PDA

View Full Version : How to Improve RF quality in GSM hopping network?



bsnlrf
2011-11-01, 06:04 PM
Hello,

Can any one help me with poor RF quality in hopping network during Traffic hours. How can I reduce that...

We are using non cyclic hopping, with HSN, 1-63

We are using 14 BCCH and 17 TCH for hopping with MAIO

Sector 1 (Alpha) 0-6-12,
Sector 2 (Beta) 2-8-14
Sector 3 (Gamma) 4-10-16

We have maximum 4 TRx configuration per sector. With 1 TRX dedicated for BCCH and is non hopping.

I have found that RF quality is poor in traffic hours. I don't understand how to improve RF quality.

We regularly perform Drive TEST. Our BCCH planning is also very nice with no co-channel interference and very minimal adjacent channel interference (That adj. ch interference to because of available BCCH freq. are just 14). I have also observed that quality is bad irrespective of any specific location. We are using synthesized hopping.

In clean n/w with less Traffic at early morning and late night RF quality is significantly good, but as traffic grows Rf quality goes bad. I have also observed that Cells having 10-13 erlang with 4 TRX also have bad RX Quality.

This problem is not for any particular site, it is evenly observed in all city sites.

Is there any parameter I can play with..

Please help me out.

mohyedeen_alkousy
2011-11-01, 08:05 PM
sometime i have similar issues if the Tilting is not well optimized
if u have too much overlapping between neighboring with strong level this will increase collisions
the fact u have mentioned about bad quality at busy hours may confirm this theory

did u try SFH add hoc planning using Atoll or Asset , i think it will give better results

gageloz
2011-11-01, 09:47 PM
Hi bsnlrf ;

In busy hours we also faced this kind of bad quality problems during the drive tests in hot spot areas.And i think it is mostly related with the overlapping of the cells. If your area is flat and if you dont have coverage hole you shoul tilt the cells to overcome the overlapping. For example if you ar using antennas with the 15 degree vertical beamwidth you shoul give the tilt at least 8 degree.:)

jan74
2011-11-01, 10:05 PM
Hello,

Can any one help me with poor RF quality in hopping network during Traffic hours. How can I reduce that...

We are using non cyclic hopping, with HSN, 1-63

We are using 14 BCCH and 17 TCH for hopping with MAIO

Sector 1 (Alpha) 0-6-12,
Sector 2 (Beta) 2-8-14
Sector 3 (Gamma) 4-10-16

We have maximum 4 TRx configuration per sector. With 1 TRX dedicated for BCCH and is non hopping.

I have found that RF quality is poor in traffic hours. I don't understand how to improve RF quality.

We regularly perform Drive TEST. Our BCCH planning is also very nice with no co-channel interference and very minimal adjacent channel interference (That adj. ch interference to because of available BCCH freq. are just 14). I have also observed that quality is bad irrespective of any specific location. We are using synthesized hopping.

In clean n/w with less Traffic at early morning and late night RF quality is significantly good, but as traffic grows Rf quality goes bad. I have also observed that Cells having 10-13 erlang with 4 TRX also have bad RX Quality.

This problem is not for any particular site, it is evenly observed in all city sites.

Is there any parameter I can play with..

Please help me out.

You can consider reducing your number of BCCHs even further to increase the number of hopping channels and thereby reducing collisions. If the network is relatively clean, you should not have too many problems. You have 31 channels and use 14 for BCCH. I have worked on a network that had 36 channels and used only 12 for BCCH. You could start by going to 12 BCCHs, giving you 19 hopping. If your network is not very heavily loaded, the BCCHs will hardly be used even in the busy hour so Adjacent interference should not be a big deal.

Also, are you using Half Rate? If so, this will add to your problems.

jayk
2011-11-01, 11:11 PM
I will suggest to go for physical optimization downtilts/azimuth changes to reduce overlapping/overshoots. Do check your power control parameters & DTX.
More aggressive power control can also help to reduce interference and improve rx qualitiy.

D33T0X
2011-11-02, 01:49 AM
Reduce overlap - tilt/reduce power where you can.

mail2paturi
2011-11-02, 02:04 AM
Hello,

Can any one help me with poor RF quality in hopping network during Traffic hours. How can I reduce that...

We are using non cyclic hopping, with HSN, 1-63

We are using 14 BCCH and 17 TCH for hopping with MAIO

Sector 1 (Alpha) 0-6-12,
Sector 2 (Beta) 2-8-14
Sector 3 (Gamma) 4-10-16

We have maximum 4 TRx configuration per sector. With 1 TRX dedicated for BCCH and is non hopping.

