[srslte-users] Open loop power control parameters

[Andrea Valori]

Dear All,



Did someone dig into the parameters for open loop power control?

I looked into the theory on this website
http://www.techtrained.com/lte-power-control/ and all seems fairly clear to
me, however, putting theory in practice is less straightforward.



I understand that there are some key parameters in SRS, but after running
some experiments, I have a lot of confusion.



alpha: this seems the clearest parameter. alpha =1 is full power control
(all path loss compensated by the UE increasing its power out), alpha=0 is
no path loss compensation, 0<alpha<1 is fractional path loss compensation.
All this is only in the region where the UE did not reach full power.
[NOTE: this is only the slope and therefore we will also need a parameter
to define the intercept to fully determine the relationship.]

rs_power: this is the power the eNB declares to have sent (required for the
UE to compute the pathloss). What is the reference for this? E.g., if I add
an amplifier on the SDR transmitter, I assume that I will have to adjust
this (since neither the eNB or the UE do know of the presence of the
amplifier). Without an amplifier the default is =0 does this mean that has
to be set to the gain of the front end (antenna and in case amplifiers?) Is
this connected with tx_gain, or the parameter tx_gain is already taken into
account automatically? Can/should this be calibrated with measurements on
the actual TX channel depending on the SDR used?



As stated before, after defining the slope of the UE tx power vs pathloss
(alpha) we will need some other parameters to fully define the
relationship. Which are this one? Mathematically, we need the intercept of
the power control line and the UE out power in the flat region. This could
be a combination of the required power at zero pathloss, or could be the
value of the pathloss at the transition, the power transmitted by the UE at
the flat region, or the power received at the transition between power
control and max power…



Looking at the sib.conf I see some candidates, but missing comments is
difficult to determine the right ones. Below the candidates (with default
values)…

q_rx_lev_min = -65

preamble_init_rx_target_power = -104

…



I assume that the pathloss involved in all the calculations is the one
estimated by the UE (and therefore affected by the parameter rs_power,
which is probably the first parameter to have right)



Lastly, if the system is not balanced (different gains on the TX and RX due
to different external amplifiers), where does this need to be accounted
for? For the TX I believe in the rs_power (as stated above) but what about
the RX?



Thanks for reading this long email, for the insight you may be able to
give, and best regards.


Andrea
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <https://lists.srsran.com/pipermail/srsran-users/attachments/20210804/13eea413/attachment.htm>