You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
This is a fix for V3 after some discussion in discord.
In RadioInterface.h we set some initial values for slotTimeMsec, preambleTimeMsec, and maxPacketTimeMsec, but the second and third of those are recalculated on startup in order to match the actual modem settings in use. However, slotTimeMsec is not, so it uses the fixed values for spreading factor and bandwidth that appear in RadioInterface.h (sf 9, bw 125) regardless of radio settings. We should update this since the actual LongFast values would create a longer slot time (~77ms rather than ~42), so this could in theory lead to some failures in channel activity detection (since we're not leaving enough time for the process to happen).
The formula may also need or want adjustment, or formulas that use this value. Possibly the maximum contention window size will be too large with the longer slot time, or other issues. This should be tested/simulated, probably.
Relevant log output
No response
The text was updated successfully, but these errors were encountered:
Category
Other
Hardware
Not Applicable
Firmware Version
2.x
Description
This is a fix for V3 after some discussion in discord.
In
RadioInterface.h
we set some initial values forslotTimeMsec
,preambleTimeMsec
, andmaxPacketTimeMsec
, but the second and third of those are recalculated on startup in order to match the actual modem settings in use. However,slotTimeMsec
is not, so it uses the fixed values for spreading factor and bandwidth that appear inRadioInterface.h
(sf 9, bw 125) regardless of radio settings. We should update this since the actual LongFast values would create a longer slot time (~77ms rather than ~42), so this could in theory lead to some failures in channel activity detection (since we're not leaving enough time for the process to happen).The formula may also need or want adjustment, or formulas that use this value. Possibly the maximum contention window size will be too large with the longer slot time, or other issues. This should be tested/simulated, probably.
Relevant log output
No response
The text was updated successfully, but these errors were encountered: