RF Parameter & Optimization Document

Brian - CommServ Radio Frequency Management & Optimization
Education Division
GSM
GSM
RF Parameter and Optimization
RF Parameter and Optimization
Education Division
Course Objectives Course Objectives
Understand the principal of optimization and types of problems
experienced in a GSM network and why they occur.
Understand the variety of different tools available to the optimisation
engineer and how to analyze and apply the parameter.
Develop and explain an optimisation process for GSM networks.
Identify suitable KPIs which could be used to highlight poorly performing
cells.
Use statistics to identify performance problems in the BSS and methods
of rectifying these problems.
How to identify service affecting BSS database and neighbour relation
issues.
Dimension GSM networks with a view to Multi-Service deployment.
Understand the characteristics of NSS Topology and Signalling.
Know how non-BSS issues can affect the network and be identified.
Education Division
1 Day
1 Day
Course Outline Course Outline
1. Review of GSM Principles
2. Network Characteristics & Problem Types
3. Introduction to Performance Management
4. Performance Management Metrics
5. Measuring Network Performance - Drive Testing
6. Measuring Network Performance The OMC
7. The Optimisation Process
8. BSS Optimisation Activities
9. NSS Characteristics
10. NSS Parameter Evaluation and Optimisation
11. Dimensioning Network for new Services
Education Division
Section 1
Section 1
Review of GSM Principles
Review of GSM Principles
Education Division
Section 1
Section 1
1.2 Physical Channel Structure 1.2 Physical Channel Structure
1.6 Timing Advance 1.6 Timing Advance
1.3 Logical Channel / Multiframe Formats 1.3 Logical Channel / Multiframe Formats
1.7 Cell Reselection 1.7 Cell Reselection
1.9 Power Control 1.9 Power Control
1.8 Handovers 1.8 Handovers
1.5 Paging Procedures 1.5 Paging Procedures
1.4 Review of the BCH Carrier 1.4 Review of the BCH Carrier
1.10 Frequency Hopping 1.10 Frequency Hopping
1.1 Introduction 1.1 Introduction
Education Division
P P- -GSM Physical Channels GSM Physical Channels
Uplink Downlink
890 915 935 960 MHz
Duplex spacing = 45 MHz
Guard Band
100 kHz wide
Channel Numbers (n) (ARFCN)
200 kHz spacing
Range of ARFCN:
1 - 124
1 n
Guard Band
100 kHz wide
Fu(n)
2 3 4
0 1 2 3 4 5 6 7
4.615 ms
timeslot = 0.577 ms
1 frame period
Raw data rate =
33.75kbps per traffic channel
270kbps per carrier channel
Education Division
DCS DCS - - 1800 Spectrum 1800 Spectrum
Uplink Downlink
1710 1785 1805 1880 MHz
Duplex spacing = 95 MHz
Guard Band
100 kHz wide
Channel Numbers (n) (ARFCN)
200 kHz spacing
Range of ARFCN:
512 - 885
1 n
Guard Band
100 kHz wide
Fu(n)
2 3 4
Education Division
GSM Logical Channels GSM Logical Channels
TCH
TCH
Traffic
Traffic
TCH/H
TCH/H
TCH/F
TCH/F
CCCH
CCCH
Control
Control
BCH
BCH
DCCH
DCCH
FCCH
FCCH
SCH
SCH
BCCH
BCCH
PCH
PCH
RACH
RACH
AGCH
AGCH
CBCH
CBCH
NCH
NCH
SDCCH
SDCCH
SACCH
SACCH
FACCH
FACCH
Two types of logical channel are defined; traffic and control channels
Each is further sub-divided as shown:
Education Division
T T
Traffic Channel Multiframe Traffic Channel Multiframe
The TCH multiframe consists of 26 timeslots.
This multiframe maps the following logical channels:
TCH Multiframe structure:
T T T T T T T T T T T T T T T T I T T T T T
0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 24 25 16 17 18 19 20 21 22 23
TCH
SACCH
FACCH
T = TCH S = SACCH I = Idle
FACCH is not allocated slots in the multiframe. It steals TCH slots when required -
indicated by the stealing flags in the normal burst.
T S
Education Division
Control Channel Multiframe Control Channel Multiframe
The control channel multiframe is formed of 51 timeslots
CCH multiframe maps the following logical channels:
0 1 42-45 46-49 50 32-35 36-39 40 41 22-25 26-29 30 31 12-15 16-19 20 21 2-5 6-9 10 11
RACH
Uplink
Downlink
Downlink:
FCCH
SCH
BCCH
CCCH (combination of PCH and AGCH)
Uplink:
RACH
F = FCCH S = SCH I = Idle
S BCCH F CCCH S CCCH F CCCH S CCCH F CCCH S CCCH F CCCH S CCCH F CCCH I
Control channel messages generally require 4 timeslots.
Education Division
Control Channel Configuration Control Channel Configuration
On the downlink, CCCH consists of paging (PCH) and access grant (AGCH)
messages
A combined multiframe has only 3 CCCH blocks to allow for SDCCH and
SACCH:
A non-combined multiframe has 9 CCCH blocks:
A complete paging or access grant message takes four bursts (timeslots),
i.e. one CCCH block.
S BCCH F CCCH S CCCH F CCCH S
SDCCH
0
F S F S F I
SDCCH
1
SDCCH
2
SDCCH
3
SACCH
0
SACCH
1
S BCCH F CCCH S CCCH F CCCH S CCCH F CCCH S CCCH F CCCH S CCCH F CCCH I
I
SACCH
0
SACCH
1
SDCCH
0
SDCCH
1
SDCCH
2
SDCCH
3
SDCCH
4
SDCCH
5
SDCCH
6
SDCCH
7
SACCH
2
SACCH
3
I I
Education Division
Control Channel Configuration Control Channel Configuration
TS0 (non-combined),TS2,TS4,TS6 4 110
TS0 (non-combined),TS2,TS4 3 100
TS0 (non-combined),TS2 2 010
TS0 (combined) 1 001
TS0 (non-combined) 1 000
Configuration
Number of CCCH
Timeslots
CCCH_CONF
Education Division
BCCH Carrier Characteristics BCCH Carrier Characteristics
Each cell has a designated BCCH carrier carrying the control
physical channel and BCCH logical channel
All BCCH carriers transmit continuously on full power by filling all
timeslots with traffic, signalling or dummy bursts
BCCH carriers use maximum frequency re-use distance to reduce
interference
TS 0 of the BCCH carrier contains the cells physical control
channel
The physical control channel comprises a number of logical
control channels
Additional physical channels on the BCCH may be required for
extra control channel capacity.
Remaining physical channels on BCCH carrier carry traffic
channels
Education Division
Paging Procedures Paging Procedures
Paging locates MS to cell Level for call
routing
Three paging message types:
Type 1 - 2 MSs using IMSI/TMSI
Type 2 - 3 MSs (1xIMSI, 2xTMSI)
Type 3 - 4 MSs using TMSI only
Paging message requires 4 bursts (1 CCCH
block)
Paging messages may be stored at BSS
Transmitted on PCH
If DRX is implemented MS listens only to
allocated paging group
Education Division
Timing Advance Timing Advance
Signal from MS1 takes longer to arrive at BTS than that from MS2
Timeslots overlap - collision
1 2 3
MS1 - Timeslot 1
1 2 3
MS2 - Timeslot 2
1 2 3
1 2 3
time
Timing Advance signal causes mobiles further from base station to transmit
earlier - this compensates for extra propagation delay
time
time
1 2 3
MS1 - Timeslot 1
1 2 3
MS2 - Timeslot 2
1 2 3
1 2 3
time
time
time
Timing Advance
Education Division
Timing Advance is calculated from delay of data bits in the access burst
received by the base station - long guard period allows space for this delay
Access burst data
delay
Guard Period Access burst data
TA signal is transmitted on SACCH as a number between 0 and 63 in
units of bit periods (0 and 219 for GSM 400).
TA value allows for round trip from MS to BTS and back to MS
Each step in TA value corresponds to a MS to BTS distance of 550
metres
Maximum MS to BTS distance allowed by TA is 35 km (GSM 900)
Education Division
Timing Advance value reduces the 3 timeslot offset
between downlink and uplink
0 1 2 3 4 5 6 7
Delay
3 timeslots
Downlink
Uplink
Timing
Advance
Actual delay
Uplink
TA value updates are sent to the MS on the SACCH
The Timing Advance technique is known as adaptive frame
alignment
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
Education Division
BCCH Allocation (BA) Lists BCCH Allocation (BA) Lists
Contains a list of BCCH carriers available in a certain
PLMN region
Passed to MS and stored in the SIM during IMSI Attach
procedure
MS monitors all carriers in BA list for signal strength.
In Idle mode this is used to reselect cell whilst moving
In dedicated mode, the measured values of the six best
neighbouring cells are reported to the serving BSC for
handover purposes
Education Division
CAMPING ON WITH A STORED BA LIST
CAMPING ON WITHOUT A STORED BA LIST
STORAGE OF BA LIST
THE BA(SDCCH) LIST
BA LIST INFORMATION
Education Division
BA Range List BA Range List
Octet n
Octet 13 RANGE4_HIGHER
Octet 12 RANGE4_HIGHER RANGE4_LOWER
Octet 11 RANGE4_LOWER RANGE3_HIGHER
Octet 9 RANGE3_LOWER
Octet 8 RANGE2_HIGHER
Octet 6 RANGE2_LOWER RANGE1_HIGHER
Octet 5 RANGE1_HIGHER (ARFCN) RANGE1_LOWER (ARFCN)
Octet 4 RANGE1_LOWER
Octet 3 Number of Ranges
Octet 2 Length of BA Range Contents
Octet 1 BA Range IEI
1 2 3 4 5 6 7 8
Education Division
BA ARFCN List BA ARFCN List
001 002 003 004 005 006 007 008
ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN
CA CA CA CA CA CA CA CA
113 114 115 116 117 118 119 120
ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN ARFCN
CA CA CA CA CA CA CA CA
121 122 123 124
ARFCN ARFCN ARFCN ARFCN spare spare Format ID
CA CA CA CA 0 0 0 0
Cell Channel Description IEI
1 2 3 4 5 6 7 8
CA = 0 not part of BA list
CA = 1 included in BA list
Education Division
BCCH Carrier Measurements BCCH Carrier Measurements
BCCH carrier measurements are used for:
Idle Mode:
Cell Selection/Reselection
Dedicated Mode
Handovers
Power Control
Call Maintenance
Education Division
BCCH Carrier Measurements BCCH Carrier Measurements Idle Mode Idle Mode
R
x
L
e
v
/ B
S
IC
R
x
L
e
v
/ B
S
IC
RxLev / BSIC
RxLev
110d
B
m
85Bm
11
2B
m
96dBm
Education Division
BCCH Carrier Measurement BCCH Carrier Measurement Idle Mode Idle Mode
Neighbouring Cells:
MS scans all carriers listed in the BA list and identifies the 6
strongest
Signal level averaged over at least 5 measurements
Result stored in RXLEV (n) parameter
MS must attempt to decode the BCCH channel of 6 best
neighbours at least every 30 seconds
Serving Cell
Measurements taken during allocated paging block
Measurements averaged over 5 consecutive paging blocks or
5 seconds (whichever greater)
Education Division
BCCH Carrier Measurements BCCH Carrier Measurements Dedicated Mode Dedicated Mode
R
x
L
e
v
/ B
S
IC
R
x
L
e
v
/ B
S
IC
RxLev / BSIC
RxLev
RxQual
Measurement
Reports
Education Division
BCCH Carrier Measurement BCCH Carrier Measurement Dedicated Mode Dedicated Mode
MS measures neighbouring cells for:
BCCH Receive Signal Level (Rx_Lev)
MS measures serving cell for:
BCCH Receive Signal Level (Rx_Lev_Full)
BCCH Receive Signal Quality (Rx_Qual_Full)
Idle TCH Receive Signal Level (Rx_Lev_Sub)
Idle TCH Receive Signal Quality (Rx_Qual_Sub)
MS sends a Measurement Report to the BSC on SACCH
every 4 TCH Multiframes
BSC measures MS uplink Rx_Qual and Rx_Lev
Education Division
Cell Selection Procedure Cell Selection Procedure
MS powers-up
MS starts measuring received power level of the BCCH carrier for all
cells in range
MS calculates average power level received from each cell:
Stored in RxLev(n) parameter
MS calculates a C1 parameter for each measured carrier based on the
RxLev(n) values
Mobile compares cells which give a positive value of C1 and camps-on
to the cell with the highest C1 value
Education Division
Cell Reselection Causes Cell Reselection Causes
C1 indicates that the path loss to the serving
cell has become too high
Downlink signalling failure
Serving cell has become barred
Better cell (in terms C2 value)
Random access attempts unsuccessful after
"Max retrans" repetitions
Education Division
Cell Re Cell Re- -selection selection GSM Phase 1 Mobiles GSM Phase 1 Mobiles
For GSM Phase 1 mobiles, cell reselection is achieved by comparing current cell C1
with neighbouring C1 cell measurements:
Between cells within a Location Area:
C1 (new) > C1 (old)
(for more than 5 seconds)
Between cells on a Location Area boundary:
C1 (new) > C1 (old) + OFFSET
(for more than 5 seconds)
Education Division
Cell Re Cell Re- -selection selection GSM Phase 2 Mobiles GSM Phase 2 Mobiles
GSM Phase 2 introduced a separate cell re-selection
parameter, C2
Intended to:
Prevent multiple handovers for fast-moving mobiles
Ensure MS camps on to cell with greatest chance of successful
communications
The C2 calculated is:
C2 = C1 + OFFSET (TEMPORARY_OFFSET x H(PENALTY_TIME T)
Education Division
C2 Cell Selection Parameters C2 Cell Selection Parameters
C2 = C1 + OFFSET (TEMPORARY_OFFSET x H(PENALTY_TIME T)
Offset:
optional parameter to encourage or discourage cell selection
H(Penalty_Time-T):
when a cell is added to list of strongest cells, a negative Temporary Offset
offset is applied for a Penalty time:
If timer expires, offset is removed making cell more attractive
If cell is removed from list, timer is reset
Used to prevent fast-moving MSs from selecting the cell
Temporary Offset:
Value of the negative offset applied
Education Division
Handover Processes Handover Processes
There are four different processes for handing over
within a GSM system, each requiring differing
procedures:
Channels (time slots) in the same cell
Cells within the same BSS (same BSC)
Cells in different BSSs (different BSCs) but under
the control of the same MSC
Cells under the control of different MSCs
GSM handovers are all hard i.e. mobile only
communicates with one cell at a time during the
process
Internal
External
BSC
MSC
VLR
BSC
MSC
VLR
BSC
Education Division
Handover Types Handover Types
Handovers can be initiated by either MS or MSC
Handover decisions are based on the following parameters
(in priority order):
UL/DL Signal Quality
UL/DL Signal Level
Interference
Power Budget
Distance of MS from BTS
Can be up- or down-link specific
Each parameter has operator-defined threshold parameters
Handover decisions can be based on one or a combination
of these parameters
Education Division
Measure serving and neighbouring
cells performance
Compare measurements with
enabled thresholds
Initiate handover procedure
Determine best candidate
neighbouring cell
Select cell and perform handover
Handover Initiation Handover Initiation
Education Division
Handover Procedure Handover Procedure - - Command Message Command Message
Structure of the message sent to MS by original BSS:
Message Type
Cell Description
Handover Reference
Power Command
Channel Description
Frequency List
or
Mobile Allocation
Includes Frequency Hopping
information if required
Non - Frequency Hopping
Frequency Hopping
MS BSS
Handover Command
Education Division
Handover Margin Handover Margin
N
o
m
i
n
a
l

c
e
l
l

b
o
u
n
d
a
r
y
BTS 1 BTS 2
Handover to BTS 2 Handover to BTS 1
Mobile remains with
BTS 1 or BTS 2
Hysteresis due to handover margin
Education Division
Handover Procedure Handover Procedure Signalling Example Signalling Example
Signalling for a basic Inter-BSC handover involving only one
MSC (Intra - MSC):
MS MSC BSS 1 BSS 2
Measurement report
Measurement report
Measurement report
Measurement report
Handover Required
Handover Request
Handover Command
Acknowledgement
Handover Command
Handover Access
Handover Detection
Physical Information
Handover Complete
Handover Complete
Clear Command
Clear Complete
Measurement report
Measurement report
Education Division
GSM Power Control Functions GSM Power Control Functions
Prevent unnecessary power emissions to:
Increase life of battery-powered devices
Reduce network interference
Equalise power levels received at BTSs
Adjustments to cell coverage
Methods Include:
Adaptive Power Control
Discontinuous Transmission (DTX)
Discontinuous Reception (RTX)
Education Division
Mobile Station Power Classes Mobile Station Power Classes
36 (4W) 480 DCS class 3
24 (.25W) 30 DCS class 2
30 (1W) 120 DCS class 1
29 (0.8W) 96 GSM class 5
33 (2W) 240 GSM class 4
37 (5W) 600 GSM class 3
39 (8W) 960 GSM class 2
dBm Power mW
Full Rate
MS Class
Power (mW) = Nominal maximum mean power output (milliwatts)
GSM class 1 deleted under GSM Phase 2 Specification
Source: ETSI GSM 02.06 (Version 4.5.2)
Power (dBm) = Maximum power output in dBm (+watts)
Education Division
BTS Power Classes BTS Power Classes
22.57 8
5 7
10 6
20 5
2.5 4 40 4
5 3 80 3
10 2 160 2
20 1 320 1
Maximum
O/P Power
(W)
TRX
Power
Class
Maximum
O/P Power
(W)
TRX
Power
Class
DCS 1800 GSM 900
Education Division
Receiver Sensitivity Levels Receiver Sensitivity Levels
-92dBm M3 Micro BTS GSM 1800
-102 dBm MS Class 3 DCS 1800
-100 dBm MS Class 1 or 2 DCS 1800
-104dBm Macro BTS GSM 900
-104dBm Other MS GSM 900
-102dBm Small MS GSM 900
Sensitivity Class System
Education Division
Adaptive Power Control Process Adaptive Power Control Process
Compulsory in MS, optional in BTS
32 power levels separated by 2dBm
Power changes are commanded using:
Reduction: POW_RED_STEP_SIZE (2, 4 dB steps)
Increase: POW_INC_STEP_SIZE (2, 4, 6 dB steps)
Commands issued on SACCH
One 2dB step change every 60mS
Education Division
Discontinuous Transmission (DTX) Discontinuous Transmission (DTX)
In a conversation, a person generally only speaks for
about 30% to 40% of the time
DTX makes use of this by reducing transmission when no
voice signal is detected
Uses a Voice Activity Detection (VAD) unit
Advantages:
Reduces interference
Prolongs battery life of mobile
Education Division
Discontinuous Reception (DRX) Discontinuous Reception (DRX)
Allows MS to power down parts of its circuitry in idle
mode
MSs within a Location Area divided into paging groups
MS only listens paging requests within its own group
Increases battery life of MS
Education Division
Power Control Power Control
Education Division
Frequency Hopping Concept Frequency Hopping Concept
TDMA Frame 1
F1 F1 F1 F1 F1 F1 F1 F1
TDMA Frame 2
F1 F1 F1 F1 F1 F1 F1 F1
TDMA Frame 3
F1 F1 F1 F1 F1 F1 F1 F1
TDMA Frame 1
F1 F1 F1 F1 F1 F1 F1 F1
TDMA Frame 2
F2 F2 F2 F2 F2 F2 F2 F2
TDMA Frame 3
F3 F3 F3 F3 F3 F3 F3 F3
TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7 TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7 TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7
TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7 TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7 TS0 TS1 TS2 TS3 TS4 TS5 TS6 TS7
Non-Frequency Hopping Carrier:
Frequency Hopping Carrier:
Education Division
Frequency Hopping Sequence Frequency Hopping Sequence
The frequency changes follow either a sequential or pseudo-random pattern
GSM defines 1 sequential pattern and 63 pseudo-random patterns
Each pattern is defined by a Hop Sequence Number (HSN)
Hop Sequence
One TDMA frame is 4.6 ms long
Rate of hopping = 1/ (4.6 x 10
-3
) = 217 hops / second
F1 F2 F3 F4 F1 F2 F3 F4 F1 F2 F3 F4 Sequential:
Hop Sequence
F1 F4 F3 F2 F1 F4 F3 F2 F1 F4 F3 F2 Pseudo-Random:
TDMA Frame
Education Division
Channel Timeslot Hopping Sequence Channel Timeslot Hopping Sequence
TDMA Frame 1
0 1 2 3 4 5 6 7 F1 0 1 2 3 4 5 6 7
TDMA Frame 2 TDMA Frame 3 TDMA Frame 4
0 1 2 3 4 5 6 7
0 1 2 3 4 5 6 7
F2
F3
Time
Education Division
Reasons for Frequency Hopping (1) Reasons for Frequency Hopping (1)
GSM radio signals are affected by multi-path interference, causing fading
Changing frequency (wavelength) moves the position of the fade
Frequency hopping cycles through many fade positions
This reduces the effect of the fades when the mobile is moving slowly
Tx
Rx
Fade position depends on
path difference in terms of
wavelengths
x
y
d
Fade when : (x +y) - d = n /2 where n = odd number
Education Division
Reasons for Frequency Hopping (2) Reasons for Frequency Hopping (2)
Cells are subject to interference
from other cells using the same
carriers at the re-use distance
If the cells hop through a set of
frequencies in different
sequences, the effect of this
interference is reduced
C/I ratio is increased
A1
A1
A3
A3
A2
A2
C1
C3 C2
B1
B3 B2
Education Division
Frequency Hopping at the BTS Frequency Hopping at the BTS
BCCH carrier will not hop - mobiles must be able to access
this for neighbour cell power level measurements
Only TRXs used for traffic channels will hop through set
sequences
The set of carrier frequencies assigned to the sequence
(Mobile allocation MA) will normally be from current cell
allocation
Hopping sequence for each TRX must be different or have a
different Mobile Allocation Index Offset (MAIO)
Education Division
Example of 4 Example of 4- -Frequency Hopping Frequency Hopping
BCCH carrier remains on single frequency
TCH carriers must start at different points in sequence (MAIO) to avoid
co-channel (C/I) interference
Above example uses same HSN for each TRX but different MAIOs
F4 F3 F2 F4 TRX 4
F3 F2 F4 F3 TRX 3
F2 F4 F3 F2 TRX 2
F1 F1 F1 F1 TRX 1
Hop 4 Hop3 Hop 2 Hop 1 Transmitter
Education Division
Baseband Frequency Hopping Baseband Frequency Hopping
Fixed TRX
Combiner
Antenna
Switch controller
Baseband
Data Signal
Fixed TRX
Fixed TRX
The baseband signal is fed to one of several TRXs in turn by a switch
The TRX outputs must be combined to be fed to the antenna
The combiner must be able to handle a wide bandwidth of signals
This can be achieved using either:
hybrid combiners - several stages causing large loss
cavity filters - one associated with each TRX - maximum loss ~ 5 dB
Education Division
Synthesiser Frequency Hopping Synthesiser Frequency Hopping
Baseband
Data Signal
Synthesiser
TRX
Tuning controller
Antenna
Education Division
Frequency Hopping at the MS Frequency Hopping at the MS
All mobiles must be capable of SFH in case it enters a
cell in which it is implemented
SFH is implemented to allow time to continue to take
measurements from adjacent cells
On connection/handover, the MS needs to know:
Frequencies used for hopping (Mobile Allocation)
Hop Sequence Number (HSN)
Start frequency (Mobile Allocation Index Offset - MAIO)
The MS uplink HSN is the same as the TRX downlink
HSN but offset by 45MHz
Education Division
Layer 3 Messaging Layer 3 Messaging
Management messages passed at network
layer 3, using layer 2 and 1 services for
transport.
Layer 3 messages divided in 3 groups:
Radio Resource (RR) Management
Mobility Management (MM)
Connection Management (CM)
Education Division
CM
MM
RR
LAPDm
TDMA
MS
RR
LAPDm
TDMA
BTSM
LAPD
G.703
BTSM
LAPD
G.703
BSSMAP
DTAP
BSSMAP
DTAP
CM
MM
Layer 1
Layer 2
Layer 3
U U
m m
A A
bis bis
A A
GSM Core GSM Core
Network Network
PSTN PSTN MSC
BTS
BTS
BSC
GSM Layer 3 Messaging GSM Layer 3 Messaging
BTS
BSC MSC
RR
SCCP SCCP
MTP MTP
Education Division
Layer 3 Message Format Layer 3 Message Format
IE Content n+2
IE Content n+1
Information Element ID (IEI) n
IE Content k+2
IE Content k+1
Information Element ID (IEI) k
Message Type 0 2
Protocol Discriminator Transaction ID 1
1 2 3 4 5 6 7 8
Protocol Discriminator: 0 0 1 1 - Call Control
0 1 0 1 - Mobility Management
0 1 1 0 - Radio Resource Management
1 0 0 1 - SMS messages
A Length Indicator (LI) may be used with variable length IEs
Message Header
Group
Payload
Group
Octets
Education Division
Section 2
Section 2
Network Characteristics
Network Characteristics
and
and
Problem Types
Problem Types
Education Division
Section 2
Section 2
2.3 BSS Capacity Issues 2.3 BSS Capacity Issues
2.2 BSS Coverage Issues 2.2 BSS Coverage Issues
2.4 Quality of Service (QoS) Issues 2.4 Quality of Service (QoS) Issues
Education Division
Main Planning Constraints Main Planning Constraints
The general requirements when designing a mobile network are to
maximise:
Coverage
Capacity
Quality of Service (QoS)
Cost-effectiveness - ROI ( Max Return Min Invest)
Education Division
BSS and Non BSS and Non- -BSS Issues BSS Issues
Network problem types can be divided into two distinct areas:
Those arising at the BSS (BTS - BSC)
Those arising in the transmission and the NSS
Although often resulting in the same effect, problems in each
area require a different approach:
BSS issues typically relate to frequencies, radio resource
dimensioning and/or maintenance of BSS database parameters
Non-BSS issues may relate to transmission links availability, call set-
ups, location updates and paging attempts
Education Division
Main Planning Constraints Main Planning Constraints
The general requirements when designing a mobile
network are to maximise:
Coverage
Capacity
Quality of Service (QoS)
Microsoft PowerPoint Presentation
Education Division
Section 3
Section 3
Introduction to
Introduction to
Performance Management
Education Division
Section 3
Section 3
3.3 The Performance Management Cycle 3.3 The Performance Management Cycle
3.2 Purpose of Performance Management 3.2 Purpose of Performance Management
3.4 Initial Network Design and Implementation 3.4 Initial Network Design and Implementation
3.6 Measuring Network Performance 3.6 Measuring Network Performance
3.5 Network Monitoring Phase 3.5 Network Monitoring Phase
3.7 Data Analysis Phase 3.7 Data Analysis Phase
Education Division
What is Performance Management? What is Performance Management?
Performance Management is a process
by which the network is continuously
monitored to ensure optimum
performance and to identify & rectify
problems areas.
Education Division
Purpose of Performance Management Purpose of Performance Management
Performance Management:
Ongoing process to monitor network performance
Sustains network quality throughout its lifecycle by proactive maintenance
Reduces risk of network degradation
Improve network performance through optimisation techniques
Supports business interests of the network operator
Poor maintenance procedures lead to network performance degradation
resulting in:
Reactive fault rectification
More complaints from customers
Customer churn
Reduced profit margins
Weakening of operators business position
Education Division
Performance Management Cycle Performance Management Cycle
Monitor
Network
Analyse Data
Yes
Identify
Problems
Implement
Changes
No
Initial Network
Design and
Implementation
Optimisation
QoS
Targets
Met?
Education Division
Initial Network Design and Implementation Initial Network Design and Implementation
Part of network design and implementation process
