Professional Documents
Culture Documents
B10 Dimensioning Methods
B10 Dimensioning Methods
in B10
Eugen Marza
December 2008
Agenda
Overview
Assessment of
traffic values for
SDCCH, TCH and
PDCH channels
Assessment of
basic and bonus
Abis nibbles from
TRX and BTS
configuration
TC
Assessment of
CS and PS traffic
over AterMux
MT120
SMU
speech
MT120
Air
SMU
A-bis
CS
traffic
TRX 1
TRX 2
PS
traffic
MxBSC
TCH
TCH
G C H B asic
M -E G C H lin k 1
TRX 3
M -E G C H lin k 2
TRX n
M -E G C H lin k n
D yna m ic
A b is
a lloc a tion
B TS
CCP
board
CCP
board
SSW
board
TP
board
LIU
board
LIU
board
G C H B asic
G C H B on u s
G CH Extra
G CH Extra
CS
CS
A-ter mux
CS+ PS
MxMFS
GP board
PS
data
SGSN
GP board
Evaluate the
required number
of GPU/GP
boards per BSC
Gb
Evaluate the
required number
of Gb 64K TS
per GPU
Target
Find the needed number of RTCH and TRX for one Cell
Input (counters values):
CS Traffic
PS Traffic
SDCCH Traffic
CS Congestion
PS Congestion
SDCCH Congestion
Customer requirements for blocking probability
Nb of SDCCH channels
Nb of TCH
Nb of TRX
Real blocking
Output:
METHOD
OUTPUT
Erlang B
Nb of required
SDCCH sub-channels
/ timeslots
Required
SDCCH Traffic
GoS:
% SDCCH blocking
%SDCCH _ cong =
MC 400
3600
MC 04
100 %
MC 04 + MC148
CS_FR service
CS_HR service
P62x (x=a-c),
P438c, P507,
P91x (x=a-f),
Required Traffic
PS_DL service
getMax
PS_UL service
P505.
KaufmannRoberts
Algorithm
Nb RTS
needed for
CS and PS
traffic
CS_Real_Co ng =
RTCH_Assig n _ Cong
MC812
=
RTCH_Assig n _ Request MC 812 + MC 703
MC 380 a
CS _ Cong
MC 380 a + MC 380 b
HR _ Cong =
MC 380 b
CS _ Cong
MC 380a + MC 380 b
Re quired _ FR _Traffic =
FR _ Successful _Traffic
MC380acell
=
1 FR _ Cong
3600 ( 1 FR _ Cong )
Re quired _ HR _Traffic =
HR _ Successful _Traffic
MC380bcell
=
1 HR _ Cong
3600 ( 1 HR _ Cong )
P 451b
3600
Measured _ UL _ PS _ traffic =
P 451a
3600
P14
100 %
P 91a + P 91b + P 91c + P 91d + P 91e + P 91f + P 505
Re quired _ PS _ traffic =
P 27
100 %
P 62 a + P 62 b + P 62 c P 438 c + P 507
Measured _ PS _ traffic
1 Min(%TBF _ radio _ cong , 30%)
Final value for Required PS traffic is Max between required DL and UL PS traffic.
Final aggregation:
Total nb of RTCH =
= 1 TS for BCCH + 1 TS for CCCH (if case) +
+ Nb of Required SDCCH TSs +
+ Nb of Required TSs for FR, HR and PS,
coming from Kaufmann-Roberts algorithm.
Output:
DL PS Throughput
UL PS Throughput
d PS _ DL [ kbps ] =
Data_DL
=
Transmision_Time_DL
m
n
d
=
1000 P 38e
d PS _ UL [ kbps ] =
Data _ UL
=
Transmisio n _ Time _ UL
m
n
d
=
1000 P 38f
Abis Interface
Dimensioning
Target
Find the required number of ExtraAbis TS
Check RSL congestion
Methods are applied at BTS level
Based on Abis allocation priorities
MAX_PDCH_High_Load Zone
Basic nibbles
of cell
Priority N1
Priority N3
Priority N2
Extra + Bonus
nibbles of BTS
PS Traffic
Extra-Bonus Traffic
Available Extra-Bonus nibbles
PS Congestion (Abis)
Customer requirements for QoS (delay and quantile)
Customer requirements for number of basic and bonus nibbles
Output:
Nb of extra Abis nibbles
Nb of ExtraAbisTS
INPUT
Nb of required
extra & bonus
Abis Nibbles
Required extra
& bonus Abis
nibble Traffic
Erlang C
GoS:
% Quantile of x
delay sec
OUTPUT
METHOD
Nb of bonus Abis
Nibbles
Nb of required
extra Abis Nibbles
Note:
The counters retrieved at cell level must be aggregated per BTS.
