PDA

View Full Version : Push 3G CS Traffic to co 2G cell



fahmi
2015-11-26, 07:39 PM
Dear Huawei experts,

Which is the best way to push 3G CS traffic to CO 2G cell ?

BR,
Fahmi

RF engineer
2015-11-26, 08:52 PM
try directed retry from 3G to 2G

Directed Retry to GSM:

Set loadSharingGsmThreshold to 95
Set loadSharingGsmFraction to 50

BR

sade
2015-11-27, 11:04 PM
Alternatively you can use LDR algorithm (Load Reshuffling)

While directed retry works when bearers are established / reconfigured, LDR can shift traffic which is already established (if proper load tresholds are crossed)

miles12
2015-12-06, 06:57 PM
Alternatively you can use LDR algorithm (Load Reshuffling)

While directed retry works when bearers are established / reconfigured, LDR can shift traffic which is already established (if proper load tresholds are crossed)


Dear Sade
Could you please explain with parameter name with value.

Thanks in advance

gudubettin
2015-12-07, 03:45 PM
You can raise CIO for irat neighborhood to get easier handover to 2g,

senjamir
2015-12-07, 03:52 PM
Hi,

Try LDR, the trigger should be power resources so that you can still get traffic from 3G and once it is trigger the action should be Inter-RAT load handover.

NBMLDCALGOSWITCH = ULLDR, DLLDR (Enabled)

NodeBLdcAlgoSwitch = IUB_LDR (Enabled)

ULLDRTRIGTHD = 55%
ULLDRRELTHD = 45%
DLLDRTRIGTHD = 50%
DLLDRRELTHD = 40%

Check for the suitable values for your network for this depend on the current traffic that you are getting.

I hope this help.

Best regards,
Sen


Dear Huawei experts,

Which is the best way to push 3G CS traffic to CO 2G cell ?

BR,
Fahmi

samuraial
2015-12-07, 09:57 PM
UCELLINTERRATHOCOV for cell level and UINTERRATHOTCOV for network level.

Inter-RAT CS Measure Start RSCP THD = -104(same as the one below)
Inter-RAT CS Measure Stop RSCP THD = -101(same as the one below)
Inter-RAT CS Measure Start Ec/No THD = -13(higher the value more traffic pushed from 3G to 2G for CS)
Inter-RAT CS Measure Stop Ec/No THD = -11(higher the value more traffic pushed from 3G to 2G for CS)

This will increase your CM attemps but push traffic to 2G from 3G.

B/R

sade
2015-12-08, 02:06 AM
senjamir is correct, what else I could add:

1) Tresholds mentioned by senjamir (ULLDRTRIGTHD, ULLDRRELTHD, DLLDRTRIGTHD, DLLDRRELTHD) can be tuned up and down - depending if you want to increase offloading traffic to 2g (then reduce tresholds) or keeping more traffic in 3g (then increase thresholds)
2) When it comes to LDR actions, they are defined separately for DL and UL. Pamaters which control them are listed below. For each of those parameters you can choose value. Values which enable LDR inter-rat handover are: CSInterRatShouldBeLDHO, CSInterRatShouldNotBeLDHO. What the difference between those 2? If during Rab Assigment you will get "CS Inter Rat HO should not be performed" - LDR can trigger HO for such RABs only when CSInterRatShouldNotBeLDHO is selected as one of the actions. By analogy, RABs which have "CS Inter Rat HO should be performed" will be handovered by LDR only when CSInterRatShouldBeLDHO is selected.
To maximize gain from this feature, set those 2 actions as the highest priority ones (as first and second action - order does not matter in my opinion).
UCELLLDR\DlLDRFirstAction
UCELLLDR\DlLDRSecondAction
..
UCELLLDR\DlLDRTwelfthAction

UCELLLDR\UlLDRFirstAction
UCELLLDR\UlLDRSecondAction
..
UCELLLDR\UlLDRNinthAction

sade
2015-12-08, 02:11 AM
UCELLINTERRATHOCOV for cell level and UINTERRATHOTCOV for network level.