involves:
Defining quality of service (QoS) levels.
Defining KPIs for each reporting level
Defining custom network performance reports
Network performance KPIs will be measured against
the QoS targets.
QoS targets may require adjustment in light of
updated demographic/geographical planning data
and/or introduction of new services.
Education Division
Monitor
Network
Analyse Data
Yes
Identify
Problems
Implement
Changes
No
Initial Network
Design and
Implementation
Optimisation
QoS
Targets
Met?
Education Division
Network Monitoring Phase Network Monitoring Phase
Monitoring current network performance at various
levels:
Network, Regional MSC, BSS, NSS transmission levels
Performance data sources include:
Drive Tests
Statistical Measurements:
MSC, Databases (HLR/VLR), BSS Data (BTS/BSC)
Customer Complaints
Field Engineer Reports
Some sources generate information in the form of
performance metrics
Education Division
Performance Metrics Performance Metrics
Technical functionality:
Failed establishment, dropped calls, failed handovers etc
Traffic volume:
Number of Subscribers, offered traffic, erlangs per cell or
per subscriber, switch processor load etc
System availability:
Channel % availability, switch outage times, channel
outage times etc
System efficiency:
% utilisation of resources, are GoS targets being met? %
blocking of cells and core network channels etc
Education Division
Performance Measurement Methods Performance Measurement Methods
Generally, three methods:
Drive Testing
OMC Statistical Testing
Protocol Analyser Testing
Apresentao do
Microsoft PowerPoint
Apresentao do
Apresentao do
Education Division
Drive Testing Data Drive Testing Data
When Used:
Fault Identification from performance data
analysis or customer complaints
First phase of network audit prior to optimisation
Validate work carried out by engineers
Used to identify:
Coverage Gaps
Abnormal interference levels
Missing Neighbour relationships
Messaging protocol performance
Education Division
BCCH Carrier Measurement BCCH Carrier Measurement Idle Mode Idle Mode
Neighbouring Cells:
MS scans all carriers listed in the BA list and identifies the 6
strongest
Signal level averaged over at least 5 measurements
Result stored in RXLEV (n) parameter
MS must attempt to decode the BCCH channel of 6 best
neighbours at least every 30 seconds
Serving Cell
Measurements taken during allocated paging block
Measurements averaged over 5 consecutive paging blocks or
5 seconds (whichever greater)
Education Division
BCCH Carrier Measurement BCCH Carrier Measurement Dedicated Mode Dedicated Mode
MS measures neighbouring cells for:
BCCH Receive Signal Level (Rx_Lev)
MS measures serving cell for:
BCCH Receive Signal Level (Rx_Lev_Full)
BCCH Receive Signal Quality (Rx_Qual_Full)
Idle TCH Receive Signal Level (Rx_Lev_Sub)
Idle TCH Receive Signal Quality (Rx_Qual_Sub)
MS sends a Measurement Report to the BSC on SACCH
every 4 TCH Multiframes
BSC measures MS uplink Rx_Qual and Rx_Lev
Education Division
OMC Statistical Data OMC Statistical Data
OMC uses counters embedded with the
network software
Uses counters to record events
Counters are activated and results logged
Events relate to errors or quality
Analyses collected data to identify faults
Education Division
Protocol Analyser Characteristics Protocol Analyser Characteristics
Used for measuring signalling traffic
Intrusive or non-intrusive network connection
Should have hardware capable of:
Connecting to the required transmission medium
Displaying measurements for immediate analysis
Storing data for subsequent analysis
Exporting data to other {storage} devices
Should have software capable of:
Searching for protocol-specific data and parameters
Applying filters
Translating raw data into a readable display format
Providing remote access to the measuring device
Education Division
Protocol Analysers
Protocol Analysers
Tektronix K-1103
Tektronix K-1205-1279
Education Division
Comparison of Measurement Methods Comparison of Measurement Methods
OMC
Central position in network network-wide overview
Capability exists within standard network software
Continuous Monitoring capability
Speedier response to network problems
Drive Testing:
Can only provide data for limited geographical region
Can only provide a snap-shot of network characteristics
More accurate local picture
Can identify specific faults
Manpower and equipment resource intensive
Expensive
Protocol Analyser Testing
Can test uplink and down link more effectively than test Mobile.
Can test transmission links
Expensive hence cost-limited in number
Education Division
Monitor
Network
Analyse Data
Yes
Identify
Problems
Implement
Changes
No
Initial Network
Design and
Implementation
Optimisation
QoS
Targets
Met?
Education Division
Analysis of Data Analysis of Data
Data can be analysed under several headings:
Call success
evaluating the outcome of call attempt in terms of set-up time, clear
down success, assignment success etc
Statistical distributions
RxLev, RxQual
Handover analysis
showing success rate of attempted handovers
Neighbours
comparing neighbour cells found by signal level measurements with
the neighbour list in the site database
Coverage Analysis
Analysing the coverage threshold levels using serving cell/neighbour
cell comparison to identify problem areas
Quality
gives a comparison of signal quality from serving and neighbouring
cells
Education Division
Drive Test Data Analysis Screens Drive Test Data Analysis Screens
Examples of analysis screens
Call Success
Handover
Education Division
OMC Data Analysis Screens OMC Data Analysis Screens
Example of analysis screen
Education Division
Section 4
Section 4
Performance Measurement
Performance Measurement
Metrics
Metrics
Education Division
Section 4
Section 4
4.2 Key Performance Indicators 4.2 Key Performance Indicators
4.3 BSS KPI Definitions 4.3 BSS KPI Definitions
Education Division
KPI Introduction KPI Introduction
Education Division
KPI Introduction (2) KPI Introduction (2)
Education Division
Education Division
Education Division
Key Performance Indicators Key Performance Indicators
KPIs are established by the operator as a benchmark
against which to judge the performance of a network
Which KPIs the operator decides to use will depend on:
the nature of the network
operator assessment of priorities for their business
Counters implemented by vendors
KPIs are generally based on three sets of data:
Network statistics from the OMC
Drive test statistics using test mobile equipment
Protocol Analyser statistics
Education Division
General Key Performance Indicators General Key Performance Indicators
KPIS vary from operator to operator but the following
are generally common to all KPI statistics:
TCH Usage
SDCCH Usage
Handover Statistics
Connection Establishment
Apresentao do
Apresentao do
Apresentao do
Education Division
KPIs KPIs TCH/SDCCH Usage TCH/SDCCH Usage
The KPIS commonly monitored for TCH and SDCCH
usage include:
Traffic Flow
Mean holding time
Congestion
Busy time
RF Losses
Education Division
KPIs KPIs Handover Statistics Handover Statistics
The KPIS commonly monitored for Handover performance
monitoring include:
Success rate
Failure rate
Handovers due to poor quality
Handovers due to poor level
Power budget handovers
Handover due to interference
Inter-cell handovers
Inter-BSS handovers
Intra-BSS handovers
Education Division
KPIs KPIs Connection Establishment Connection Establishment
The KPIS commonly monitored for connection establishment
include:
Successful immediate assignment procedure
Attempted immediate assignment procedure
Cell TCH assignments
Education Division
Speech Quality Measurements Speech Quality Measurements
Speech QoS represented by:
Received Signal Level RxLev
Received Signal Quality RxQual
Factors affecting these values include:
BTS power output
MS power output
Current BTS loading
Education Division
KPI Definitions
KPI Definitions
Education Division
Call Setup Success Rate (CSSR) Call Setup Success Rate (CSSR)
Measures proportion of call setup attempts resulting in a
successful TCH assignment, excluding those setup
attempts for reasons that do not require a TCH (eg.
Location Update, IMSI Detach etc.)
CSSR = Successfully Completed Call Setups
Total Valid Call Setup Attempts
Education Division
Dropped Call Rate (DCR) Dropped Call Rate (DCR)
Measures the proportion of successfully established calls
that are terminated abnormally
) _ _ ( _ _ _
_ _ _ _
handovers incoming successful Setups Call Completed ly Successful
Drop Fail HO Losses RF TCH
DCR
+
+
=
Education Division
Call Success Rate (CSR) Call Success Rate (CSR)
Measures the proportion of call setup attempts that are
successfully established and terminated normally (ie. not
dropped)
) 1 ( DCR CSSR CSR =
Education Division
Handover Success Rate (HOSR) Handover Success Rate (HOSR)
Measures the proportion of total handover attempts
resulting in a successfully completed handover, including
inter-cell, Intra-BSS and Inter-BSS handovers.
Attempts Handover Total
handovers Successful Total
HOSR
_ _
_ _
=
Education Division
Handover Failure Rate (HOFR) Handover Failure Rate (HOFR)
Measures the proportion of handover attempts that result
in handover failure with dropped call (ie. excluding
handover attempts that fail and recover to the source cell)
Attempts Handover Total
Drop Fail HO Total
HOFR
_ _
_ _ _
=
Education Division
Traffic Channel Blocking ( Traffic Channel Blocking (TCH_Cong TCH_Cong) )
Measures the proportion of attempts to allocate a TCH that
are blocked due to lack of TCH resources. This should
exclude blocked TCH assignment attempts that
subsequently succeed due to queuing procedures.
Attempts Assignment TCH Total
s Assignment TCH Blocked
Cong TCH
_ _ _
_ _
_ =
Education Division
SDCCH Channel Blocking ( SDCCH Channel Blocking (SD_Cong SD_Cong) )
Measures the proportion of SDCCH Assignment Attempts
that are blocked due to lack of SDCCH resources.
Attempts Assignment SDCCH Total
s Assignment SDCCH Blocked
Cong SD
_ _ _
_ _
_ =
Education Division
Dropped Call Dropped Call RF Failure ( RF Failure (TCH_RF_Loss TCH_RF_Loss) )
Measures the proportion of successfully established calls
that drop due to TCH RF loss (as distinct from
HO_Fail_Drop)
Setups Call Completed ly Successful
Losses RF TCH Total
Loss RF TCH
_ _ _
_ _ _
_ _ =
Education Division
TCH Assignment Failure Rate ( TCH Assignment Failure Rate (TCH_Ass_Flr TCH_Ass_Flr) )
Measures the proportion of TCH assignment attempts that
fail after TCH allocation and before completion of the call
setup procedure.
Attempts Assignment TCH Total
Failures Assignment TCH
Flr Ass TCH
_ _ _
_ _
_ _ =
Education Division
SDCCH Access Success Rate ( SDCCH Access Success Rate (SD_Acc_Suc SD_Acc_Suc) )
Measures the proportion of Immediate Assignment
Attempts that result in successful assignment to SDCCH,
for all call setup reasons including location update, IMSI
detach etc.
Attempts Assignment mediate Total
s Assignment mediate Successful
Suc Acc SD
_ _ Im _
_ Im _
_ _ =
Education Division
Non Non- -BSS Call Setup Failure Rate (Non BSS Call Setup Failure Rate (Non- -BSS_CSFR) BSS_CSFR)
Measures the proportion of Call Setup Attempts that fail
due to Non-BSS reasons (eg. MSC rejections, subscriber
error, etc.)
Attempts Setup Call Valid
Failures Setup Call BSS Failures Setup Call Total
CSFR BSS Non
_ _ _
_ _ _ _ _ _
_