P 466
3600
P105 i
100 %
P 91a + P 91b + P 91c + P 91d + P 91e + P 91f + P 505
P105 j
100 %
P 62a + P 62b + P 62c P 438 c + P 507
%TBF _ Abis _ cong = Max (%DL _ TBF _ Abis _ cong , %UL _ TBF _ Abis _ cong )
GCH Method
(Method 2)
INPUT
Nb of required
Abis Nibbles
Required Data
Traffic on Abis
Erlang C
GoS:
% Quantile of x
delay sec
OUTPUT
METHOD
Nb of basic &
bonus Abis Nibbles
Nb of required
extra Abis Nibbles
GCH Method
Needed counters for Data traffic:
P100c or GAAGCHUST (GPRS_transmission_GCH_busy_time)
Note:
The counters retrieved at cell level must be aggregated per BTS.
GCH Method
Abis congestion:
P105 i
100 %
P 91a + P 91b + P 91c + P 91d + P 91e + P 91f + P 505
P105 j
100 %
P 62a + P 62b + P 62c P 438 c + P 507
%TBF _ Abis _ cong = Max (%DL _ TBF _ Abis _ cong , %UL _ TBF _ Abis _ cong )
GCH Method
Measured traffic:
Required traffic:
P100c
3600
GCH Method
Abis Dimensioning
Finally, at BTS level for Abis dimensioning
Total Number of Abis TS =
= 2*Total nb of TRX +
+ Nb of TS for RSL and OML (depending on MCB type)
+ Nb of required Extra Abis TS,
and
Number of required Abis Links =
= Roundup[(Total Number of Abis Ts)/31].
Formula:
LAPD efficiency [%] = L1.1/( L1.1+L1.15)*100
BSC
Dimensioning
Target
Find the total CS traffic
Find the related CS signaling traffic
Decide what solution is required for signaling traffic: LSL or HSL
Check the HDLC occupancy to decide what type of TP GSM board is required:
TP GSM V1 or TP GSM STM1 (V3)
Find the required number of AterMux CS links
Find the total PS traffic
Find the required number of AterMux PS links
Find the required number of GPU/GP boards
CS Traffic
SS7 Traffic
PS Traffic
AterMux PS congestion
Customer requirements for CS and SS7 blocking and QoS for PS
Nb
Nb
Nb
Nb
Nb
Output:
of
of
of
of
of
MC380a + MC380b
3600
The required value of TCH traffic must be below the BSC Erlang capacity
(depending on BSC configuration), else the BSC capacity must be extended if
possible by adding one CCP board.
103
If the required number of SS7 TSs is below or equal to 16, the LSL solution
may be used.
If the required number of SS7 TSs is above 16, the HSL solution is mandatory.
41 | Presentation Title | Month 2008
UserLabel
MuxRate
MuxRule
{ amecID 1, moiRdn 1}
G62B011
{signallingLoadBts high}
statisticMux64
G62B083
{signallingLoadBts high}
G62B271
Total #
TRE
Total #
TRX DR
Total #
TRX FR
Total #
HDLC
statisticMux64
{signallingLoadBts dummy}
statisticMux64
G62B122
{signallingLoadBts normal}
statisticMux64
12
11
11
G62B091
{signallingLoadBts dummy}
noMux
G62A12_3
{signallingLoadBts dummy}
staticMux
10
10
from AlcatelBtsSiteManager.csv
42 | Presentation Title | Month 2008
Total #
LAPD
from AlcatelBts_Sector.csv
P100c
3600
P 383a
100 %
3600
P 384
100 %
3600
Nb of Required GCH
Nb of Re quired GPU/GP board =
Max ( P 41, P 42 )
28800
At least 2 GSLs are recommended to be defined per GPU due to security reason.
