You are not logged in.
I need to use sequencing to keep track of lost and incoming packets...specifically, I needed to know which packet number arrived or got lost. I looked at the enableSequencing() option but it only prevents duplicate packets from being sent over the line. Could you point the exact location in warpmac.c where the sequence number is actually being added to header?
Offline
We briefly tried to insert sequencing into the system to prevent duplicate packets being received over-the-wire, but there was a problem that still needs some debugging. You're right to notice that the sequence number is not currently being added to the header.
However, if you are just trying to keep track of which retransmission was successfully received, just take over any of the unused (zero-valued) header elements and fill it in with a value that is incremented with every backoff.
Offline
1) Thanks. I already tried adding the sequence number to the empty field, but there is a lot of other housekeeping stuff left to do..... will work on it.
2) Another question: You use an AGC estimate of -85 dBm for calculations. Is it some pre-calculated value or were you just using a noise floor estimate? I checked out the noise floor level in my lab, and it was somewhere around -59 dBm. Does this affect the AGC algorithm in any way, since the target remains the same at -10dBm?
3) Using only the API provided for WARP, and not digging into Simulink/SysGen, is it possible to do some BER calculations using WARP as a testbed?
Thanks again for your help.
Offline
kshitji wrote:
Another question: You use an AGC estimate of -85 dBm for calculations. Is it some pre-calculated value or were you just using a noise floor estimate? I checked out the noise floor level in my lab, and it was somewhere around -59 dBm. Does this affect the AGC algorithm in any way, since the target remains the same at -10dBm?
The noise estimate is used to calculate thresholds in the AGC algoritm and depends only on properties of the radio and RF front-end, not the environment's noise level.
kshitji wrote:
Using only the API provided for WARP, and not digging into Simulink/SysGen, is it possible to do some BER calculations using WARP as a testbed?
Let's make this its own thread: BER Testing.
Offline