I have found that RF quality is poor in traffic hours. I don't understand how to improve RF quality.

We regularly perform Drive TEST. Our BCCH planning is also very nice with no co-channel interference and very minimal adjacent channel interference (That adj. ch interference to because of available BCCH freq. are just 14). I have also observed that quality is bad irrespective of any specific location. We are using synthesized hopping.

In clean n/w with less Traffic at early morning and late night RF quality is significantly good, but as traffic grows Rf quality goes bad. I have also observed that Cells having 10-13 erlang with 4 TRX also have bad RX Quality.

This problem is not for any particular site, it is evenly observed in all city sites.

Is there any parameter I can play with..

Please help me out.

Dear Bro,

Physical optimization is one point,
apart from them you can do the following exercise
1. Check if any BCCH and BSIC clashes
2. Do power check if more reduce the trx power
3. Check the HSN clashes
4. Check stats for TRX wise and identify wthr all trx are having the same issue or any particular trx is giving that issue.
5. Down tilts are must required to avoid overshooting and overlapping
6.AMR codecs can be fine tuned kindly check which coding scheme you are using
7. plan different BSIC for each cell if possible.
8. Try to avoid - in 1 particular LAC there should be no same BCCH, BSIC combination means no two cells in metro sites should not have same BCCH frequency and BSIC.

srilasitha
2011-11-02, 01:26 PM
if you balance traffic by configuring multiple layers in the same frequency band rx quality would be poor.

In 1X1 hopping serving cell should have best SS with at least 3db higher to the best neighbor. otherwise RX quality would be poor.

check DTX, Power control parameters and the physical site modifications as well

bsnlrf
2011-11-02, 07:47 PM
sometime i have similar issues if the Tilting is not well optimized
if u have too much overlapping between neighboring with strong level this will increase collisions
the fact u have mentioned about bad quality at busy hours may confirm this theory

did u try SFH add hoc planning using Atoll or Asset , i think it will give better results


Hey thanks for the reply.

You are right, we are having overlapping between neighbouring with strong Level. Now the city I am talking about have Antenna placed on top of very high buildings. Also, Antennas are not having Electrical tilt provision. Lowering of antenna is also not possible because it is not feasible. Mechanical Tilt we give max 4 degree because sites are very much near to each other so I don't want any back lobe issue.

Any good Ideas about physical optimization for such situation?

We are using Atoll as planning tool. So, can you tell me what is SFH add hoc and how to plan it using Atoll?

bsnlrf
2011-11-02, 07:58 PM
Hi bsnlrf ;

In busy hours we also faced this kind of bad quality problems during the drive tests in hot spot areas.And i think it is mostly related with the overlapping of the cells. If your area is flat and if you dont have coverage hole you shoul tilt the cells to overcome the overlapping. For example if you ar using antennas with the 15 degree vertical beamwidth you shoul give the tilt at least 8 degree.:)


Thanks for the reply...

You are right, we are having overlapping between neighbouring with strong Level. Now the city I am talking about have Antenna placed on top of very high buildings. Also, Antennas are not having Electrical tilt provision. Lowering of antenna is also not possible because it is not feasible. Mechanical Tilt we give max 4 degree because sites are very much near to each other so I don't want any back lobe issue.

Any good Ideas about physical optimization for such situation?

Is 8 degree Mechanical tilt is good practice, antenna we are using don't have Electrical tilt provision? Please explore. Also tell me How much tilt are good as per vertical beamwidth of antenna.

bsnlrf
2011-11-02, 08:06 PM
You can consider reducing your number of BCCHs even further to increase the number of hopping channels and thereby reducing collisions. If the network is relatively clean, you should not have too many problems. You have 31 channels and use 14 for BCCH. I have worked on a network that had 36 channels and used only 12 for BCCH. You could start by going to 12 BCCHs, giving you 19 hopping. If your network is not very heavily loaded, the BCCHs will hardly be used even in the busy hour so Adjacent interference should not be a big deal.

Also, are you using Half Rate? If so, this will add to your problems.


Thanks for the reply. Earlie we were using 12 BCCH. But we had same problem at that time also. At that time we were having problem in Adjacent Channel interference in planning. So, I was the among one who advocate 14 BCCH in network.

You said Adjacent Channel Interference is not big deal. How much does Adjacent Channel Interference affect in network. I read in many documents that Co & Adjacent Channel Interference are having adverse affect on network.