Up to 4 GSLs can be defined per GPU.
Ater CS
Dimensioning
Target
Find the required number of AterMux CS links
Method: ErlangB
Input (counters values):
CS Traffic
SS7 Traffic
Customer requirements for CS and SS7 blocking and QoS for PS
Output:
Nb of SS7 connections, links and real blocking
Nb of TCH, CS dedicated links and real blocking
Method
Find the max number of Ater CS links between TCH requirements and
signaling requirements (SCCP)
INPUT
METHOD
OUTPUT
Signalling Traffic
Required
SDCCH
Traffic
Erlang B
Nb of
required SS7
per BSC
GoS:
% SS7
blocking
Choose
Max
value
User Traffic
Required
TCH
Traffic
GoS:
% TCH
blocking
Nb of
required
AterMUX-CS
links per
BSC
Erlang B
Nb of
required
TCH per BSC
Nb of
required
AterMUX-CS
links per
BSC
Final nb of
required
AterMUX-CS
links per
BSC
CS Traffic
Method for TCH
Measured _TCH _ traffic =
MC380a + MC380b
3600
The number of required AterMux CS links can be obtained by dividing the number of
AterMux CS channels obtained from the dimensioning process by the AterMux CS link
capacity.
The total capacity of n AterMUX CS links is:
111TCH(link#1) + 111TCH(link#2) + 116TCH(link#3) + 116TCH(link#4) +
116TCH(link#5) + 116TCH(link#6) + 115TCH(link#7) + 115TCH (link#8) + 116TCH(link
#9) + 116TCH(link #10) + 116TCH(link #11)+ + 116TCH(link #N),
where N =< MaxNbAterMuxCS.
53 | Presentation Title | Month 2008
CS Signaling Traffic
Method for SCCP
Measured SCCP Traffic = SCCP traffic in Erlang during Busy Hour =
= ((MC400/MC390)*(MC01+MC02+MC10) + (MC380a+MC380b))/3600
103
CS Signaling Traffic
Method for SS7 based on volume of data transferred
Counters for N7 at link level (Type 9):
N3.1 (NB_N7_SIF_SIO_SENT)
Number of octets of Signalling Information (SIF) and Service Information Octets
(SIO) transmitted on the N7 SL.
N3.3 (NB_N7_MSU_SENT)
Number of Message Signalling Units (MSU) transmitted on the N7 SL.
N3.4 (NB_N7_SIF_SIO_REC)
Number of octets of Signalling Information Field (SIF) and Service Information
Octets (SIO) received on the N7 SL.
N3.5 (NB_N7_MSU_REC)
Number of Message Signalling Units (MSU) received on the N7 SL.
N3.10 (NB_N7_MSU_DISC_DUE_CONG)
Number Message Signalling Units (MSU) discarded due to a message buffer overflow
caused by an extended period of Signalling Link congestion.
CS Signaling Traffic
Method for SS7 based on volume of data transferred
Traffic evaluation in UL
Counter values must be aggregated for the link set
Measured _ SS 7 _ traffic =
%SS 7 _ cong =
N 3. 1 + 6 * N 3 . 3
28800000
N 3.10
100 %
N 3.3 + N 3.10
Re quired _ SS 7 _ traffic =
Measured _ SS 7 _ traffic
1 Min (%SS 7 _ cong ; 30%)
CS Signaling Traffic
Method for SS7 based on volume of data transferred
Traffic evaluation in DL
Counter values must be aggregated for the link set
Measured _ SS 7 _ traffic =
N 3 . 4 + 6 * N 3 .5
28800000
The final number of SS7 TSs (links) is the max value between the results for
UL and DL.
If this number is greater than 16, HSL usage is mandatory.
CS Signaling Traffic
Method for SS7 based on volume of data transferred
N7 link efficiency based on couters (in UL):
N3.3 (NB_N7_MSU_SENT)
N3.10 (NB_N7_MSU_DISC_DUE_CONG)
Formula:
N7 efficiency [%] = N3.3/( N3.3+N3.10)*100
CS Signaling Traffic
Method for SS7 link load based on volume of data transferred
LSL case
%N7 Utilization in UL = (N3.1 + (6 * N3.3)) / (3600*8,000)
%N7 Utilization in DL = (N3.4 + (6 * N3.5)) / (3600*8,000)
Explanation:
MSU contains 6 bytes (fix part) + SIF&SIO (variable part).