Inter-RAT CS Measure Start RSCP THD = -104(same as the one below)
Inter-RAT CS Measure Stop RSCP THD = -101(same as the one below)
Inter-RAT CS Measure Start Ec/No THD = -13(higher the value more traffic pushed from 3G to 2G for CS)
Inter-RAT CS Measure Stop Ec/No THD = -11(higher the value more traffic pushed from 3G to 2G for CS)

This will increase your CM attemps but push traffic to 2G from 3G.

B/R

This is good solution when you want to offload CS traffic to 2g unconditionally - no matter what's the load in current 3g cell.
I would say that what operators typically want to achieve is to serve CS users in 3g, but when there is a congestion - offload them to 2g to save 3g resources for PS users (CS users handovered to 2g won't feel such a great quality degradation). For that purpose I would consider LDR better than changing 2d/2f thresholds - of course as long as it is efficient enough

radio_network
2015-12-08, 11:46 AM
Thanks for the above explanation.
Huawei 3g problems:

1. In some of the Sectors/NodeB Huawei, in Auto mode the phone is frequently moving to 2g even near to the sector. Is there any suggestion pl.

2. In some sectors, (using DSP CELLCHK ) uplink load seems to be zero and Noise fig is also observed -70 something diff from -67. Is there any suggestion pl.

sade
2015-12-09, 05:07 AM
Thanks for the above explanation.
Huawei 3g problems:

1. In some of the Sectors/NodeB Huawei, in Auto mode the phone is frequently moving to 2g even near to the sector. Is there any suggestion pl.

2. In some sectors, (using DSP CELLCHK ) uplink load seems to be zero and Noise fig is also observed -70 something diff from -67. Is there any suggestion pl.

Hi,
I think it would be better if you posted those question in new topic, as they're not linked directly to "pushing traffic to 2g" topic. Nevertheless I will try to give you some answer

1. In general, UE is deciding to move between intrafreq/interfreq/interrat based on EcNo and RSCP values. Situation you described may happen in both Idle and Connected mode - principle is the same, but different parameters govern this process.
I will assume that this happens in idle (idle, pch or fach - they're all controlled by the same parameter. Peremeters which control reselections in idle are:
UCELLSELRESEL\qRxLevMin (lets assume -114 dBm)
UCELLSELRESEL\sSearchRat (lets assume 2 dB)
There is a lot of others params which control UE behavior, but since you mention specifically moving to 2g, I have narrowed it down to those 2. So to reselect to 2g assuming values above, your signal would have to drop to -114 dBm + 2 dB = -112 dBm.

EcNo is used rather for intrafreq/interfreq reselection - I don't remember if it can be used also as criteria for interrat measurement - In my opinion event if it can - it shouldnt be used.

So to troubleshoot your case, make some measurements on your phone - see what is the signal level and quality when the phone is switched to 2g. Then troubleshoot why the level or quality is low. If switching to 2g happens in connected - check values of e2d, e2f thresholds in UCELLINTERFEQHOCOV / UCELLINTERRATHOCOV - or on appropriate objects on RNC. After checking parameters make some measurements as well to check if it's quality or level that causes the HO, and troubleshoot the problem accordingly.


2. To help with these one, you should give us some more details. Maybe the following questions will help to narrow it down:
- What's common for sectors having such problems?
- What's the site type? Stack (feeders?) distributed (no feeders, RRU and jumpers used instead?)
- What technologies and frequencies are on those sectors? Do they share feeders? antennas? any other elements?
- What happen when you tilt this sector antenna down or change the azimuth? Do the interference go down?
- Are there any obstacles in front of antennas which could reflect emitted waves?
- In case you share multiple bands/technologies: what happens when you turn down all other cells in the same sector? Do the interference go down?
- Is the interference constant in time, or do they happen and disappear?
- Can you measure the spectrum with any external equipement in this azimuth?