Question related to optimization UMTS cells borders

[Mr_Piongly]

Hi guys,

I heard about the existence of different possible parametrization for WCDMA cells that are border of a GSM network in order to optimize the cell reselection ( and the ISHO also ?) 3g-> 2G when a user leave the umts coverage ( for example headways from town where there are many umts sites to outside in country direction in which there are only gsm sites network)
Someone has made experience in this kind of scenario ?
2 different set:
1)set: of umts internal cells sites in town (which parametrization ?)
2)set: of cells that point in direction outside the town (which parametrization ?)
Some suggestions?
I hope that someone could make light on this subject
Many thanks

[TZUN TZU]

Hi,

Personally I would tune the IRATHO parameters on a cell by cell basis based on DCR and type of drop.

Coverage ul/dl limited drops are where I would target my areas of concern and then tune 2d/6a/6b/3a parameters accordingly.

Also consider how much traffic you are going to lose to 2g when reducing 3g footprint

[boring]

In theory, ISHOs at network edge are mainly triggered by RSCP whereas within coverage by EcNo.

Typical value for RSCP = -105, but i have used up to -103 for some difficult areas (faster activation which however pushes traffic to 2G so it must be used carefully).

EcNo of -12 to -14 for within coverage handovers seems to work. You should bear in mind again that within coverage ISHO happen because of excessive interference (coverage is fine). Nonetheless, you will very often find that ISHOs are triggered because of coverage as well (e.g. in building). in this case RSCP of -105 is OK.

thanks

[TZUN TZU]

Its also worth considering motorway cells where UE speed effects the ability of power control creating deeper fades, coverage can be lost faster thereby creating a need to start the HO process earlier.

For exact values it depends what your starting value is and how much traffic you accept you will lose to 2g. I also agree with Boring, EcNo should be looked at for city centres where as out of town would probably be more likely RSCP.

Decide if the problem is interference or coverage or both and tune accordingly. In most operators its a simple change so you can tune easily until you meet the criteria of DCR vs Traffic loss.

[tukangoptim]

my network uses global -101 dBm for RSCP doing fine <0.5% RNC wise.
anyway, back to topic.., i dont think tuning ISHO threshold will affected much your performance nor decrease dcr for instance.
its not worth the effort, and looking at the trend where 3G coverage expanded with new sites coming. Considering as well signal fluctuation in 3G due to its near-far effect.


In theory, ISHOs at network edge are mainly triggered by RSCP whereas within coverage by EcNo.

Typical value for RSCP = -105, but i have used up to -103 for some difficult areas (faster activation which however pushes traffic to 2G so it must be used carefully).

EcNo of -12 to -14 for within coverage handovers seems to work. You should bear in mind again that within coverage ISHO happen because of excessive interference (coverage is fine). Nonetheless, you will very often find that ISHOs are triggered because of coverage as well (e.g. in building). in this case RSCP of -105 is OK.

thanks

[cococrunch]

I think this will be an issue of which site is a better server in an area. If the UMTS cell edge is already congested compared to 2G cells in the area, then I think you should trigger ISHO a bit earlier so the 2G cells can handle and vice versa..