SIF is between 2 and 34 bytes.
SIO is 1 byte.
Note:
1 hour of usage of 1 E1 Time Slot (for N7) = 1hour * 64Kbps =
= 3600s * 8000 bytes/s = 28,800 Kbytes/h
Warning:
The load on one N7 link shall not exceed 40% (recommended).
CS Signaling Traffic
Method for SS7 link load based on volume of data transferred
HSL case
In B10, if HSL is actived, N7 can have 2 different frame formats:
Normal frame format
Extended frame format
Parameter: MTP_Sequence_Number_Format
(MtpSequenceNumberFormat in ACIE AlcatelManagedElementComplex.csv)
Range / default value: (0=Normal, 1=Extended) / 0
Note:
In case of Normal frame format
MSU contains 6 bytes (fix part) + SIF&SIO (variable part).
CS Signaling Traffic
Method for SS7 link load based on volume of data transferred
HSL case with Normal frame format
%N7 Utilization in UL = (N3.1 + (6 * N3.3)) / (3600*8,000*31)
%N7 Utilization in DL = (N3.4 + (6 * N3.5)) / (3600*8,000*31)
Explanation:
32 TSs of 64Kbps are available for one HSL link.
Note:
1 hour of usage of 1 E1 Time Slot (for N7) = 1hour * 64Kbps =
= 3600s * 8000 bytes/s = 28,800 Kbytes/h
For 32 TS per HSL link we get
3600s * 8000 bytes/s * 31TS = 892,800 Kbytes/h
Warning:
The load on one N7 link shall not exceed 40% (recommended).
CS Signaling Traffic
Method for SS7 link load based on volume of data transferred
HSL case with Extended frame format
%N7 Utilization in UL = (N3.1 + (9 * N3.3)) / (3600*8,000*31)
%N7 Utilization in DL = (N3.4 + (9 * N3.5)) / (3600*8,000*31)
Explanation:
32 TSs of 64Kbps are available for one HSL link.
Note:
1 hour of usage of 1 E1 Time Slot (for N7) = 1hour * 64Kbps =
= 3600s * 8000 bytes/s = 28,800 Kbytes/h
For 32 TS per HSL link we get
3600s * 8000 bytes/s * 31TS = 892,800 Kbytes/h
Warning:
The load on one N7 link shall not exceed 40% (recommended).
Ater PS
Dimensioning
Target
Find the required number of AterMux PS links
Method: ErlangC
Input (counters values):
PS Traffic
Number of Available GCH
Number of Max Busy GCH
AterMux PS congestion
Customer requirements for PS QoS
Output:
Nb of GCH
Nb of GSL links per BSC and PS real quantile
Nb of PS links per BSC and PS real quantile
Method
Find the max number of Ater PS links between GCH requirements and
signaling requirements (GSL)
INPUT
OUTPUT
METHOD
Signaling Traffic
Required
GSL
Traffic
Erlang C
Nb of
required
GSL per BSC
QoS:
quantile
delay
Choose
Max
value
User Traffic
Required
GCH
Traffic
Erlang C
Nb of
required
GCH per BSC
QoS:
quantile
delay
Nb of
required PS
links per
BSC
Nb of
required PS
links per
BSC
Final nb of
required
AterMUX-PS
links per
BSC
P100c
3600
P 383a
100 %
3600
P 384
100 %
3600
28
Max ( P 41, P 42 )
28800
GSL _ load =
It is recommended to increase the number of GSL per GPU if GSL load is greater than
80%.
The final number of required AterMux PS links is the maximum value between
the number of links needed for signaling traffic (GSL) and for user traffic (GCH).
