CDMA Radio Network Optimization Plot Analysis (using Actix)

[kenshinz]

a good tutorial from mr. pram




In this Post I wanna answer a question from a friend of mine who has commented . Hope this post will useful for us. Sorry if I post a little bit late, because I have a little work in Ketapang
Lets begin.
First I wanna share a little bit about Optimization Process. Optimization process consists of 3 main steps ,i.e.:

1. Preparation (Tools: hardware and software setting, scenario setting)
2. Data Collecting
3. Post Processing and Analysis

Actix is a well known post processing software which is powerful enough to help Optim Engineer analyze the network problem and create the Optimization recommendation. In this post, I just wanna share about Plot analysis (requested by Vikaz via comment and email).
On Actix, there are 3 main standard procedures in Optimization analysis, i.e.:

1. Site Verification (single cell Optimization)
2. Cluster Verification (a group of Cell Optimization: 10 Cell more or less)
3. Network Verification (combination of 2 previous stages to verify whole network)

Let’s go to Plot Analysis.
Actually on CDMA, there are 4 Performance Indicator that we have to analyze:

1. Forward FER
2. Rx Level (Mobile Received Power)
3. Ec/Io
4. Handset Transmit Power (TxPO)

you can read those definition here (http://pram.web.id/blog/2008/03/31/cdma2000-network-optimization-fundamentals/)
Below are vikaz actix plot, that I’ve received from him via email. I assume these Plots below, is an example for site verification because the drive test route mostly on the Sec-3 Ballabh garh site coverage area. CMIIW
Plot Analysis (Site Verification)

Forward FER Plot
http://pram.web.id/blog/wp-content/uploads/2009/01/ffer.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/ffer.jpg)
From visual check, at least there are eight spots with bad FFER. On those spots, voice communication quality isn’t good enough. So we have to find out why it can be occurred on those spots with following plot analysis below.

Rx Level Plot

http://pram.web.id/blog/wp-content/uploads/2009/01/rx-lev.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/rx-lev.jpg)
From Rx Plot above, we can conclude: there are no Rx Level Problem a long the drive test area. We can go to the next plot.

Ec/Io Plot

http://pram.web.id/blog/wp-content/uploads/2009/01/ecio-combined.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/ecio-combined.jpg)


From visual check, we can conclude that there’s a spot which has bad Ec/Io. We have to check Mobile Transmit Plot for deeper analysis.

TxPO Plot

http://pram.web.id/blog/wp-content/uploads/2009/01/txpo.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/txpo.jpg)


Above picture display Mobile Transmit Power along the drive test area. We can see that TxPO is good enough to provide good quality communication.

So… What’s going on with the BAD FFER’s?

Well before we goo deeply in to next steps, I think it’ll be helpful if we discuss about CDMA Signature Table.

[kenshinz]

CDMA Signature Table

According to CDMA performance Indicator (you can read on my previous post here (http://pram.web.id/blog/2008/03/31/cdma2000-network-optimization-fundamentals/)), below I’ve made a little table which is useful to guide us to the next step Analysis.

1. Normal Signature ( my own value, just find another source to compare the value ^_^)


Normal FFER : 0 to 5 (in percentage)
Normal Receive Power (Rx Lev): -30 to -100 (in dBm)
Normal Transmit Gain Adjust (TxGA): depends on site configuration, loading and NOM_PWR Setting
Normal Ec/Io: more than -12 dB
Normal Transmit Power: less than 20dBm
Normal system parameter messages on forward and reverse

2. Abnormal Signature, consists of several condition i.e., as you see in this table below:

http://pram.web.id/blog/wp-content/uploads/2009/01/signature-table.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/signature-table.jpg)

According to the Plot Analysis and Signature Table above, we may conclude that vikaz’s network suffer “Forward Link Interference” although the Ec/Io wasn’t too bad (according to the plot). But we can go deeply to Ec/Io plot analysis.

