Re: [Discuss-gnuradio] IEEE 802.11 a/g/p transreceiver module - Number of symbols per frame?

[Bastian Bloessl]

Hi,

On 05/21/2017 11:45 PM, Qurat-Ul-Ann Akbar wrote:
> Hi,
>
> I want to look at the number of symbols copied for *each frame received 
> *in wifi_rx flow graph. I have been trying to understand how the start 
> of each frame is calculated and by looking at the code it seems like its 
> being done in the wifi sync_short block where the auto correlation 
> coefficient is being checked against a threshold and then the d_plateau 
> is being updated until it reaches MIN_PLATEAU.
>
> However, for the number of symbols per frame, there is a MAX_SAMPLES 
> which is set to 540*80 where 80 is the number of samples per symbol but 
> I don't understand why is the total number of symbols per frame set to 540?

IIRC, that corresponds to a 1500 byte BPSK 1/2 frame. It just pipes in 
the maximum number of samples, since this block doesn't know the actual 
size of the frame.


> Whats the role of MIN_GAP? Why is it being used to detect a shorter frame?


The autcorrelation has a plateau at the beginning of the frame. To avoid 
triggering multiple times, the receiver waits at least MIN_GAP samples.

But even if the receiver still copies samples, it can already 
resynchronize after MIN_GAP samples, i.e., if it's not one large frame, 
but short ones with low interarrival time.


> Moreover, does it mean that the number of symbols per frame can be 
> either MAX_SAMPLES or MIN_GAP because in both cases it starts searching 
> for new frames. What if one frame has less number of symbols than 
> MIN_GAP and MAX_SAMPLES?

It will receive frames with __up to__ MAX_SAMPLES and short frames 
back-to-back, if they are spaced __at least__ MIN_GAP samples.

Best,
Bastian