GPU/GP
Dimensioning
Target
Find the load and the required number of AterMux PS links per GPU/GP board
Method: ErlangC
Input (counters values):
PS Traffic
MS context average
MS context max
DSP congestion
GSL traffic
Average & Max CPU load
BSS failures rate
Output:
GPU/GP load
Check if GPU/GP board is not overloaded
INPUT
METHOD
MS_context_handling (= 0/1)
User Traffic
Required
GCH
Traffic
QoS:
quantile
delay
Erlang C
Needed
GCH
GCH
capacity
of GPU
Power_Limitation (= 0/1)
OUTPUT
Number
of GPU
boards
GPU/GP load
GPU/GP board limitation control
GPU limitation
PMU
PTU
PPC/CPU
DSP
Power Limitation Memory Limitation Power Limitation Memory Limitation
P76a
P392a
P201 (thr_1 )
Dimensioning
P77a
P384
P392b
P202 (thr_2 )
indicators
=> P402 (thr )
QoS
P105e
UL TBF estab
P203
P105c
indicators
P105f
BSS Failure
P204
P105d
(TBF establ)
CPU Cong
BSS Fail
DSP Load
MFS parameters:
Thr = PMU_CPU_Overload (Default=95%)
Thr_1 = DSP_Load_Thr_1 (Default=85%)
Thr_2 = DSP_Load_Thr_2 (Default=95%)
DSP Cong
GPU/GP load
User traffic load (GCH)
Needed counters:
P100c or GAAGCHUST (GPRS_transmission_GCH_busy_time),
P383a or GQAGALCTT (GPRS_GPU_Ater_cong_time),
P384 or GQRGPUCDT (GPRS_GPU_DSP_cong_time),
P201 or GTRGPULDLT (GPRS_DSP_CPU_load_time),
P202 or GTRGPULDOLT (GPRS_DSP_CPU_overload_time),
GQRDTELDR (GPRS_DL_TBF_estab_fail_dsp_load_rate)
or P203 / (P91a + P91b + P91c + P91d + P91e + P91f + P505)
GQRUTELDR (GPRS_UL_TBF_estab_fail_dsp_load_rate)
or P204 / (P62a + P62b + P62c - P438c + P507)
GPU/GP load
Signaling load
Needed counters:
P392a or GTRGPUMCA (GPRS_MS_idle_context_avg),
P392b or GTRGPUM (GPRS_MS_idle_context_max),
P76a or GTRGPUPBA (GPRS_PMU_CPU_power_budget_avg),
P77a or GTRGPUPBM (GPRS_PMU_CPU_power_budget_max),
P402 or GTRGPUOT (GPRS_PMU_overload_time),
GQRDTECCR (GPRS_DL_TBF_estab_fail_cpu_cong_rate)
or P105e / (P91a + P91b + P91c + P91d + P91e + P91f + P505)
GQRUTECCR (GPRS_UL_TBF_estab_fail_cpu_cong_rate),
or P105f / (P62a + P62b + P62c - P438c + P507)
GPU/GP load
Required number of GPU/GP boards based on user traffic
Measured _ GCH _ traffic =
%GCH _ Ater _ cong =
%DSP _ load =
P100c
3600
P 383a
100 %
3600
P 384
100 %
3600
P 402
x100 %
3600
%GCH _ cong = Max(% Ater _ cong ,%DSP _ cong ,%DSP _ load ,%CPU _ overload )
Re quired _ GCH _ traffic =
GPU/GP load
Required number of GPU/GP boards based on user traffic
Nb of Required GCH
Nb of Re quired GPU/GP board =
where the GPU board capacity is 480 GCH and the GP board capacity is 1560 GCH.
The result must be compared with the existing number of GPU/GP boards per BSC.
If the required number of GPU/GP boards is less than the existing one, there is no
need for change.
If the required number of GPU/GP boards is greater than the existing one, then the
number of GPU/GP boards per BSC must be increased by one unit.
For a BSC in multi GPU configuration, the method (GCH traffic) may be applied per
each GPU/GP board separately to check the load on each board.
