need to think about in order to make this parameter of my simulation more accurate?

To add more context, I have built a 'representative' LNB in the software from: a data source/random bit generator (for which I would need the bit rate parameter value), a PSK digital modulator component, a sin amplitude modulator (to modulate the generated QPSK signal with the laser) and the laser itself being a Continuous Wave Lorentzian laser.

I have uploaded a screen shot of this section of the model for more clarity.
 

Attachments

  • LNB_sim.PNG
    LNB_sim.PNG
    33.2 KB · Views: 103

Harald Kapp

Moderator
Moderator
Nov 17, 2011
13,700
Joined
Nov 17, 2011
Messages
13,700
I'm sorry, that's above my horizon.
I think (just a guess), that using 2 random bit generators (PSRG), one for each componnet of the quadrature signal would be suitable. Each PSRG running at the symbol rate (22Msym/sec) which then is the same as the bit rate for each generator as the two bits make up one symbol.
 

Lawrenciumbc

Nov 2, 2015
41
Joined
Nov 2, 2015
Messages
41
Ok no problem, that's given me something to think about and experiment with. Thanks for your thoughts!
 
Top