Ec/Io Plot Deep Analysis

Plot analysis below just an example of deeper Ec/Io Plot analysis. Without mind to give incorrect article, but because of one and another reason I just can provide you these plot to give some illustration about Ec/Io Analysis deeply:

1. Ec/Io PN Combined (different log file)

http://pram.web.id/blog/wp-content/uploads/2009/01/ecio-combined1-800x487.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/ecio-combined1.jpg)

well.. let’s zoom it, so we can see Pilot Pollution area more detail.

http://pram.web.id/blog/wp-content/uploads/2009/01/pilot-polution-800x565.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/pilot-polution.jpg)

Above picture shows us the desired PN on site ID 239 are: PN105, PN273, and PN441.All of these PN’s are potentially interfered by it’s neighbor(s), i.e. :


PN105 is potentially interfered by PN201 and PN051 (in case there are somethig wrong in BTS logical and physical configuration)
PN273 is potentially interfered by PN30.
PN441 is potentially interfered by PN123.


let’s find out which PN’s caused Pilot Polution. Below are Ec/Io per PN analysis.

2. Ec/Io PN105

http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn105.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn105.jpg)

[kenshinz]

3. Ec/Io PN201

http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn201-800x503.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn201.jpg)


From the picture above, we can see that PN201overshoot to PN105 coverage area and cause coverage hole on it. We may purpose some recommendation to repair the physical and logical parameter at PN201sector including Tx_Power, and Antenna Tilting and orientation (vertical and horizontal).

4. Ec/Io PN273

http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn273-800x454.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn273.jpg)



5. Ec/Io PN30

http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn30-800x534.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn30.jpg)

And then.. let’s analyse another site239’s sector.

6. Ec/Io PN441

http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn441-800x488.jpg (http://pram.web.id/blog/wp-content/uploads/2009/01/eciopn441.jpg)



PN123 analysis which has potential to interfere PN441 is skipped, because I don’t have enough data to do that. I’m sorry… at least, from this post we can conclude that Overshooting Sector with certain PN, could cause forward link interference. Forward Link interference trigger bad FFER on the network.

Pilot Polution is a case which is on a certain sector, there are too much PN active, and lack of dominance. This could happen when there are too many sectors overshooted. Pilot Polution could cause drop call, block call, bad FFER. This is the worst scenario problem in CDMA air interface caused by bad Planning, (or) Installation and (or) commissioning.

The Recommendation to fix this Forward Link Interference problem is:


Check the BSS Parameter of the site. If there’s anything wrong such as BTS Tx_Power, Neighbor List, and Closed Loop Power Control Parameter.
Check the Antenna Orientation and Tilting of each site and repair it.
In case there’s a repeater on the coverage, check the Tx and Rx gain on Uplink path.

Maybe enough for this Post, and I hope it will useful for us who want to learn how to Optimize CDMA air Interface. I believe this article is far away from enough, so if there any mistakes.. please correct it together. Thanks. ^_^
ps for Vikaz : thanks for giving me such a nice material, and share it to us. Is there any questions please let me know.

[kaka.enine]

good tutorial ...
btw can yu create on pdf or ppt file?
that's more easier to read n learn i guess ...

thanks for share ...

BR,

-kaka-

[kurdi]

Good analysis mr kenshinz, thank you 4 share knowledge

[da_manole]

http://www.howcdmaworks.com/
a very good site with tutorials we should thank to him(BAXTER)

[da_manole]

you should check from the begining Ec power . the Rx level will cheat you
Rx is bigger during the day (at busy hour for example)and smaller during the night .The Ec (pilot power) it's same all the time.
Ec =Rx+Ec/io( Rx=-65dbm Ec/io=-13 Ec=-65-13=-78)
It's same in wcdma other names

[da_manole]

http://www.4shared.com/file/80983549/9cae936c/signature.html

[Darko]

Good analysis !!!

[justdream]

http://pram.web.id/blog/2009/01/14/radio-network-optimization-plot-analysis-using-actix/

[Bulldogwireless]

Question that i have is , do you use ZK cell testing equipment?And if you do , what should normal FER levels be when doing a drive test so i can get better data? I had a problem with my post processing tool and it crashed so i need to get another, this unit had all of my info for FER levels. What do you suggest?

[pram]

@kenzhin
thanks for share this tutorial here... it will be very helpfull if you turn it to PDF file :)

@de_manole
yup.. I've learnt much from Scott Baxter's literature.

@Bulldogwireless
those log file I've got from vikas, a friend of mine that commented on my blog and send those file to me via email. the normal FER you can see on previous page about CDMA Network optimization fundamental

btw.. for all..
if you visit pram.web.id, please leave a message
and if you have a blog.. I'll be very happy to do link exchange

[anwar]

Im glad My Brother Join This web too.....

sibuk opo kowe kang saiki??Ko ora ana kabare maning??:)


@kenzhin
thanks for share this tutorial here... it will be very helpfull if you turn it to PDF file :)

@de_manole
yup.. I've learnt much from Scott Baxter's literature.