GPU/GP load
Required number of GPU/GP boards
GPU power limitation from PTU (DSP/CPU)
If
MAX(P201;P202)/3600 > 0.1%
and
dsp_load_fail_rate = MAX(GQRDTELDR; GQRUTELDR) > 1%
then
PTU_Power_Limitation = 1
else
PTU_Power_Limitation = 0
GPU/GP load
Required number of GPU/GP boards
GPU power limitation from PMU (CPU)
If
P402/3600 > 0.1% or P76a > 70%
and
cpu_cong_fail_rate = MAX (GQRUTECCR; GQRDTECCR) > 1%
then
PMU_Power_Limitation = 1
else
PMU_Power_Limitation = 0
GPU/GP load
Required number of GPU/GP boards
GPU memory limitation from PMU (CPU)
If P392b = 1000 and P392a > 900
for MxMFS
then
MS_context_handling = 1
else
MS_context_handling = 0
GPU/GP load
Final result
INPUT
OUTPUT
METHOD
Signaling Traffic
Required
GSL
Traffic
Erlang C
Nb of
required
GSL per BSC
QoS:
quantile
delay
Choose
Max
value
User Traffic
Required
GCH
Traffic
Erlang C
Nb of
required
GCH per BSC
Nb of
required
GPU boards
per BSC
QoS:
quantile
delay
Nb of
required PS
links per
GPU board
Nb of
required PS
links per
GPU board
Final nb of
required
AterMUX-PS
links per
GPU
INPUT
OUTPUT
METHOD
Signaling Traffic
Required
GSL
Traffic
QoS:
quantile
delay
Erlang C
Nb of
required
GSL per BSC
Nb of
required
GPU boards
per BSC
Nb of
required
GSL links
per GPU
Max ( P 41, P 42 )
28800
GSL _ load =
GPU
K_GSL
BSC
GSL messages
buffer
87 | Presentation Title | Month 2008
Nb_PS_paging/s+ Nb_CS_paging/s
2.5
3.5
x Nb_TBF_Data_req/s
(TBF req)
per TBF_data_req
Available
High
Low
High
3.5
Low
2.5
3.5
GCH
If Ater congestion observed
The values provided in the table above are relibale if T_GCH_INACTIVITY = 3s and
T_GCH_INACTIVITY_LAST = 20s (default setting).
Otherwise (I.e. T_GCH_INACTIVITY = 2s and T_GCH_INACTIVITY_LAST = 6s) GSL load will be
greater due to the consequent increase of deallocation/allocation messages.
89 | Presentation Title | Month 2008
GSL _ load =
GSL load in terms of treated messages per second on a given GPU/GP should not
exceed 75%
It is recommended to increase the number of GSL per GPU/GP if GSL load is greater
than 75%.
If the result is greater than 4 (8), we must increase the number of GPU (GP)
boards with 1, and redo the calculations.
The final number of required AterMux PS links per GPU/GP boards is the
maximum value between the number of links needed for signaling traffic
(GSL) and for user traffic (GCH).
Gb Interface
Dimensioning
Output:
Nb of Gb TS
INPUT
Required Gb
Traffic
QoS:
% Quantile of x
delay sec
METHOD
Erlang C
OUTPUT
Nb of required
Gb Timeslot per
link
max( P 45,P 46 )
28800
The maximum number of Gb links from one GPU/GP board to the SGSN is 8.
Maximum 31 Gb TS (TS no. 1 to 31) can be configured per one Gb link.
Tc (Measurement interval) = 1s
ACCESS_RATE_BC
EIR
EBS
NIR
CIR
CIR = (CBS*8)/Tc
NIR (Normal Information Rate)
CBS
NIR CIR
EBS (Excess Burst Size)
Tc
time
Example:
Parameter
value
unit
Tc
CBS
124
kbyte
CIR
992
kbps
NIR
992
kbps
EBS
kbyte
EIR
1008
kbps
AC
1008
kbps
NbGbTS
16
number
Example:
For CBS = 0; CIR = 0 and NIR = 0:
Parameter
value
unit
Tc
EBS
160
kbyte
AC
1280
kbps
EIR
1280
kbps
NbGbTS
20
number
Gb over IP
Using Gigabit Ethernet link to the aggregation Router, the max load is:
For MxMFS: 1560 GCHperGP * 16 Kbps = 24.96 Mbps
Assuming all this traffic coming from SGSN for one MxMFS with 21 GP boards we get:
21 boards*24.96 Mbps = 524.16 Mbps Half of the GigaEthernet link capacity
Congestion is not expected at the level of the link to the aggregation router.
Any Congestion seen afterwards is dependent on bandwidth management of the
core IP network.
