Re: [Discuss-gnuradio] Question on the Packet Decoder

[Zhiwen He]

I tried 4 QAM and 16 QAM.

I used a FFT Sink to see the Amplitude in dB after the kanal, which include the signal and the noise, I think :

SNR = Amplitude of Singal - Amp. of Noise,  is this right?

On Wed, Jan 23, 2013 at 4:23 PM, Tom Rondeau <address@hidden> wrote:

On Wed, Jan 23, 2013 at 10:20 AM, Zhiwen He <address@hidden> wrote:

On Wed, Jan 23, 2013 at 4:11 PM, Zhiwen He <address@hidden> wrote:

you mean that i need to add an access code for the system, or just leave blank?

There is a default 64-bit access code that is used when you leave this field blank.

 

how to make the system better? I mean that, for example, when the SNR< 20, there are too many errors for the packet decoder, it doesn't work anymore, but for practise, i think it should keep working, although  there are much more errors. for the simulation, how can I get the data for SNR<20, so that i can compare with the source data to  calculate the BER?

Thanks,

Zhiwen

It's possible that you are calculating your SNR incorrectly.I know that under standard channel conditions (that is, AWGN) the modulators work pretty close to their theoretical performance. So when you say SNR < 20, which modulation are you trying to use? GMSK, DBPSK, and DQPSK should work under these conditions.

Tom

 

On Wed, Jan 23, 2013 at 3:56 PM, Tom Rondeau <address@hidden> wrote:

On Tue, Jan 22, 2013 at 11:07 AM, Zhiwen He <address@hidden> wrote:

Hello all,

i'm working on the digital modultaion, and i build a system like this:

random source -> packet encoder -> qam mod -> (adder and noise source) -> qam demod -> packet decoder ->file sink

I found that,  when the amplitude of the noise is low, there will be some data in file sink, but when the amplitude of noise is high, the packet decoder doesn't output any more. Who can tell me why? does the noise destroy the structure of the packet so that i can not be decoded any more?

I also tried psk mod/demod and gmsk mod/demod, it comes the same problem.

thanks,

Zhiwen

Yes, you are correct. The packet decoder looks for an access code. If the access code is valid to some number of bits, it will pull out a frame. If there are too many garbles in the access code, the packet decoder will never see it as a valid frame. By default, the access code is 64 bits and, I think, there can be up to 12 bit errors in it (this is the threshold value).

Tom