UL Load Factor (%)

[_lisi4king]

Hello, did anybody calculate UL Load Factor for 3G Huawei?
I've found 2 formulas (depends, which CAC algorithms do i use on my Network):

a) UL load factor = 1 - (cell background noise/UL RX total wideband power)

I can get RTWP (from m2000), but cell background noise from the system - Background noise[0.1dBm] = 61 (constant). For idle cell UL load factor will be 1-6,1/-104=(1+0,06)*100%= ???? And what do we have.....UL Load more than 100%???


b) The ratio of the current number to the maximum number of UL equivalent users to measure the load of the cell. Maximum number is a constant = 80, but current i cen get only from Trace viewer. It is not a sollution.

Do anybody know, how to calculate UL Load Factor%?

[qaqa]

Hello, did anybody calculate UL Load Factor for 3G Huawei?
I've found 2 formulas (depends, which CAC algorithms do i use on my Network):

a) UL load factor = 1 - (cell background noise/UL RX total wideband power)

I can get RTWP (from m2000), but cell background noise from the system - Background noise[0.1dBm] = 61 (constant). For idle cell UL load factor will be 1-6,1/-104=(1+0,06)*100%= ???? And what do we have.....UL Load more than 100%???


b) The ratio of the current number to the maximum number of UL equivalent users to measure the load of the cell. Maximum number is a constant = 80, but current i cen get
only from Trace viewer. It is not a sollution.

Do anybody know, how to calculate UL Load Factor%?

Hi
Correct you formula for Backgroud noise as
Back_noise=-112+(Step-1)*0.1. in your case it will be Back_nose=-112+(61-1)*0.1=-106 dBm.
now to calculate your UL utilization you must first convert the dBms for backnoise and RTWP to Watts and do the maths. we don't devide or multiply dB or dBm :-) remember that :p

2- for your second formula, you can get the EUN from the counters. look for the BH only or get the Mean of the max because you want to look in the worst case.

[_lisi4king]

Hi
Correct you formula for Backgroud noise as
Back_noise=-112+(Step-1)*0.1. in your case it will be Back_nose=-112+(61-1)*0.1=-106 dBm.
now to calculate your UL utilization you must first convert the dBms for backnoise and RTWP to Watts and do the maths. we don't devide or multiply dB or dBm :-) remember that :p

2- for your second formula, you can get the EUN from the counters. look for the BH only or get the Mean of the max because you want to look in the worst case.

Thanks a lot for response, but can you clarify for me some issues from your explanation:

1) Im i right, at first i need to calculate Back_noise in accordance with formula Back_noise=-112+(Step-1)*0,1 than convert it to Watts, the next step -> convert RTWP to Watts and than divide Back_noise(W)/RTWP (W)

2) Can you explain, where did you get this formula? May be you have description? Back_noise=-112+(Step-1)*0,1

3) Did you work with Huawei equipment? Did you see counter which count EUN, because i really can not find it :(.

Thanks for your help! :handshake

[agenov]

Hello, did anybody calculate UL Load Factor for 3G Huawei?
I've found 2 formulas (depends, which CAC algorithms do i use on my Network):

a) UL load factor = 1 - (cell background noise/UL RX total wideband power)

I can get RTWP (from m2000), but cell background noise from the system - Background noise[0.1dBm] = 61 (constant). For idle cell UL load factor will be 1-6,1/-104=(1+0,06)*100%= ???? And what do we have.....UL Load more than 100%???


b) The ratio of the current number to the maximum number of UL equivalent users to measure the load of the cell. Maximum number is a constant = 80, but current i cen get only from Trace viewer. It is not a sollution.

Do anybody know, how to calculate UL Load Factor%?


To be more precisely you have counters and real time traces for both. Does anyone have doc/info about relation between RTWP and ENU??? THis information will be very important and appreciated.

BR
Alex

[_lisi4king]

To be more precisely you have counters and real time traces for both.

BR
Alex

This is not suitable for me. I need to have UL load factor for all cells inside cluster (more than 100). And if we speack about BH, i need to have one value for day for every cell, for Month. If i will use real time trace, i will waste 1 year for this job :rolleyes:

[qaqa]

Hi
see my comments in Red


Thanks a lot for response, but can you clarify for me some issues from your explanation:

1) Im i right, at first i need to calculate Back_noise in accordance with formula Back_noise=-112+(Step-1)*0,1 than convert it to Watts, the next step -> convert RTWP to Watts and than divide Back_noise(W)/RTWP (W)==> yes

2) Can you explain, where did you get this formula? May be you have description? Back_noise=-112+(Step-1)*0,1 If you have the LMT check command ADD CELLCAC and look for Back_noise, from the explanation you can derive the formula

3) Did you work with Huawei equipment? Did you see counter which count EUN, because i really can not find it :(. if you have the HW performance decument look "ALGO.Cell"--> VS.RAC.DL.EqvUserNum

Thanks for your help! :handshake

[_lisi4king]

Hi
see my comments in Red

;) Thanks a lot.

