cellular-infrastructure
/
osmo-bsc
mirror of https://gerrit.osmocom.org/osmo-bsc
9
0
Fork 0
Code Issues Releases Wiki Activity

handover: apply meas report BS Power to RXLEV, fix ho oscillation 

This fixes handover oscillation in the presence of power reduction on
the downlink: apply the reduced power to the cell's RXLEV in order to
not rate the current cell as weaker than it actually is.

Related: SYS#5339
Change-Id: Ifcf59964b5e2d550d79e4ba14d90962808f79dae
This commit is contained in:
Neels Hofmeyr 2021-05-26 20:47:34 +02:00 committed by laforge
parent cdaf3ec006
commit be161380a3
2 changed files with 9 additions and 10 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

8
src/osmo-bsc/meas_rep.c
View File

@ -31,11 +31,15 @@ static int get_field(const struct gsm_meas_rep *rep,
case MEAS_REP_DL_RXLEV_FULL:
if (!(rep->flags & MEAS_REP_F_DL_VALID))
return -EINVAL;
return rep->dl.full.rx_lev;
/* Add BS Power value to rxlev: improve the RXLEV value by the amount of power that the BTS is reducing
* transmission. Note that bs_power is coded as dB/2, a positive value indicating the amount of power reduction
* on the downlink; rxlev is coded in dB, where a higher number means stronger signal. */
return rep->dl.full.rx_lev + (rep->bs_power * 2);
case MEAS_REP_DL_RXLEV_SUB:
if (!(rep->flags & MEAS_REP_F_DL_VALID))
return -EINVAL;
return rep->dl.sub.rx_lev;
/* Apply BS Power as explained above */
return rep->dl.sub.rx_lev + (rep->bs_power * 2);
case MEAS_REP_DL_RXQUAL_FULL:
if (!(rep->flags & MEAS_REP_F_DL_VALID))
return -EINVAL;

11
tests/handover/test_bs_power.ho_vty
View File

@ -6,11 +6,6 @@ create-bts trx-count 1 timeslots c+s4 TCH/F TCH/F TCH/F TCH/F TCH/F TCH/F PDCH
set-ts-use trx 0 0 states * TCH/F - - - - - *
meas-rep repeat 10 bspower 20 lchan 0 0 1 0 rxlev 20 rxqual 0 ta 0 neighbors 30
# FAIL: there should be no handover, because the bspower reduction of 20 with an rxlev of 20 (= 40) is stronger than the
# neighbor at 30. But since osmo-bsc fails to take the BS Power into account, it hands over to the neighbor.
#expect-no-chan
expect-ho from lchan 0 0 1 0 to lchan 1 0 1 0
# When the neighbor then reports the stronger original cell, we get handover oscillation
meas-rep repeat 10 bspower 0 lchan 1 0 1 0 rxlev 30 rxqual 0 ta 0 neighbors 40
expect-ho from lchan 1 0 1 0 to lchan 0 0 1 0
# there should be no handover, because the bspower reduction of 20 with an rxlev of 20 (= 40) is stronger than the
# neighbor at 30.
expect-no-chan