=
Education Division
Utilisation Factor Utilisation Factor
The ratio of actual traffic carried to the theoretical traffic
that can be carried for a 2% Grade of Service, based on
the number of available TCHs and Erlang-B formula.
GOS Erlangs l Theoretica
Carried Erlangs
n Utilizatio
% 2 _ _
_
=
Education Division
Section 5
Section 5
Measuring Network Performance
Drive Testing
Drive Testing
Education Division
Section 5
Section 5
5.2 Benefits and Limitations of Drive Testing 5.2 Benefits and Limitations of Drive Testing
5.3 Drive Test Equipment 5.3 Drive Test Equipment
5.4 Test Mobile Data 5.4 Test Mobile Data
Education Division
Drive Test Benefits and Limitations Drive Test Benefits and Limitations
Benefits:
Replicates subscriber conditions (QoS perspective)
Able to provide comparative performance between different operators
Focus on specific parameter set or geographical region
Limitations:
Difficult to replicate subscriber usage patterns
Area access restriction may limit realistic testing
Difficult to achieve network-wide snapshot
(limited time and/or geographical area)
Primarily down-link analysis only
Resource intensive and expensive
Ascom QVoice
Test Mobile Box
Education Division
Drive Testing Equipment Drive Testing Equipment
System consists of:
Test mobiles - either one or two for benchmarking against another
network
GPS and differential receivers to provide location information
Logging box to interface the measurement equipment to a laptop
computer
Computer running logging and analysis software (Neptune/Ranopt)
Education Division
The Logging Box The Logging Box
A logging box can be used to interface the test mobile
outputs and GPS signals to the laptop computer system.
The logging box is designed for use with Processing SW:
Connector for phone 1 (MS1)
Connector for handset 2 (MS2)
Charging switch (on/off)
for second phone (MS2)
GPS Receiver
Connector
Charging switch (on/off)
for first phone (MS1)
Cradle for handset 1 Cradle for handset 2
Handset 1 to
laptop connector
Handset 2 to
laptop connector
GPS to Laptop
connector
12V Power Supply
Status Indicator
Then use of logging boxes is diminishing as PC software
and hardware becomes more sophisticated and adaptable
Education Division
Test Mobile Data Test Mobile Data
Typical data collected is shown
here on the All Data screen in
Post Processing SW:
Serving cell information - BCCH,
BSIC, TCH, RxLev and RxQual,
Timing Advance, Tx Power
Neighbour cell information - BCCH
and RxLev for six best neighbour
cells
Education Division
For each measurement taken
the:
BCCH
BSIC
TCH
are identified
Education Division
the
RxLev Full
RxLev Sub
are identified
Full = monitored TCH
Sub = DTX TCH
Education Division
the
RxQual Full
RxQual Sub
are identified
Full = monitored TCH
Sub = DTX TCH
Education Division
the Timing Advance Value (TA)
value is recorded
Education Division
For each of the best six
neighbouring cells, the following
are recorded:
BCCH n
RxLev n
BSIC n
Education Division
Neighbour Lists Neighbour Lists
Many handover problems, dropped calls and so on are due to incorrect
neighbour lists.
Each time a change is made to a network, the neighbour relationships
should be rigorously updated.
Failure to maintain neighbour lists leads to problems such as:
Unwanted legacy neighbours
Neighbour lists reaching the maximum allowable (typically 32)
Neighbours defined on co-channel frequencies
Missing neighbours
Unintended one-way neighbour definitions
Depending on the vendor, the OMC may support automated neighbour
detection via A-bis interface traces.
Education Division
Neighbour Data Neighbour Data
RxLev data can be displayed
for the six best neighbour cells
and compared with the serving
cells RxLev
This can be used to identify
neighbouring cells that do not
currently exist in the BA list of
the serving cell
Education Division
Layer 3 Messages Layer 3 Messages
Neptune displays all layer 3
messages recorded by the test
mobile using the following fields:
Time of message
Direction
UL = Uplink
DL = Downlink
Message Category
RR = Radio Resource
MM = Mobility Management
CC = Call Control
Message Type
More detailed description
Education Division
Graphical Display Graphical Display
Data can be displayed
in a variety of graphical
forms
The display here
shows:
RxLev for the serving cell
RXLev for six best
neighbours
RxQual for the serving
cell
Frame Erasure Rate
(FER) of the current
measurement
Education Division
Test Mobile Data Test Mobile Data vs vs BSC Statistics BSC Statistics
A test mobile logging tool such as Actix or DeskCat allows data to be
collected by drive testing a live network
Advantages of test mobile data over BSC statistics:
BSCs provide counters with the number of dropped calls but do not
indicate why or where
BSCs do not collate information on poor downlink quality
BSCs cannot give information on areas without network access from
either poor signal or quality
BSCs do not store detailed information on calls
Test mobiles are the only solution for diagnosing localised network
performance issues
Education Division
Section 6
Section 6
The OMC
The OMC
Education Division
Section 6
Section 6
6.3 Statistical Data 6.3 Statistical Data
6.2 Role and Function of the OMC 6.2 Role and Function of the OMC
6.4 OMC Communications 6.4 OMC Communications
Education Division
Role of the OMC Role of the OMC
Perform operations and maintenance (O&M) activities
on GSM PLMN Elements
Provide central network overview
Often regionally-based under NMC hierarchy
Manually ordering or remotely configuring
disconnections or restoration of network elements
Manage the traffic loads at various points in the
network
Education Division
OMC Functions OMC Functions
Fault and Alarm Management
Configuration and Operations Management
Apresentao do
Apresentao do
Apresentao do
Education Division
OMC Fault and Alarm Management OMC Fault and Alarm Management
Monitoring System Alarms
Specific fault identification
Executing Diagnostic testing
Prioritise fault rectification
Implement corrective procedures based on
test results
Record faults to identify patterns/trends
Selective testing for pre-emptive
maintenance purposes
Education Division
OMC Configuration and Operations Management OMC Configuration and Operations Management
Permits physical and logical overview of
network elements
Remote reconfiguration of network elements
Controlling software upgrades
Introduction of new network features
Controlling network expansion
Network Auditing
Education Division
OMC Performance Management OMC Performance Management
Collection of traffic statistics and other
performance-related data
Scheduling of data collection periods
Determining composition of performance reports
Analysis of collected data to:
Identify performance trends
Identify fault trends
Identify pre-emptive maintenance tasks through
selective reporting
Education Division
Statistical Data
The network is capable of generating a huge amount of statistical
data. There is therefore a danger that information overload could
obscure some fundamental problems. Hence, it is important to
categorise the data in terms of importance and relevance. To
achieve this, performance counters are divided into groups by
network element and frequency of sampling.
Education Division
Main QoS Parameters Main QoS Parameters
Dropped Call Rate (< 2%):
Blocking (GoS) (2%-5%):
Traffic Throughput:
Traffic for one subscriber = no of calls per hour per subscriber * MHT
3600
100 =
tiveCalls NumberOfAc
oppedCalls NumberOfDr
lRate DroppedCal
100
=
llAttempts NumberOfCa
s llSuccesse NumberOfCa llAttempts NumberOfCa
GOS
Education Division
Periodic Counters Periodic Counters
Daily:
Cell Level
Rapid Fault Identification
Includes Call setup rate, TCH/SDCCH congestion, TCH drop call
rate
Weekly:
Cell Level
Performance Trends
Includes traffic trends, cell retention, cell accessibility
Monthly:
BSC Level
Performance Trends
Includes processor loading, BSS call setup, handover success rates
Apresentao do
Apresentao do
Apresentao do
Education Division
Daily Cell Measurements Daily Cell Measurements
Call Setup Success Rate
Dropped Call Rate
Handover Failure Rate
SDCCH Blocking
TCH Blocking
Education Division
Call Setup Success Rate Call Setup Success Rate
No access to SDCCH
Congestion prevents SDCCH allocation
CM service reject
Incompatible algorithms/cipher keys
Requested service not available
TCH assignment failure
Congestion prevents SDCCH allocation
Excessive Interference on TCHs
Hardware Problems
Education Division
Call Setup Success Rate Call Setup Success Rate
site 20B has a high PTFAIL level (23.4% TCH setup failure)
Education Division
Dropped Call Rate Dropped Call Rate
Drop on Handover
Interference on target cell
Incorrect serving cell
Low Signal Level/Quality
RxLev/RxQual thresholds exceeded
MS enter high attenuation area (building/tunnel etc)
Requested service not available
High Co-, Adjacent- and/or Extraneous Interference
Excessive interference on TCHs
Excessive interference on CCHs
Link Imbalance
Hardware degradation
Education Division
Education Division
site 130 A has large disconnection failure rate (411) due to bad signal level (DISBSS)
Education Division
Handover Failure Rate Handover Failure Rate
High Neighbour Interference/congestion
Interference on neighbouring cells
Lack of capacity in neighbouring cells
No Dominant Server
Difficult cell selection decision
Database Parameters
Incorrect threshold parameter settings
Link Imbalance
Education Division
SDCCH Blocking SDCCH Blocking
No Access on SDCCH
Interference on serving cell
Lack of capacity on serving cell
Failure before Assignment of TCH
Invalid service request
High Signalling Load
High location update signalling
Contention with SMS signalling
Education Division
SDCCH Blocking SDCCH Blocking
Table presenting the cell counters for the last hour. ( site 130A showing high 0.4 PCCONG)
Education Division
TCH Blocking TCH Blocking
Handover Margin
Low margin values may reduce handovers causing
congestion
Cells Too Large
Antenna too high
Excessive power
Excessive traffic capture
Capacity Limitations
Poor capacity dimensioning
Unanticipated traffic capture
Education Division
Weekly Measurements Weekly Measurements
Traffic Trends
Cell Retaining
Cell Accessibility
Education Division
Weekly Measurements Weekly Measurements Traffic Trends Traffic Trends
TCH Traffic
SDCCH Traffic
Derived mainly from peak-hour statistics
Education Division
Traffic trend for cell site 285B
Education Division
weekly busy hour analysis
Education Division
Weekly Measurements Weekly Measurements Cell Cell Retainability Retainability
Measures ability to retain calls
Frequently monitored using drops through
poor RxQual or RxLev
Education Division
Education Division
Percentage of idle channel measurement in the fifth band
Education Division
Weekly Measurements Weekly Measurements Cell Accessibility Cell Accessibility
Measures ability to establish calls
Uses PCFAIL, PCCONGS, PCONGS, PTFAIL
counters
Education Division
Traffic failure analysis graph
Education Division
Education Division
Monthly Measurements Monthly Measurements
Processor Load
Call Success Rate
Handover Success Rate
Education Division
Monthly Measurements Monthly Measurements Processor Load Processor Load
Processor Load of BSC CMSC1
Education Division
Monthly Measurements Monthly Measurements Call Success Rate Call Success Rate