[_lisi4king]

Hi
see my comments in Red

Today i've checked your responce and i it is for me still is not clear:

1) -112 - what this? Because i've found only Background noice abnormal threshold (-100dBm). We use RAN 11.

2) And the second quastion is about VS.RAC.DL.EqvUserNum. I did not find it in M2000. May be it appeared in RAN 12?

[gprastomo]

Hi convert the dbm to mW or watt before calculating the formula

[_lisi4king]

Hi convert the dbm to mW or watt before calculating the formula

What do you mean? :rolleyes:

[_lisi4king]

Today, i've get information that -112dBm this is NodB sensetivity. :rolleyes:

[agenov]

Today i've checked your responce and i it is for me

2) And the second quastion is about VS.RAC.DL.EqvUserNum. I did not find it in M2000. May be it appeared in RAN 12?


Until RAN12 it is called VS.RAC.DL.TotalTrfFactor.

BR

[_lisi4king]

Before RAN12 it is called VS.RAC.DL.TotalTrfFactor.

BR

Thanks, i've found it,but i need UL :p. Another step will be to campare 2 methods of calculating of % UL Load Factor , and compare it.

I don't like method with User number, because we play with MaxEqvUserNumber - it helps to reduse power congestion on some high loaded cell. But on another hand if i will increase MaxEqvUserNumber - UL load Factor will decrease....looks strange.

For example current ULEqvUserNum - 40, Max 80 %UL load Factor - 50%. But if i set Max to 120 - %UL load factor will be 33%?

If i will use RTWP, it will be more usefull, because RTWP will show real load on cell.

[agenov]

Thanks, i've found it,but i need UL :p. Another step will be to campare 2 methods of calculating of % UL Load Factor , and compare it.

I don't like method with User number, because we play with MaxEqvUserNumber - it helps to reduse power congestion on some high loaded cell. But on another hand if i will increase MaxEqvUserNumber - UL load Factor will decrease....looks strange.

For example current ULEqvUserNum - 40, Max 80 %UL load Factor - 50%. But if i set Max to 120 - %UL load factor will be 33%?

If i will use RTWP, it will be more usefull, because RTWP will show real load on cell.


Ok just change DL to UL in the counter name and you got it. Also if you want it so much you can change how your CAC algorithm to work and use based on RTWP i.e this is CAC algorithm type 2.

BR

[_lisi4king]

Ok just change DL to UL in the counter name and you got it. Also if you want it so much you can change how your CAC algorithm to work and use based on RTWP i.e this is CAC algorithm type 2.

BR

I need not to change algorithm type. I need to calculate %UL load factor for all cell inside one cluster. :victory:

[agenov]

Thanks, i've found it,but i need UL :p. Another step will be to campare 2 methods of calculating of % UL Load Factor , and compare it.

I don't like method with User number, because we play with MaxEqvUserNumber - it helps to reduse power congestion on some high loaded cell. But on another hand if i will increase MaxEqvUserNumber - UL load Factor will decrease....looks strange.

For example current ULEqvUserNum - 40, Max 80 %UL load Factor - 50%. But if i set Max to 120 - %UL load factor will be 33%?

If i will use RTWP, it will be more usefull, because RTWP will show real load on cell.


This is interesting. Based on what/how you do calculation of how many ENUs you need to add in order to avoid congestion, please share some info if you have experience

br

[_lisi4king]

This is interesting. Based on what/how you do calculation of how many ENUs you need to add in order to avoid congestion, please share some info if you have experience

br

On our network we have some cells, which cover hostels. It means that in one place we have more than 100 users in BH and of course high PS RAB Failure due power congestion. We already have done everything what we can - we have added second carrier (we can not add more than 2 :rolleyes:), we increased tilt (to avoid extra users), but congestion still exists.

But, when we have changed ULEqvUserNum from 80 (default Huawei value) to 120 -140 - congestion becames near 0. Of cause we can guess, that if we increase ULEqvUserNum we loose something else. But I don't know exactly -what. And on another hand we can increase user perception and KPI's. :cool:

[agenov]

On our network we have some cells, which cover hostels. It means that in one place we have more than 100 users in BH and of course high PS RAB Failure due power congestion. We already have done everything what we can - we have added second carrier (we can not add more than 2 :rolleyes:), we increased tilt (to avoid extra users), but congestion still exists.

