0
$\begingroup$

With the current lockdown, I'm spending some time in understanding L1 PHY development and have some questions on the start of packet detection in an IQ capture

Let's say I have a WiFi IQ capture which consists of 1000 packets and want to extract indexes of start and End of each WiFi Packet.

When I plot abs(IQ_samples), it is like

WiFiPacket|NULL|WiFiPacket|NULL|NULL|NULL|WiFiPacket|NULL|WiFiPacket ....

Since the energy of NULL is very much close to zero compared to the max_value of abs(IQ_sample of each packet). Can I keep some kind of threshold to find the start and end of the packet.

This is what I found after searching over internet and going through some posts *) find the minimum value of IQ capture .. that would give me noise value (NVAL) *) take some threshold which is a factor of NVAL (Threshold = factor x NVAL)

My main confusion is what should be the factor value or how should I find that?

Improve this question
$\endgroup$
4
  • $\begingroup$ Do you know anything about the packets? Is there a known symbol sequence (preamble)? $\endgroup$
    – Engineer
    Commented Mar 17, 2020 at 17:22
  • $\begingroup$ Yes.. it does have a STF ,LTF and the data part. $\endgroup$
    – Saira
    Commented Mar 17, 2020 at 17:24
  • $\begingroup$ which appraoch is better in detecting Start of packet ? Is it energy based or autocorr of preamble and then find the SOP ? $\endgroup$
    – Saira
    Commented Mar 17, 2020 at 17:25
  • $\begingroup$ I say go with the auto-correlation based method with the reason being that you should use the information that you have to your advantage always. I asked my earlier question because if you truly had no other knowledge of the signal then the energy based detector is not a bad approach because it wouldn't be throwing away any useful information (there still may be better approaches that I didn't mention) $\endgroup$
    – Engineer
    Commented Mar 17, 2020 at 17:31

1 Answer 1

Reset to default
0
$\begingroup$

Let us say you detected noise level power at -100dBm. You keep additional 10dB threshold to detect energy. Means if you detect >= -90dBm power, you try to detect preamble of Wifi packet. If you detect -95dBm power, you ignore it assuming it may not be a valid Wifi packet. This really works assuming you are very far from base station/access point(AP). Now suppose say you want a higher threshold so that you don't want too many false alarms (may be in order to save battery). So you raise the threshold to -60dBm. Here, if you are too far from AP, you may not trigger at all. If you are close to say within 10 meters, you start to detect energy more than -60dBm and try to detect Wifi packets.

So, you can decide on threshold on how frequently you want to try to detect Wifi signal.

If your threshold is too low, you may more frequently try to detect packets. Advantage - you can detect packets if you are far away from AP. Disadvantage- you may consume more power trying to detect packets.

If you increase threshold, you will detect packets less frequently. Advantage - you can filter out signals which are weak and can use higher modulation rates. Disadvantage - you cannot move too far away from the AP.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.