call setup analysis
Education Division
Monthly Measurements Monthly Measurements Call Success Rate Call Success Rate
call drop analysis
Education Division
Monthly Measurements Monthly Measurements Handover Success Rate Handover Success Rate

Education Division
Monthly Counters Monthly Counters Handover Success Rate Handover Success Rate
Education Division
Customised Queries / Reports Customised Queries / Reports
Most OMC Software tools allow customer-
configurable reports
Enables ad-hoc queries
Enables targeting of specific parameter set
Examples of useful customised reports
include:
BSS Capacity / Traffic Report
BSC Performance Report
Education Division
BSC capacity report
Education Division
BSC performance report
Education Division
BSC performance report in excel format
Education Division
OMC Communications OMC Communications
OMC must communicate with other network elements
Uses a separate TMN to communicate with network
components frequently utilising PSTN lines
OMC data passed using:
SS7 protocols for short/medium length control messages
within GSM network
X.25 protocols for large external data transfers such as
downloads or subscriber database transfers between
subscription centres
Education Division
OMC Communications OMC Communications
BSC
MSC
NMC
VLR VLR
BTS
BTS
BSC
BTS
BTS
BSC
BTS
BTS
BSC
BTS
BTS
MSC
VLR VLR
HLR HLR
OMC
Relational Relational
database database
Analysis Analysis
Tools Tools
SS7 / X.25 Links
Education Division
Section 7
Section 7
Introduction to Optimisation
Introduction to Optimisation
Education Division
Section 7
Section 7
7.2 Requirements for Optimisation 7.2 Requirements for Optimisation
7.4 Network Audit Phase 7.4 Network Audit Phase
7.3 Outline Optimisation Process 7.3 Outline Optimisation Process
7.5 Network Performance Review Phase 7.5 Network Performance Review Phase
7.6 Activity Phase 7.6 Activity Phase
Education Division
Network Optimisation Network Optimisation
No prescribed methodology
Often network architecture dependant
Often vendor equipment-dependant
Often engineer-experience dependant
An overview course
Insight into one approach to processes
Optimisation is an art as much as a science
A tool-box approach
Education Division
Dictionary Definition of Optimisation Dictionary Definition of Optimisation
Determining the best compromise
between potentially conflicting
requirements in order to plan and
implement an activity with maximum
efficiency.
Education Division
Definition of Quality Definition of Quality
Education Division
A Mobile Network Definition of Optimisation A Mobile Network Definition of Optimisation
The identification and rectification of
performance affecting problems within
the constraints of an existing network
infrastructure.
Education Division
Optimization Issue Optimization Issue
Education Division
Radio Network Planning Process Radio Network Planning Process
Education Division
Optimization Process Optimization Process
Education Division
Communication
Interference
Quality
Capacity
Coverage
Planning Requirement Planning Requirement
Cost
Education Division
Education Division
Reason for Optimisation Reason for Optimisation
Goal:
Ensure optimum network performance within QoS constraints
Process:
Identify network problems through analysis of KPIs
Implement corrective actions and/or procedures to rectify problems
Monitor results of modifications
Purpose:
Maintain/improve QoS
Reduce churn rate
Attract new customers
Maximise revenue-generating service
Maximize efficiency of network functional elements
Education Division
Requirements for Optimisation Requirements for Optimisation
Network issues that may generate a requirement for optimisation include:
Perceived reduction in network quality
Indications from network performance monitoring
Subscribers experience of using the network
Maximising the use of existing infrastructure
operator wants to ensure best return on investment
Introduction of new services
maximise existing resources to accommodate new services such as GPRS
Change in original design parameters
Flawed original design information
Original design information has changed
Regardless of the initiating event, the optimisation procedures remain
generally the same
Education Division
Requirements for Optimisation (2) Requirements for Optimisation (2)
GSM
Education Division
Monitor
Network
Analyse Data
Yes
Identify
Problems
Implement
Changes
No
Initial Network
Design and
Implementation
Optimisation
QoS
Targets
Met?
Education Division
Outline Optimisation Process Outline Optimisation Process
Initial network design
and roll-out phase
Monitoring/
Network Audit Phase
Optimisation Activity
Phase
Design Review and
growth phase
Education Division
Network Audit Phase of Optimisation Network Audit Phase of Optimisation
Decide on KPIs,
Measurement Strategy and
Tools
Measure Performance,
Establish Benchmark
Performance Review to
Identify Major Performance
Affecting Issues
Decide on Strategy,
Establish Action Plan Feedback from
Optimisation
Activities
Start Optimisation Activity
Education Division
Network Audit Process Network Audit Process
Performance Review Processes:
Network performance statistics (OMC)
A-Interface performance statistical analysis
Call Trace Analysis
Alarms and events
GSM performance drive test
GPRS Performance drive test
Competitive drive test benchmarking
Education Division
For each category the following structure can be
applied:
Objective. What are we trying to measure?
Description. Why measure it and its relevance to
performance?
Report Form. How should the results be presented?
Interpretation. What conclusions can be drawn from
results?
Recommendations. How can the problem be resolved?
Education Division
OMC Statistics Review:
Call Success Rate
Call Setup Rate
TCH Blocking
TCH Assignment Failure (RF)
SDCCH RF Loss
SDCCH Access performance
Dropped Calls
Handovers
Education Division
A Interface Analysis:
Call Setup Failures
Location Update Success Rate
Handover Causes
Education Division
Call Trace Analysis:
DL receive level and BTS power
UL receive level and MS transmit power
UL and DL RxQual distributions
Education Division
Events and Alarms:
DL receive level and BTS power
UL receive level and MS transmit power
UL and DL RxQual distributions
Education Division
Drive Test Metrics - Route :
Coverage route plot
Quality route plot
Frame erasure (FER) route plot
MS transmit power route plot
Speech Quality Index (SQI) route plot
Education Division
Drive Test Metrics - Events :
Dropped call events
Call setup failure events
Handover failure events
Successful handover events
Education Division
GPRS-Specific Drive Test Metrics:
RLC throughput
LLC throughput
RLC BlkER
RLC retransmission rate
Coding scheme usage (CS1-4)
Allocated timeslots
Packet loss rate
Latency/Jitter
PDP Context activation failure
PDP context loss (GPRS call drop)
Education Division
Activity Phase of Optimisation Activity Phase of Optimisation
Start Optimisation Activity
Identify and Fix Hardware Problems
Ongoing
Performance
Measurement
Process
Review Process and Results
Feedback to modify
strategy decisions
BSS Database
Parameters Review:
Settings
Consistency
Change Control
Design Review and Growth Plan
Identify and Fix Neighbour Problems
Identify and Fix Frequency Plan Problems
Network Audit Procedures
Feedback
to
Audit
Procedures
Education Division
Section 8
Section 8
Optimisation Activities
Optimisation Activities
Education Division
Section 8
Section 8
8.3 Identifying and Fixing Hardware Problems 8.3 Identifying and Fixing Hardware Problems
8.2 BSS Database Parameter Review 8.2 BSS Database Parameter Review
8.4 Identifying and Fixing Neighbour Problems 8.4 Identifying and Fixing Neighbour Problems
8.5 Identifying and Fixing Frequency Plan Problems 8.5 Identifying and Fixing Frequency Plan Problems
Education Division
BSS Configuration Parameter Review BSS Configuration Parameter Review
Review current settings as related to performance
measurement results
Recommend changes to improve/optimise performance of
specific features.
Review includes:
Handover parameters, timers, thresholds and margins
Power control thresholds
Averaging mechanisms for handover and power control
Call setup parameters to maximise resource utilisation
C1/C2 cell reselection parameters
Any vendor-specific advance traffic management algorithms
Education Division
BSS Database Parameter Consistency BSS Database Parameter Consistency
Within a network, different site types are defined (e.g. urban