Gb over IP
Target
Find the DL and UL traffic during BH
Input (counters values):
Ip Endpoint downlink data volume
Ip Endpoint uplink data volume
Output:
DL and UL data rate
Gb over IP
Data rate calculation in DL (SGSN to MFS) and in UL (MFS to SGSN)
Data rate DL/UL, at IP Endpoint level or aggregated at GP or MFS level:
DL Data rate [Kbps] = 1.233*(P45a*8 )/(3600-P34a)
UL Data rate [Kbps] = 1.233*(P46a*8)/(3600-P34a)
Explanation:
Assuming LLC PDU and BSSGP payload are 300 bytes, NS, UDP, IP and Fast Ethernet
overhead are 70 bytes (i.e. 4+28+38),
SNS layer (where Kbytes counts is done) overhead ratio at this layer is 0.23% (i.e.
70/300).
Gb over IP
GboIP traffic estimation based on actual GboFR traffic
Estimated Traffic:
Estimated DL IP Data rate [Kbps] = (370/306)*(P45*8)/(3600-P34)
Estimated UL IP Data rate [Kbps] = (370/306)*(P46*8)/(3600-P34)
Explanation:
Assuming LLC PDU and BSSGP payload are 300 bytes, NS, UDP, IP and Fast Ethernet
overhead are 70 bytes (i.e. 4+28+38), NS and FR overhead are 6 bytes (i.e. 4+2).
Packet size increase ratio = Total IP packet (i.e. 370) / Total FR packet (i.e. 306)
Location Area
Dimensioning
CCCH Sharing
Air interface
Pagings sent in DL on CCCH
Sharing between PCH & AGCH within one CCCH M-51 frame (235 mS
duration)
CCCH_CONF
Combined mode
3 blocks
Non combined mode
9 blocks
Nb of AGCH blocks = BS_AG_BLKS_RES
CCCH 1
CCCH 2
CCCH 3
CCCH 4
CCCH 5
CCCH 6
CCCH 7
Example of AGCH and PCH organisation when BS_AG_BLKS_RES = 3 (Non combined mode)
CCCH 8
PCH content
Air interface
Nb of pagings per CCCH block, depends on
Paging request type
TYPE 1
TYPE 2
- IMSI
- (P)TMSI, (P)TMSI, (P)TMSI
- (P)TMSI
- (P)TMSI, (P)TMSI,IMSI
- IMSI,IMSI
- IMSI, (P)TMSI
- (P)TMSI, (P)TMSI
TYPE 3
- (P)TMSI, (P)TMSI, (P)TMSI, (P)TMSI
Paging efficiency
between 1.5 3
2.5 pagings / PCH block considered usually in dimensioning
PCH content
Paging request
message
(type 1, 2 or 3)
MS
PCH blocks
on CCCH
Paging command
Message
(RSL)
Cell
TC
MSC
AterMUX-CS
BTS
Paging request
or
Packet downlink
assignment
message
CS Paging
(SS7)
Gs
Abis
BSC
Paging command
or
Immediate assign
command message
(RSL)
AterMUX-PS
(GSL)
PS/CS Paging
or
Channel assignment
downlink message
Gb
SGSN
MFS
GPU
PCH capacity
Non-Combined cell
BS_AG_BLKS_RES = 4 (AGCH blocks)
PCH capacity
Combined cell
BS_AG_BLKS_RES = 1
3 - 1 = 2 PCH blocks
45 957 paging/hour
12.76 paging/s
AGCH content
Access grant
message
(IA, IAE or IAR)
AGCH blocks
MS on CCCH
Immediate assign
command message
(RSL)
Immediate assignment
or
Immediate assignment
reject message
BTS
Cell
Access grant
message
(PUA, PDA or IAR)
Abis
BSC
Immediate assign
command message
(RSL)
AterMUX-PS
(GSL)
Channel assignment uplink
(EGPRS Uplink assign.),
Channel assignment downlink
(Downlink assign.)
or
Channel assignment uplink
message (Assign. reject)
MFS
GPU
Note: IA= Immediate Assign; IAE=Immediate Assign Extended; IAR= Immediate Assign Reject;
PUA= Packet Uplink Assignment; PDA= Packet Downlink Assignment.