But, when we have changed ULEqvUserNum from 80 (default Huawei value) to 120 -140 - congestion becames near 0. Of cause we can guess, that if we increase ULEqvUserNum we loose something else. But I don't know exactly -what. And on another hand we can increase user perception and KPI's. :cool:


I did the same. the question is what the relation between ENU and RTWP is? do u have some info?

[_lisi4king]

I did the same. the question is what the relation between ENU and RTWP is? do u have some info?

I don't think, that they have direct relation (ENU and RTWP). I mean, that you have only 3 algorithm types. 2 of them based on RTWP and TCP. 1 - based on ENU. As i understand Huawei has his own experience, how many ENU corresponds to each service. In accordance with algorithm - Eqvuser (total)+Eqvuser (new) should be less than MaxULEqvUserNum.....otherwise admission will be rejected.

[_lisi4king]

Today, i've finished calculation for all cells inside one cluster with granularity -30 min (i use formula with RTWP and Back_noice). Result looks good. Values from 0 to 93%. But one problem, some of them has negative value. It means that RTWP lower than Back_noise. Is this normal? What do you think?:rolleyes:

[arisondang]

Today, i've finished calculation for all cells inside one cluster with granularity -30 min (i use formula with RTWP and Back_noice). Result looks good. Values from 0 to 93%. But one problem, some of them has negative value. It means that RTWP lower than Back_noise. Is this normal? What do you think?:rolleyes:

What do you mean back_noise? Is it Noise Floor or (Thermal noise+ Noise figure)?
If yes, i don't think its possible RTWP is lower than Back Noise..

As I know, RTWP= Noise Floor + Noise Rise
so RTWP must be higher than Noise Floor..

If you want to know relation between EUN and RTWP,
you have to calculate UL load per user, and then calculate UL Load Cell based on RTWP, and you can get max EUN number..

Thanks

[_lisi4king]

What do you mean back_noise? Is it Noise Floor or (Thermal noise+ Noise figure)?
If yes, i don't think its possible RTWP is lower than Back Noise..

As I know, RTWP= Noise Floor + Noise Rise
so RTWP must be higher than Noise Floor..

If you want to know relation between EUN and RTWP,
you have to calculate UL load per user, and then calculate UL Load Cell based on RTWP, and you can get max EUN number..

Thanks

I mean, formula from Huawei HedEx

UL load factor = 1 - (cell background noise/UL RX total wideband power)

And cell background noise - -112+(Step-1)*0,1, Step - this is value from settings. On my Network - it is 6,1 dBm

[arisondang]

I mean, formula from Huawei HedEx

UL load factor = 1 - (cell background noise/UL RX total wideband power)

And cell background noise - -112+(Step-1)*0,1, Step - this is value from settings. On my Network - it is 6,1 dBm

Basicly,
your formula comes from basic UL Load Formula = 1- Pn/Itotal...
(please check in book WCDMA for UMTS page 261,edited by Harri Holma)

Pn= Background and Receiver Noise= Noise Thermal+Noise Figure= Noise Floor
Itotal=Total Received Power= RTWP= Pn+Iown+Iother= Noise Floor + Noise Rise

So
RTWP must be HIGHER THAN Background noise

Thanks

[rimoucha]

Hello Expert,

Could you please tell me how to calculate noise rise based on RTWP (in dBm)?
Do you think that we should consider the Maximum or the mean value for RTWP per cell ?

Best regards,

[tukangoptim]

Thanks, i've found it,but i need UL :p. Another step will be to campare 2 methods of calculating of % UL Load Factor , and compare it.

I don't like method with User number, because we play with MaxEqvUserNumber - it helps to reduse power congestion on some high loaded cell. But on another hand if i will increase MaxEqvUserNumber - UL load Factor will decrease....looks strange.

For example current ULEqvUserNum - 40, Max 80 %UL load Factor - 50%. But if i set Max to 120 - %UL load factor will be 33%?

If i will use RTWP, it will be more usefull, because RTWP will show real load on cell.
i have found many many cases where rtwp increased with no relation to load at all. thats why huawei prefers to apply enu.
anyway, that is really interesting your UL load reduced to 33% due to that parameter ?? :loveliness: :loveliness:

[jayk]

thats not correct formula to calculate uplink load, _lisi4king has already mentioned that..and in fact that was an example just for making people understand..

just to correct you in both algorithms, RTWP effects the network same way. ENU Algorithm has many parameters to control effect of high RTWP but not algorithm1. thats why ENU algorithm is used in uplink CAC