@Bulldogwireless
those log file I've got from vikas, a friend of mine that commented on my blog and send those file to me via email. the normal FER you can see on previous page about CDMA Network optimization fundamental

btw.. for all..
if you visit pram.web.id, please leave a message
and if you have a blog.. I'll be very happy to do link exchange

[kenshinz]

@pram

this is yours ;-). thanks to you. welcome to forum

[krobelux]

Dear All Master
have anyone here used ZTE CDMA BTS equipment, i found a lot of problem with ZTE eq, i have power overload alarm [its effect from too many subs access the BTSes resopurces] and our recommendation is to increase the rating power, carrier power [Cell Power].
my question is
is that will broke ZTE equipment??i,m using CDMA 2000 1x with freq 1900 MHz.

[ithink]

Dear All Master
have anyone here used ZTE CDMA BTS equipment, i found a lot of problem with ZTE eq, i have power overload alarm [its effect from too many subs access the BTSes resopurces] and our recommendation is to increase the rating power, carrier power [Cell Power].
my question is
is that will broke ZTE equipment??i,m using CDMA 2000 1x with freq 1900 MHz.

Krobelux,

ZTE has 2 BTS Types,
for AE BTS, the max Rating power 20 W, CBTS 30/60 W,
you can try to increase maxOverloadPwr with adjust pilotOverloadratio for solution

[ithink]

Dear all,

When i do drivetest packet Data, throughput data Drop, and is detected other PN from other Channel served, whereas we just use 1 channel.
why it can be happen?

[pram]

Dear all,

When i do drivetest packet Data, throughput data Drop, and is detected other PN from other Channel served, whereas we just use 1 channel.
why it can be happen?

maybe you have to check FFER and Ec/Io...
by the way.. when it was occurred? is it on the center of cell or on the edge of cell..

[Armagadon]

Dear all,

When i do drivetest packet Data, throughput data Drop, and is detected other PN from other Channel served, whereas we just use 1 channel.
why it can be happen?

I will try to help you if you get me the answers for the following:

1- Do you mean that a hard handoff occurs to another channel?
2- which drive test software do you use? can you send a sample from the drive test log?
3- can you send a capture for the situation?

[andyDgreat]

very helpful and very useful tutorial
thanks a lot for sharing...

[krobelux]

Krobelux,

ZTE has 2 BTS Types,
for AE BTS, the max Rating power 20 W, CBTS 30/60 W,
you can try to increase maxOverloadPwr with adjust pilotOverloadratio for solution

yah thats right, i think its cannot be a solution, cause if we increase MaxOverload power, without adjust rating and cell power, its never increase the capacity, cause maxoverload power is depend on rating, cell power, and pilot ration.



Dear all,
When i do drivetest packet Data, throughput data Drop, and is detected other PN from other Channel served, whereas we just use 1 channel.
why it can be happen?

ithink, is that CDMA 2000 1x packet data drive test? or EvDO?
is that happen always or just temporary?
drop throudhput data is can be happen cause change value from DRC that its maintain in PCF, what about air interface in that position? is the Ec/Io in good condition?

[ithink]

maybe you have to check FFER and Ec/Io...
by the way.. when it was occurred? is it on the center of cell or on the edge of cell..

FFER and EcIo is good, and unexpectedly move to other CH, (out of PN inc),
in the center of cell

[ithink]

I will try to help you if you get me the answers for the following:

1- Do you mean that a hard handoff occurs to another channel?
2- which drive test software do you use? can you send a sample from the drive test log?
3- can you send a capture for the situation?


1.yes, but other ch is unknown
2.oke, i'll
3.dark green is our chfile:///C:/DOCUME%7E1/PRI%7E1.SUP/LOCALS%7E1/Temp/moz-screenshot-2.jpgfile:///C:/DOCUME%7E1/PRI%7E1.SUP/LOCALS%7E1/Temp/moz-screenshot.jpgfile:///C:/DOCUME%7E1/PRI%7E1.SUP/LOCALS%7E1/Temp/moz-screenshot-1.jpg

[ithink]

yah thats right, i think its cannot be a solution, cause if we increase MaxOverload power, without adjust rating and cell power, its never increase the capacity, cause maxoverload power is depend on rating, cell power, and pilot ration.

or you can increase sch setup th to be 88, but don't forget must less than csh th<softho th<powerup th



ithink, is that CDMA 2000 1x packet data drive test? or EvDO?
is that happen always or just temporary?
drop throudhput data is can be happen cause change value from DRC that its maintain in PCF, what about air interface in that position? is the Ec/Io in good condition?

