How to deal with uplink interference in UMTS ?

[yassanino]

hi all , i wish to get a help about some drop i have i tried tres faster but nothing in RTWP i found -85 wich means a serious uplink interference , wich action should i take to correct the cell in question

many thanks in advance

[tremalnike]

You have to identify main reason for RTWP - interference(constant in time) or load(rssi peak in BH),or wrong nodeB configuration(constant for day and night,check attenuation parameters,delay and etc.).
Which vendor you use?

[yassanino]

we rae using NSN vendor equipement

[RFHumam]

if its external interferance try to switch to another frequency, otherwise try to smallen the CPICH for this site in order to serve less pepole which means less internal interferance (consider not taking pepole out of coverage )

[vaampeere]

Hi,

Best is to use spectrum analyzer and find the source by means of triangulation. However, if equipment is not available then you can try any internal FFT analysis software tool to check overlap of any spectrum on the site by blocking the BRD.

[yassanino]

it appears that due to MHA ( not compatible and commisssioned like "other equipements" ) on some cells , untill now solution is to do a different parameters for sectores holding MHA

[dtvt2]

Pls check RTWP in non Busy hour. If it is high. there will be external interference

[yassanino]

thanks i'll check hourly the cells

[brianm]

Check if there are no missing neighbours or some subscriber with an illegal cell extender

In one town we had a building where people go for prayers and they had installed cell phone jammers to block calls during prayers. The regulatory officials ahd to ask them to switch them off as they were interfering with 3G and 2G signal

[samuraial]

If the rtwp is -85 all the time it is external interference most of the time.
In order to improve user experience you should find the source and avoid it. If not you can play with RTWPINITADJUST that can be found at nodeb level MML. Setting it to -150 will add -15dbm more and in total you will go -100dbm.
User experience will increase and throughput also.
Just be careful to monitor it as if the external interference will varnish you need to set RTWPINITADJUST 0 again.
Use automatic background noise also. Cell level.
Run dsp ucellchk command(RNC level) before and after to check the changes.


B/R

[yassanino]

ok will try it and comeback with a report if improvement thanks for the hint samuraial (http://www.finetopix.com/member.php?15890-samuraial) really appereciated.

[Optimustron]

Hi Yassanino, all the advices from the engineers are ok, however I recommend you a little method in order to reduce the possibilities:


Check the behaviour of the interference, is all the time, is hourly, random, it is from the beginning or it appears suddenly, etc.
Check the surrounding place, i.e., the cell is isolated because of buildings, is the las cell of the coverage, is in building, etc.
Check if other sectors have RTWP too, specially those in front of the affected node, but you need to check too if the cosite sectors are interfered and if the are how much is te RTWP of each one, for example if sector A is your worst sector and the sector B is the next and C is less or almost doesn't have interference, maybe the interference is between sectors A & B but near to the center or A.
Shut down the affected sector in order to see the other affected sector, if any, if the interference in the other sectors go away then you have a problem in the affected sector, probably INTERMODULATION, IM/IMD, or sometimes in some manufacturers the NODE B works bad because it generates interference and you need to reset it, believe me it happen to us.
You can make some tests like changing the downtilt of the site and see at that time if the RTWP changes, you can change the azimuth too and you can locate in this way the direction of the interference, of course if the antenna is directional, if the antenna is omnidirectional, almost all the time the arey small then you can change by a directional model in order to look for interference.
If you are seeing the RTWP in real time you can move/shake a little the jumpers, gentle tap the antenna in the back, in order to see if the RTWP get worst if that is true you have a defective component, and of course don't stand in front or at any of the sides of the antenna, because this changes the RTWP stand behind it.
Check the sourronding place of the antenna specially near to it, because you can have metal surfaces or structures that are producing IM/IMD, if you paint the antenna check that is the right paint because some of them have plumb and this is a source of IM/IMD. At this point you can measure the isolation of the antenna.
Check with an SA or better if you have an RTSA, and if you have an PIM TESTER its better. By the way with the SA/RTSA you must to check the neighbors of the site with an omni antenna and make a grid an locate the place were the interferece is higher then you need to locate more precisely with a yagui antenna by moving to diferent places between the zone that the interferece is higher and make a triangulation.


I hope this can help you, best regards,

[absoares]

The reason for high RSSI might be due Internal or External causes: Internal Causes:  RBS Internal Interference  Faulty or loose components Cable, connectors and/or diplexers  Faulty Antenna/TMA – Check alarms and if the TMA is set to Active.  Incorrect parameter settings – check attenuation delays and gains.  Incorrect installation External Causes:  GSM Cross Band Interference – Make sure that no GSM cells is transmitting in the GSM band.  External interference – Check the horuly stats to see if there is a pattern, ie: during working hours. Also try changing the tilts between max and min range to see how it impacts the RSSI. If it is external interference the RSSI might change depending if the interference source is near or far from the site. A dummy load at the antenna would can also be used to track external interference. When doing the tests trace the RBS stats for RSSI don’t use RNC stats.

[electron]

Hi,

there is a chance some how noise floor lock (thermal noise estimation mechanism) does not work properly. E/// uses sampling method during 24 hour then gives a noise floor value and UL RSSI accordingly will be calculated. Instead of allowing RBS calculate it automatically use a tip to disable this mechanism and give constant value and observe the result. This trick can be used in case of phenomena that even in off load hours also UL RSSI is high .


Cheers

[Ralphobafemi]

For just one sector,
you may use your electric tilt to simulate a better antenna pattern void of this interference, reduce e-tilt by 2.
if you don't have an improvement, azimuth tuning of +/-10/20 will do.
If it affects multiple cells, you will need to do geographical scanning of the band with a tool.


Hope this helps
RalphObafemi
+ClickThanksPlease