micro, rural macro etc) by a standard template
Each site type database will comprise a default parameter
set
Each site may modify default set to suit local conditions
Consistency of the default parameter sets should be
checked across BSS types
Change control management processes should be reviewed
to ensure procedural consistency
Education Division
BSS Configuration Parameter Sets BSS Configuration Parameter Sets
Each BSS operates in accordance with its software
configuration
All BSS are configured with certain standard default
parameters
Each BSS will have a subset of BSS-specific parameters
Parameters defined by:
ETSI GSM Recommendations (GSM 12.04)
Proprietary Parameter Implementation
Potentially hundreds of configuration parameters
Many parameters are inter-dependant
Often vendor-specific abbreviations/acronyms used for same parameter
Education Division
BSS Configuration Parameter Types BSS Configuration Parameter Types
Identifiers:
CI, LAI, GCI, BSIC etc
Channel Configuration:
TCH channels, Signalling channel configuration (e.g. CCCH)
Timers:
Location Updates, C2 calculations etc
Thresholds:
RxLev, RxQual for handover decisions etc
Offsets:
Hysteresis for handovers etc
Control Features:
SFH, DTX, DRX etc
Education Division
Adjusting BSS Configuration Parameters Adjusting BSS Configuration Parameters
Effected from:
PC connected directly to hardware
Remotely from OMC/NMC
Can be individually addressed or broadcast
May require hardware reset to effect change
Be aware of hierarchical changes (MSCBSCBTS)
Only implement during low-traffic periods
Use test BSS where available
Avoid simultaneous multiple parameter changes
Education Division
Identify Hardware Problems Identify Hardware Problems
Problems identified through analysis of performance
measurement counters from:
OMC Statistics
Driver \Test data
A-bis and A interface logs
Identify worst performing cells according to established KPI
benchmarks
Education Division
Typical Hardware Problems Typical Hardware Problems
Misaligned Antennas
Increased interference, coverage degradation
Wrongly/poorly connected feeders
Reduced power output, reduced coverage, cell imbalance, distortion
Poor transmission line performance (water ingress, corrosion,
physical damage etc
High VSWR, high Insertion losses, reduced power output, reduced
coverage, cell imbalance, distortion
Poorly calibrated TRXs
Inconsistent TRX performance
Incorrectly configured combiners/duplexers etc
Reduced power output, reduced coverage, cell imbalance, distortion
Education Division
Optimising Neighbour Lists Optimising Neighbour Lists
Effects of poor maintenance:
Unwanted legacy neighbours
Oversized neighbour lists
Co-channel neighbour definitions
Missing neighbours
Unintentional 1-way neighbour definitions
Optimise by:
Analyse neighbour performance form statistics.
Utilise automated neighbour detection
Identify inconsistent neighbour profiles
Modify appropriate neighbour lists
Education Division
Frequency Optimisation Activities Frequency Optimisation Activities
Identify problem areas:
Interference issues (internal and external)
Coverage issues ( excessive overlaps, coverage gaps, high sites etc)
Recommend appropriate frequency optimisation techniques
BCCH and TCH Frequency Plans
Frequency Diversity (space, polarity)
Frequency Hopping
DTX
Antenna Down tilting
Advanced Optimisation Techniques
Education Division
BCCH and TCH Frequency Plans BCCH and TCH Frequency Plans
BCCH is typically planned with low frequency re-use (long re-use
distance) in order to ensure high quality.
The BCCH plan should take into account: site design, terrain and
topography and subscriber distribution. A good BCCH plan can generally
be achieved with 14 - 15 carriers.
The TCH plan requires the same considerations as BCCH, but may also
employ:
Frequency hopping - synthesiser or baseband
Multiple Re-use Patterns - giving tighter re-use on lower TCH layers
Concentric multi-layer cell arrangements in which BCCH is only required on
one band
Education Division
BCCH and TCH Frequency Plans BCCH and TCH Frequency Plans
BCCH TCH Hopping MICRO
Guard Band Guard Band
14 ch
1 ch
26 ch 8 ch
1 ch
Education Division
High Site Problems High Site Problems
Initial roll-out tends to concentrate on higher sites
High sites potentially cause excessive interference with
subsequent lower site roll-out
A Typical process for replacement could include:
From performance parameters, identify those cell interference-
contributing the most interference to the most cells
Develop plan to lower antennas or decommission these sites. This
may require additional lower sites to cover any coverage gaps
Prioritise decommissioning and integration of new sites to minimise
disruption to services
Education Division
Antenna Down Antenna Down- -Tilting Tilting
An option for adjusting cell coverage
e.g. down tilting may direct coverage deeper into a building
Antenna tilt may be:
mechanical operator set affects directional coverage
electrical manufacturer set affects omni-directional coverage
Omni antennas may have electrical tilt but not mechanical
New technologies allow for remote electrical tilting
Education Division
Examples of Antenna Tilting Examples of Antenna Tilting
-35
-30
-25
-20
-15
-10
-5
0
-35
-30
-25
-20
-15
-10
-5
0
-35
-30
-25
-20
-15
-10
-5
0
-35
-30
-25
-20
-15
-10
-5
0
No Tilt
Mechanical
Downtilt
Electrical
Downtilt
Electrical Downtilt +
Mechanical Uptilt
Education Division
Tilting Effects Tilting Effects
Education Division
2Downtilt 0Downtilt
Education Division
6 Downtilt 4 Downtilt
Education Division
8 Downtilt
10 Downtilt
Education Division
Antenna Selection Antenna Selection
Factors to consider with Antenna selection:
Vertical and Horizontal Beamwidth
Gain
Front-to-Back Ratio
Null Fill
Downtilt (electrical/mechanical)
Education Division
Antenna Configuration Options Antenna Configuration Options
Combining:
Economical with antenna elements
Reduced coverage due to insertion loss
Diversity Choice:
Horizontal space diversity
Vertical space diversity
Polarisation diversity
Education Division
Section 9
Section 9
NSS Characteristics
NSS Characteristics
Education Division
Section 9
Section 9
9.3 Overview of SS7 Signalling in the NSS 9.3 Overview of SS7 Signalling in the NSS
9.2 NSS Topology 9.2 NSS Topology
Education Division
GMSC GMSC
MSC MSC
PLMN Configuration PLMN Configuration
HLR HLR
AuC AuC
EIR EIR
MSC MSC
VLR VLR
VLR VLR
BSC BSC
BSC BSC
BSC BSC
BSC BSC
Traffic Links/Link Sets
Signalling Links/Link Sets
Education Division
Big BSC Big BSC Configuration Configuration
Education Division
Transit Switching Layer Transit Switching Layer
TSC
1a
HLR 2
300k
HLR 3
150k
HLR 4
1200k
HLR 5
600k
HLR 1
600k
TSC
3b
TSC
4b
TSC
3a
TSC
1b
TSC
4a
TSC
2b
TSC
2a
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
MSC
IN
GPRS
VMS
SMS
HLR/SMS/VMS/IN
Layer
TSC Layer
MSC Layer
Education Division
A Interface Configuration A Interface Configuration
MSC MSC
BSC BSC
fm Trunk 2 n/u TS31 TS30 TS29 30
:
:
:
:
:
:
:
:
:
:
fm Trunk 3 TS20 TS19 TS18 TS17 19
n/u
TS20
TS20
n/u
n/u
n/u
:
:
TS8
TS4
TS4
TS4
31
18
17
16
15
14
13
:
:
8
5
4
3
2
1
0
fm Trunk 3 TS31 TS30 TS29
C7 Signalling
:
:
:
:
:
:
fm Trunk 2 Trunk 2 TS0 Data
fm Trunk 1 Trunk 1 TS0 Data
Frame Alignment
TRUNK 1
TRUNK 2
TRUNK 3
TRUNK 1
TRUNK 2
TRUNK 3
2Mbps E1 Link
Education Division
TRAU Configurations TRAU Configurations
BTS Site BTS Site
BSC Site BSC Site MSC Site MSC Site
TRAU
BTS Site BTS Site
CCU
CCU
BTS Site BTS Site
CCU
CCU
A A
B B
C C
A A
bis bis
U U
m m
A A
MSC Node MSC Node BSC Node BSC Node CCU
Channel Coding Unit Channel Coding Unit
TRAU
TRAU
16kbps 16kbps
16kbps 16kbps
16kbps 16kbps 16kbps 16kbps
64kbps 64kbps
64kbps 64kbps
64kbps 64kbps
64kbps 64kbps
16kbps 16kbps
CCU
CCU
Education Division
GMSC GMSC
MSC MSC
PLMN Interfaces PLMN Interfaces
HLR HLR
AuC AuC
EIR EIR
MSC MSC
VLR VLR
VLR VLR
A A
C C
C C
F F
F F
B B
B B
H H
D D
D D
E E
A A
A A
A A
G G
E E
E E
BSC BSC
BSC BSC
BSC BSC
BSC BSC
Education Division
Network Signalling Requirements Network Signalling Requirements
BSC to MS GSM-specific Signalling
BSC to NSS C7 Signalling:
User Data Connections:
ISDN User Part (ISUP)
Inter-NSS Node Signalling:
C7 Mobile Application Part (MAP)
A Interface Signalling:
C7 BSS Application Part (BSSAP)
Physical Layers
Message Transfer Part (MTP)
Education Division
C7 Network
GMSC GMSC
MSC MSC
PLMN Interfaces PLMN Interfaces C7 Signalling C7 Signalling
HLR HLR
AuC AuC
