- This topic has 12 replies, 1 voice, and was last updated 11 years ago by Ers.
11th November 2009 at 06:57 #59570ErsGuest
I need to discuss about engineering process needed to optimize TDR from 2% to 0.65%
Is ‘2%’ of TDR showing very very bad network performance? Or maybe any of you have experience worse than it?
Thanks in advance11th November 2009 at 10:10 #59571hichamGuest
It’s not so bad for cells with Hight TA but it’s should be under 1% for urbain cell(and less)11th November 2009 at 10:26 #59572mesferGuest
what do u mean TDR
thanks13th November 2009 at 06:23 #59573ErsGuest
TDR = TCH Drop Rate
hicham ..is it TIming Advance you mean for TA?13th November 2009 at 20:57 #59574PixGuest
as said before, 2% could be OK if the cell has an extended & weak coverage (with poor rxlev in many areas)
urban cells should be better than 1.2%.
Very High TDR (>> 5%) could also be caused by poor hardware performances (abis links, board failures, etc).16th November 2009 at 11:29 #59575ErsGuest
In condition as you mentioned (..the cell has an extended & weak coverage (with poor rxlev in many areas)), is it hard to optimize TDR value from 2% to 0.65% without new site deployment?16th November 2009 at 13:48 #59576PixGuest
Absolutely. If you are certain that the frequencies are clean ( = no interference) and that the coverage area cannot be enhanced (with high power transmitters, or better antennas, or thicker feeders, or TMA, or remove some combiners, etc)
Keep in mind that what you need is to increase the coverage.. There is no miracle :
– increase current cell coverage (see above)
– or, add a repeater
– or, add a BTS
Now, there is a trick. It’s not “good” IMO, but it will make your stats look better. By changing some parameters, you can actually forbid the user with poor RXLEV to establish a call.
Less “low rxlev” calls = less call drops
… but also = less traffic.
the 2 parameters you can use are :
RACH TA FILTER
RXLEV ACCESS MIN
But really, if anyone asks you who told you to do that, don’t mention me. The point of my job is to increase traffic, because traffic= revenues !
Just say it’s an idea from..
Nosfera-pix17th November 2009 at 06:32 #59577ErsGuest
If we do the tricky tweaky :)…would it possibly increase the SD blocking?
Ers17th November 2009 at 10:38 #59578PixGuest
The evil trick will actually decrease the traffic, SDCCH & TCH. Therefore, I don’t foresee an increase in SDCCH congestion.
What you are going to do is *logically* reduce the “cell best serving area”.
Pix17th November 2009 at 11:26 #59579ErsGuest
Just to make sure..
the trick will not effect the stat param of SDSR (SD Success Rate)?17th November 2009 at 12:43 #59580PixGuest
The only thing I can say is that it *should* not decrease QoS, except the TCH traffic and SDCCH traffic.
Of course, there are always small chances that something unpredictable happens. But in your case, I can’t think of anything. So it is a pretty safe trick, as long as you don’t care about traffic.
(sorry to emphasize so much on traffic…)
pix17th November 2009 at 16:34 #59581hichamGuest
Hi and sorry to answer late,
to Reduce ur call drop in such condition (cell with important TA) u have to :
– Check 1st ur HW (in such case most important cell are in rural area and their maintenance is so hard , pylone with more than 40 HBA) : Antenna , conbiner, TRXs ect…check alarms
– check ur frequency (in general in this rural area , there is no problem)
if u don’t have any problem from these 2 u have a planification problem u have to add a new site or repeter (u have to locate the source of ur trafic)
Some astuces (not recommended becuase u reduce ur trafic):
Reduce ur RX_LEV_ACCES
Reduce the BS_POWER_MAX
with these two parameters u are sure that u will have a good KPI but with a important reduction of trafic (be carefull :))))))))
Good Luck18th November 2009 at 07:52 #59582ErsGuest
Pix & Hicham,
Thanks a lot guys,much clear for me.
Wish you both success..always 🙂