1x, happen, DRC in EVDO, airinterface is good.

[lejun]

Everyone, share your experience please
With me, i usually analyzer CDMA logfile by some steps:
1. See KPIs:
In CDMA we have 4 essential KPI: Ec/Io, Rx, Tx, Forward FER.
Firstly, we must define the value of these KPI, you can see in this thread.
Or you can see in book: RF200, find in http://www.howcdmaworks.com/
It is a wonderfull book in CDMA optimization network field.
By seeing KPI, we know some problem in our network
In some situation, we can define problem clearly from these KPIs
For example:
Ec/Io poor, Rx poor: coverage problem, then place some sites or change antenna...
Ec/Io poor, Rx good: pilot poluttion, it is a complex problem, especially in a large city.
....
Sometime, we need see some PN in area that have problems and active set PN...
And many other situation, you can see problem and solution in RF200 or in attached file:p
2. See events data
Failed calls, dropped calls, handoff...
3. Find the cause of these events
We must understand signalling messages, critical parameters in these messages, call processing by message... RF200 Please:D
Actix have a powerful function: synchronize events data with signalling message, map... It will help us.
In next post, i will have a example for more clearly
If i have some mistake, please correct it for me
BR

[krobelux]

Dear all, i have a problem with handoff in cdma network, when mobile send PSMM and give Handoff request to BSC. but BSC cannot give a HDM to mobile and make call drop, anyone can give a solution what usually happen in BSC that HDM cannot sent from BSC to Mobile.
thnaks.
fyi
Neigbour List OK
Rx, Ec/Io, Tx, FFER, and RFER [BSC side], good condition
no PN Offset missing..
is there any problem with BSC or BTSes?

[zakzakzak]

http://pram.web.id/blog/2009/01/14/radio-network-optimization-plot-analysis-using-actix/


nice site, good info insode

[kenshinz]

please refer this for RF200 by scottbaxter, pdf file


http://www.finetopix.com/showthread.php?p=40207#post40207

[aosan]

Friend,
:eek:The thresholds you wrote here is not the right one for good network KPI
the following will maintain KPI in good situation

Normal FFER : <2% for Voice and <5% for data
Normal Receive Power (Rx Lev): <=-30 & >=-95 (in dBm)
Normal Ec/Io: >= -11 dB
Normal Transmit Power: <= 0dBm

[atilkurt]

Thanx for great post

[honhungoc]

Everyone, share your experience please
With me, i usually analyzer CDMA logfile by some steps:
1. See KPIs:
In CDMA we have 4 essential KPI: Ec/Io, Rx, Tx, Forward FER.
Firstly, we must define the value of these KPI, you can see in this thread.
Or you can see in book: RF200, find in http://www.howcdmaworks.com/
It is a wonderfull book in CDMA optimization network field.
By seeing KPI, we know some problem in our network
In some situation, we can define problem clearly from these KPIs
For example:
Ec/Io poor, Rx poor: coverage problem, then place some sites or change antenna...
Ec/Io poor, Rx good: pilot poluttion, it is a complex problem, especially in a large city.
....
Sometime, we need see some PN in area that have problems and active set PN...
And many other situation, you can see problem and solution in RF200 or in attached file:p
2. See events data
Failed calls, dropped calls, handoff...
3. Find the cause of these events
We must understand signalling messages, critical parameters in these messages, call processing by message... RF200 Please:D
Actix have a powerful function: synchronize events data with signalling message, map... It will help us.
In next post, i will have a example for more clearly
If i have some mistake, please correct it for me
BR

its a good website but its impossible to get the tutorials

[da_manole]

its a good website but its impossible to get the tutorials

http://www.howcdmaworks.com/drophelp2.htm

[ithink]

welldone, deep analize,

just add, you can try to decrese pilot gain, from 225 to 222, or you can check Cell radius, maybe too big for urban area, ( 384 or 512) :)