EIR EIR
MSC MSC
VLR VLR
VLR VLR
A A
C C
C C
F F
F F
B B
B B
H H
D D
D D
E E
A A
A A
A A
G G
E E
E E
BSC BSC
BSC BSC
BSC BSC
BSC BSC
Education Division
Generic SS7 Network Generic SS7 Network
SP
SP
SP
SP
STP
STP
SP
STP
STP
SP
SP
SP
SP
Education Division
MSC MSC
C7 in the GSM NSS C7 in the GSM NSS
BSC BSC
HLR HLR
AuC AuC
EIR EIR
BSC BSC
BSC BSC
BSC BSC
MSC MSC
VLR VLR
VLR VLR
SP
SP
SP
SP
SP
SP
SP
SP
SP
STP
STP
Education Division
MTP 1-3
A Interface Signalling A Interface Signalling
SCCP
DTAP
BSSMAP
User Data User Data
BSSAPP BSSAPP
Layer 1 Layer 1- -3 3
Over the A Interface, GSM uses C7 signalling with protocol
extensions for GSM-specific requirements.
Education Division
BSS Mobile Application Part (BSSMAP)
All messages exchanged between BSC and MSC processed by
the BSC
Direct Transfer Application Part (DTAP)
Message protocol for direct transfer of MM/CM Layer 3 messaging
between the MSC and the MS
Transparent to the BSS
Signalling Connection Control Part (SCCP)
Logical signalling connection between MSC and MS to support
Layer 3 (MM/CM) message transfers
Uses reduced set of C7 signalling instructions
Message Transfer Part Layers 1-3 (MTP 1-3)
Responsible for routing and transport of signalling messages
A Interface Signalling Protocols A Interface Signalling Protocols
Education Division
MTP
SCCP
TCAP
MAP
NSS Interface Signalling NSS Interface Signalling
HLR
Layers 4 Layers 4- -6 6
MSC MSC VLR
E E
C C
B B
D D
Layer 3 Layer 3
Layers 1 Layers 1- -2 2
Layer 7 Layer 7
C7 Protocol Stack C7 Protocol Stack
EIR
F F
Education Division
Mobile Application Part (MAP)
Mobile-specific extension of C7 standard.
All messages exchanged between NSS Elements (MSC,HLR,VLR)
Transaction Capabilities Application Part (TCAP)
A connectionless transport service provided by the SCCP
Provides core functionality to support roaming
Allows its users to access other users via worldwide C7/SS7 network
Signalling Connection Control Part (SCCP)
Logical signalling connection between MSC and MS to support Layer 3
(MM/CM) message transfers
Uses reduced set of C7 signalling instructions
Routes messages via STPs
Message Transfer Part (MTP)
C7 protocols responsible for routing and transport of signalling messages
NSS Interface Signalling Protocols NSS Interface Signalling Protocols
Education Division
Section 10
Section 10
NSS Parameter Evaluation
NSS Parameter Evaluation
and
and
Optimisation
Optimisation
Education Division
Section 10
Section 10
?.1 Introduction ?.1 Introduction
?.4 Traffic Loading Evaluation ?.4 Traffic Loading Evaluation
?.2 SS7 Signal Routing Analysis ?.2 SS7 Signal Routing Analysis
?.5 Parameter Evaluation ?.5 Parameter Evaluation
?.6 Database Usage ?.6 Database Usage
?.3 SS7 Link Load Evaluation ?.3 SS7 Link Load Evaluation
?.7 NSS Optimisation Process ?.7 NSS Optimisation Process
Education Division
NSS Optimisation Process NSS Optimisation Process
NSS Data Capture
Phase
Data Analysis Phase
Define Optimisation
Requirements
Define Implementation
Strategy
Optimisation Activity
Phase
Education Division
NSS Data Capture Phase NSS Data Capture Phase
Available Data:
C7 signal routing
C7 link loading
User traffic routing
Trunk group configuration and loading
NSS configuration parameters
Database loading
Education Division
NSS Data Analysis Phase NSS Data Analysis Phase
STEP 1
Organise collected data into functional areas
STEP 2
Validate organised data
STEP 3
Evaluate validated data
Education Division
NSS Optimisation Phase NSS Optimisation Phase
STEP 1
Review areas of concern identified as part of
the Validation Phase
STEP 2
Identify optimisation techniques suitable for
improving poorly performing areas
Education Division
NSS Implementation Phase NSS Implementation Phase
Plan the implementation of optimisation
procedures
Ensure sufficient time/manpower/equipment
resources are available of the right time
De-conflict potentially conflicting procedures
Identify monitoring procedures/periods required
to assess performance changes
Education Division
NSS Optimisation Activity Phase NSS Optimisation Activity Phase
Implement optimisation procedures:
as identified during the Optimisation Phase
in accordance with the plan designed during the
Implementation Phase
monitored as agreed during the Implementation
Phase
Education Division
Routing Evaluation Routing Evaluation
Check for:
Main and alternate routing
Presence of routing circles
Consistent implementation of routing strategy
Effective use of transit switches (where used)
Apresentao do
Apresentao do
Apresentao do
Apresentao do
Education Division
Routing Analysis Routing Analysis Main and Alternate Routing Main and Alternate Routing
Each signalling destination should be reachable by
a minimum of a main and alternate route,
particularly HLR connections
Too few routes can cause loss of signalling
messages
Too many routes is inefficient and overly-complex to
manage
Education Division
Routing Analysis Routing Analysis Routing Circles Routing Circles
Also known as tromboning
A
C B
A-C Main / B-C Main
A-C Alternate
B-C Alternate
A-C and B-C Fail
A routes via B to C
B routes via A to C
A routes back to B
Education Division
Routing Analysis Routing Analysis Strategy Consistency Strategy Consistency
Strategy for signal routing should be consistent
in order to:
Optimise signalling link resources
Ease engineering by applying a consistent approach
Spread traffic loads evenly across the network
Increase reliability
Simplify future growth planning
Education Division
Routing Analysis Routing Analysis Transit Layer Switches Transit Layer Switches
Transit Layer switching provides an additional
layer of switching that makes signal routing
more efficient
Signalling traffic should be evenly distributed
across TSCs in TSC layer
If TSCs are used, MSCs should not be used
as STPs as this reduces MSC and SS7
performance
Education Division
Possible Routing Problems Possible Routing Problems
Lack of uniformally implemented routing strategy:
Number of primary/secondary/tertiary routes
Inconsistent Routing parameters
Excessive signal loading on routes
Lack of alternate routing to key network elements
Signalling errors due to routing circles
Too many routes defined uneven spreading of load
Education Division
C7 Signalling Links:
Signalling links are logically independent of voice traffic links
Links are full-duplex
Periodicity of Monitoring:
Carried out at least monthly (preferably weekly)
Carried out on significant network configuration change
Recommended Capacity:
ITU Q.706 specifies maximum of 200mE per link in each direction
May be exceeded temporarily but is not recommended
Calculation:
Signalling Load [ mErl] = (Number of MSU * 6 + Number of octets of payload)
( 8 * length of time in seconds)
Monitoring Signalling Loads Monitoring Signalling Loads
Education Division
Signal Load Evaluation Signal Load Evaluation Link Load Sharing Link Load Sharing
Excessive link loads can occur
when:
Original signal load dimensioning was
not correct
Network characteristics have changed
Ideal Situation:
Links should not exceed 200mE
Signalling traffic should be spread
evenly across all links in link set
102 104 9
89 103 8
109 104 7
94 104 6
91 312 5
76 302 4
84 303 3
127 305 2
134 307 1
121 305 0
HLR1
Tx [mErl] Rx [mErl]
Link
No Destination
C7 Load C7 Load TSC1
Education Division
Signal Load Evaluation Signal Load Evaluation Link Load Symmetry Link Load Symmetry
Ideal Situation:
Individual links should not exceed
200mE
Tx/Rx loads should be as symmetrical
as possible
Benefits:
Simplified routing
Reduces risk of lost MSUs
Reduced risk of signalling loops
More efficient bearer usage
437 19 1
483 15 0 MSC3
Tx [mErl] Rx [mErl] Link No Destination
C7 Load C7 Load TSC1
Education Division
The table above serves as an example of unbalanced signal loading between a
TSC and an MSC in a live network. The Rx side of the two signalling links shown
is relatively unused whereas the Tx side is overloaded (exceeding the 200mE
threshold). Inefficient usage of C7 links is wasteful of C7 hardware resources.
This ideal symmetrical configuration, together with ensuring maximum
recommended thresholds are not exceeded, has the following benefits:
simplified routing
reduces the risk of lost MSU
reduces the risk of signalling loops being created.
More efficient bearer utilisation due to equal load Tx/Rx volume
spreading
Education Division
Total MSU per sec
0
500
1000
1500
2000
2500
3000
3500
4000
4500
5000
H
L
R
1