112 | Presentation Title | Month 2008
AGCH capacity
Non-Combined cell
BS_AG_BLKS_RES = 4
4 AGCH blocks/Multiframe x (3600s / 235 ms) =
61 272 AGCH blocks/ hour
1 Immediate Assign message / Block x 61 272 Blocks =
61 272 paging/hour at 100% load
If 60 % engineering limit applied:
36 763 IA message/hour
10.21 IA message/s
AGCH capacity
Combined cell
BS_AG_BLKS_RES = 1
1 AGCH block/Multiframe x (3600s / 235 ms) =
15 318 AGCH blocks/ hour
1 Immediate Assign message / Block x 15 318 Blocks =
15 318 paging/hour at 100% load
If 60 % engineering limit applied:
9 180 IA message/hour
2.55 IA message/s
PCH
blocks
6
5*
4**
AGCH
blocks
3
4
5
Paging
Request/s
(PCH
blocks/s)
25.53
21.27
17.02
Access
Grant
msgs/s
(AGCH
blocks/s)
12.75
17.02
21.27
Paging
msgs/s at
60% limit
38.29
31.91
25.53
ImmAs
msgs/s at
60% limit
7.65
10.21
12.76
PCH
msgs/h at
60% limit
137851
114893
91908
ImmAs
msgs/h at
60% limit
27540
36763
45936
Total #
msgs
sent/h at
60% limit
165391
151656
137844
B9 limitations
Abis interface limitation
max paging load of 33 paging commands per second = 118 800 paging/hour
G2 BSC limitation
max paging load of 70 paging commands per second = 252 000 paging/hour
MxBSC limitation
max paging load of 95 paging commands per second = 342 000 paging/hour
2 Location Areas
with an average of
132 cells per LA
LA
LA
Abis
33 paging/s
G2 BSC
(max 264 cells)
A
70 paging/s
MxBSC:
3 Location Areas
with an average of
88 cells per LA
LA
LA
Abis
33 paging/s
Mx BSC
(max 264 cells)
LA
A
95 paging/s
B10 limitations
Abis interface limitation
max paging load of 33 paging commands/second 118 800 paging/hour
max paging load of 66 paging commands/second 237 600 paging/hour with mCCCH
G2 BSC limitation
max paging load of 70 paging commands/second = 252 000 paging/hour
MxBSC limitation
max paging load of 120 paging commands/second= 432 000 paging/h
(expected in B10)
B10 limitations
CCCH capacity in B10 with mCCCH
(rounded figures)
Nb CCCH TS BS_AG_BLKS_RES PCH blocks
1
1
1
2
2
2
2
3
4
5
3
4
5
6
6
5
4
12
10
8
6
2 Location Areas
with an average of
132 cells per LA
LA
LA
Abis
33 paging/s
G2 BSC
(max 264 cells)
A
70 paging/s
LA
Abis
66 paging/s
G2 BSC
(max 264 cells)
A
70 paging/s
4 Location Areas
with an average of
125 cells per LA
LA
LA
LA
LA
Abis
1 CCCH
33 paging/s
Mx BSC
(max 500 cells)
A
120 paging/s
LA
LA
Abis
2 CCCH
66 paging/s
Mx BSC
(max 500 cells)
A
120 paging/s
Dimensioning Target
Verify if the Location Area size is OK
Method:
Check the BSC and Abis limits
Check the related parameters
Check the load on CCCH (PCH, AGCH, RACH)
Output
Location area load
Yes
No
Check Um Limitation
No
No
Yes
No
MC8a >114,891
OK
Yes
Change to non-combined
Re-Design LA
OK
Re-Design LA
Yes
No
Check Um Limitation
No
No
Yes
No
Yes
MC8a >114,891
OK
Yes
Change to non-combined
Re-Design LA
OK
Re-Design LA
LA
RA
Note:
Total PCH channel capacity is 191 485 paging/hour for 1 CCCH TS and
382 970 paging/hour for 2 CCCH TSs (for BS_AG_BLKS_RES = 4).
(CS+PS) Paging load = MC8a
60% represents engineering limit
129 | Presentation Title | Month 2008
RA
RA1
RA2
LA
LA
LA1
LA2
Tuning
If PCH load is below 60% and no congestion but AGCH load is above 60% and with
congestion => BS_AG_BLKS_RES may be increased by 1.
If PCH load is above 60% and with congestion but AGCH load is below 60% and without
congestion => BS_AG_BLKS_RES may be decreased by 1.
www.alcatel-lucent.com
www.alcatel-lucent.com