laforge has submitted this change. ( https://gerrit.osmocom.org/c/osmo-bsc/+/21975 )
Change subject: handover_test: show ineffective intra-cell choice ...................................................................... handover_test: show ineffective intra-cell choice In handover decision 2, choosing an lchan to move so far takes into account only the absolute current rxlev, and it fails to look at the *difference* in rxlev between source and target cell. My recent feature to favor freeing half-used TCH/H does in fact not work well. The main reason why the test for that feature passes is that all lchans in the test get identical measurements -- which is unrealistic. A test with differing ratings uncovers the flaw in comparing intra-cell candidates. Related: SYS#5032 Change-Id: Iabce1ab44b496833c19d999fc3f2903d835c357f --- M tests/handover/test_dyn_ts_favor_moving_half_used_tch_h.ho_vty 1 file changed, 36 insertions(+), 5 deletions(-) Approvals: Jenkins Builder: Verified laforge: Looks good to me, approved diff --git a/tests/handover/test_dyn_ts_favor_moving_half_used_tch_h.ho_vty b/tests/handover/test_dyn_ts_favor_moving_half_used_tch_h.ho_vty index b90236d..9d88c86 100644 --- a/tests/handover/test_dyn_ts_favor_moving_half_used_tch_h.ho_vty +++ b/tests/handover/test_dyn_ts_favor_moving_half_used_tch_h.ho_vty @@ -1,15 +1,46 @@ -# Congestion check: favor moving a TCH/H that frees a half-used dyn TS completely +# Congestion check: favor moving a TCH/H that frees a half-used dyn TS completely. +# The algorithm should notice that this is about moving an lchan within the same cell, so all candidates will remain +# with unchanged rxlev after a re-assignment; hence the current rxlev for each candidate should not make a difference. create-bts trx-count 1 timeslots c+s4 dyn dyn dyn dyn dyn - - +network + handover2 min-free-slots tch/h 6 + +# Test with identical rxlev across lchans (trivial and unrealistic) set-ts-use trx 0 0 states * pdch TCH/HH TCH/H- TCH/HH pdch - - meas-rep lchan 0 0 2 1 rxlev 30 rxqual 0 ta 0 meas-rep lchan 0 0 3 0 rxlev 30 rxqual 0 ta 0 meas-rep lchan 0 0 4 0 rxlev 30 rxqual 0 ta 0 meas-rep lchan 0 0 4 1 rxlev 30 rxqual 0 ta 0 -# pick one to move: -network - bts 0 - handover2 min-free-slots tch/h 6 congestion-check expect-ho from lchan 0 0 3 0 to lchan 0 0 1 0 expect-ts-use trx 0 0 states * TCH/F TCH/HH pdch TCH/HH pdch - - + +# clear measurements for the next run +set-ts-use trx 0 0 states * pdch pdch pdch pdch pdch - - + +# Check that a weaker rxlev coming up earlier in the congestion checking loop does not override the favored half-used +# TCH/H +set-ts-use trx 0 0 states * pdch TCH/HH TCH/H- TCH/HH pdch - - +meas-rep lchan 0 0 2 1 rxlev 30 rxqual 0 ta 0 +meas-rep lchan 0 0 3 0 rxlev 31 rxqual 0 ta 0 +meas-rep lchan 0 0 4 0 rxlev 32 rxqual 0 ta 0 +meas-rep lchan 0 0 4 1 rxlev 33 rxqual 0 ta 0 +congestion-check +expect-ho from lchan 0 0 3 0 to lchan 0 0 1 0 +expect-ts-use trx 0 0 states * TCH/F TCH/HH pdch TCH/HH pdch - - + +# clear measurements for the next run +set-ts-use trx 0 0 states * pdch pdch pdch pdch pdch - - + +# Check that a weaker rxlev coming up later in the congestion checking loop does not override the favored half-used +# TCH/H +set-ts-use trx 0 0 states * pdch TCH/HH TCH/H- TCH/HH pdch - - +meas-rep lchan 0 0 2 1 rxlev 34 rxqual 0 ta 0 +meas-rep lchan 0 0 3 0 rxlev 33 rxqual 0 ta 0 +meas-rep lchan 0 0 4 0 rxlev 32 rxqual 0 ta 0 +meas-rep lchan 0 0 4 1 rxlev 31 rxqual 0 ta 0 +congestion-check +# FAIL: should still pick lchan 0 0 3 0! +expect-ho from lchan 0 0 4 1 to lchan 0 0 1 0 +expect-ts-use trx 0 0 states * TCH/F TCH/HH TCH/H- TCH/H- pdch - - -- To view, visit https://gerrit.osmocom.org/c/osmo-bsc/+/21975 To unsubscribe, or for help writing mail filters, visit https://gerrit.osmocom.org/settings Gerrit-Project: osmo-bsc Gerrit-Branch: master Gerrit-Change-Id: Iabce1ab44b496833c19d999fc3f2903d835c357f Gerrit-Change-Number: 21975 Gerrit-PatchSet: 1 Gerrit-Owner: neels <nhofm...@sysmocom.de> Gerrit-Reviewer: Jenkins Builder Gerrit-Reviewer: laforge <lafo...@osmocom.org> Gerrit-MessageType: merged