You are also right that we are using Half rate in network. How much do you think is a heavy load. I mean to say you are saying heavy load as per TRX or sector or LAC or BSC.

bsnlrf
2011-11-02, 08:23 PM
I will suggest to go for physical optimization downtilts/azimuth changes to reduce overlapping/overshoots. Do check your power control parameters & DTX.
More aggressive power control can also help to reduce interference and improve rx qualitiy.


Hi Thanks for the reply.

As I have mentioned in my earlier replies about limitations of physical optimization. So, I not repeating them here again.

As far as PC parameter and DTX is concerned can you guide me about that. If you wish I can send you our parameter setting of our network.

bsnlrf
2011-11-02, 08:25 PM
Reduce overlap - tilt/reduce power where you can.

Thanks for the reply.

Can you suggest me how reduced power affect the network. I have feeling that if I will reduce power indoor coverage may get affected. So, how to balance both the things.

bsnlrf
2011-11-02, 08:35 PM
Dear Bro,

Physical optimization is one point,
apart from them you can do the following exercise
1. Check if any BCCH and BSIC clashes
2. Do power check if more reduce the trx power
3. Check the HSN clashes
4. Check stats for TRX wise and identify wthr all trx are having the same issue or any particular trx is giving that issue.
5. Down tilts are must required to avoid overshooting and overlapping
6.AMR codecs can be fine tuned kindly check which coding scheme you are using
7. plan different BSIC for each cell if possible.
8. Try to avoid - in 1 particular LAC there should be no same BCCH, BSIC combination means no two cells in metro sites should not have same BCCH frequency and BSIC.


I have already mentioned physical optimizations limitations in my earlier mail. So please go through them.

Respone of your points:
1. Check if any BCCH and BSIC clashes

No such cases.

2. Do power check if more reduce the trx power

Our TRX power is about 46 dBm. Can you guide me how reduced power affect the network. I have feeling that if I will reduce power indoor coverage may get affected. So, how to balance both the things.

3. Check the HSN clashes

What is best HSN plan. There is no same HSN in nearing sites. How much HSN difference is preferable in neighbouring sites. We generally use difference of 5. I have read somewhere that we can even use next HSN in neighbouring site. Please clarify.

4. Check stats for TRX wise and identify wthr all trx are having the same issue or any particular trx is giving that issue.

This problem is irrespective of any TRX. It is common in all city's sites.

5. Down tilts are must required to avoid overshooting and overlapping

Please guide me, as per distance between 2 sites, how much tilts is advisable. We have no provision of Electrical tilt in Antenna.

6.AMR codecs can be fine tuned kindly check which coding scheme you are using

We are using AMR. I am not aware of Codec scheme. But I will let you know that later.

7. plan different BSIC for each cell if possible.

We already do that practice.

8. Try to avoid - in 1 particular LAC there should be no same BCCH, BSIC combination means no two cells in metro sites should not have same BCCH frequency and BSIC.

We take care of that thing and there is no such case.

mail2paturi
2011-11-02, 08:59 PM
I have already mentioned physical optimizations limitations in my earlier mail. So please go through them.

Respone of your points:
1. Check if any BCCH and BSIC clashes

No such cases.

2. Do power check if more reduce the trx power

Our TRX power is about 46 dBm. Can you guide me how reduced power affect the network. I have feeling that if I will reduce power indoor coverage may get affected. So, how to balance both the things.



3. Check the HSN clashes

What is best HSN plan. There is no same HSN in nearing sites. How much HSN difference is preferable in neighbouring sites. We generally use difference of 5. I have read somewhere that we can even use next HSN in neighbouring site. Please clarify.

4. Check stats for TRX wise and identify wthr all trx are having the same issue or any particular trx is giving that issue.

This problem is irrespective of any TRX. It is common in all city's sites.

5. Down tilts are must required to avoid overshooting and overlapping

Please guide me, as per distance between 2 sites, how much tilts is advisable. We have no provision of Electrical tilt in Antenna.

6.AMR codecs can be fine tuned kindly check which coding scheme you are using

We are using AMR. I am not aware of Codec scheme. But I will let you know that later.

7. plan different BSIC for each cell if possible.

We already do that practice.

8. Try to avoid - in 1 particular LAC there should be no same BCCH, BSIC combination means no two cells in metro sites should not have same BCCH frequency and BSIC.

We take care of that thing and there is no such case.

1.Our TRX power is about 43 dBm
2. As per your inputs HSN plan is fine
3. Down tilt depends on the type of antenna , if there is no electrical tilt option i hope you are using 90 deg beam width antenna. so you can give a tilt of 6 - 8 it depends again on inter distance between sites