(
R
a
h
a
h
a
n
)
H
L
R
2

(
R
a
h
a
h
a
n
)
M
S
C
E
M
A
M
M
S
C
K
J
M
S
C
R
A
M
S
C
S
B

A
l
g
h
a
d
i
r

M
S
C
H
L
R
3

(
B
H
)
M
S
C
B
E
H
M
S
C
F
A
R
M
S
C
L
C
T
H
L
R
5

(
E
m
a
m
)
L
C
T
T
S
C
1
E
M
A
M
T
S
C
1
M
S
C
K
A
N
F
A
T
E
H
M
S
C
Exchange
M
S
U

p
e
r

s
e
c
Tx
Rx
Recommended Value per Node
H
L
R

1
H
L
R

2
M
S
C
1
M
S
C
2
M
S
C
3
M
S
C
4
M
S
C
5
H
L
R
3
M
S
C
6
M
S
C
7
M
S
C
8
H
L
R
4
T
S
C
1
T
S
C
2
M
S
C
9
M
S
C
1
0
Signal Load Evaluation Signal Load Evaluation Nodal MSU Capacity Nodal MSU Capacity
Recommended Nodal MSU Capacity <= 1500 MSUs/second
Education Division
Traffic Load Measurement Requirement Traffic Load Measurement Requirement
Recommended twice-daily 1-hour measurement
periods
Data gathered from MSCs at regular intervals to
assess capacity on each link
Trunk capacities assessed:
Between MSCs within the PLMN
From PLMN to PSTN
From PSTN to PLMN
Education Division
Traffic Load Data Capture Traffic Load Data Capture
Data captured should include:
Number of trunk available on each traffic link
Level of traffic (erlangs) on each link during
the busy hour
Answer Bit Ratio (ABR) measured on each
link during the busy hour
Education Division
Traffic Load Analysis Traffic Load Analysis
Analysis of the measured data should include:
Calculation of number of trunk required to support
traffic on link
Any difference between trunks required and trunks
available
Utilisation Factor for each trunk
Education Division
Recommended Traffic Parameters Recommended Traffic Parameters
Maximum erlangs per trunk = 0.7e
Maximum utilisation factor per trunk = 75%
GOS = 1%
Education Division
NSS Traffic NSS Traffic Trunk Trunk Utilisation Utilisation Example Example
180 84% 1280 1250.2 58% 98,192 1527 11-12 5
373 95% 1311 1281.3 58% 104,514 1375 11-12 4
34 87% 183 162.7 32% 12,324 210 11-12 3
5 77% 139 120.9 51% 6,783 180 11-12 2
154 83% 1207 1177.2 46% 128,394 1455 11-12 1
Required Utilisation Reqd Erlangs ABR% Attempts Trunks Time Trunk Gp
Trunks Trunk Trunks Total NSS NSS Call
Addit.
GoS = 1% Max Erlangs/trunk = 0.7 Only trunk utilisation > 75% shown
Education Division
Network Parameter Analysis Network Parameter Analysis
Parameters control equipment operation and
performance
Initial parameters determined at network planning
stage
Parameter sets should remain common where
possible to facilitate efficient network management
Parameters may require modification to suit local
conditions.
Periodic operation and performance monitoring
necessary to ensure parameters remain valid
Education Division
Parameter Analysis Parameter Analysis VLR Parameters VLR Parameters
Fax gr.3 and Alternate Speech (T61)/ Automatic fax group 3 (T62)
Short Message Mobile Terminated (T21)/Originated (T22)
TELESERVICE SUPPORT
Call Barring
Advice of charge charging (AOCC)
Call Hold (HOLD)
Call Waiting (CW)
Call Forwarding
Calling Line Identification Presentation (CLIP)
SUBSCRIBER SERVICES SUPPORT
Minimum Amount of Triplets to be kept on record (MINT)
VLR Cleaning Start Time (for Siemens If capacity above 80%)
Call Waiting Time Limits
Implicit IMSI Detach Time Limit Time
Emergency Call
Implicit Detach Usage
Authentication Usage
TMSI Usage
SUBSCRIBER PROFILE
Education Division
Parameter Analysis Parameter Analysis TMSI Parameters TMSI Parameters
Advantages of TMSI usage:
Increases paging channel (PCH) capacity
Reduces VLR/HLR signalling requirement
Provides greater security than using IMSI only
TMSI should be unique within a LA
New TMSI issued on change of LA
TMSI allocation/update occurrences set by Operator
Too many TMSI update occurrences increases signalling
capacity requirements
Use of TMSI should be uniform across network
Education Division
Parameter Analysis Parameter Analysis TMSI Parameters TMSI Parameters
N Mobile SS Operation
N Mobile Terminating USSD
N Mobile Terminating SMS
N Mobile Terminating Call
N Mobile Originating SMS COUNTERS
N Mobile Originating Call ALLOCATION
N Periodic Location Update TMSI
Y Location Update
Y IMSI Attach
Y Location Update New Visitor
REC PARAMETER
Education Division
Parameter Analysis Parameter Analysis Authentication Parameters Authentication Parameters
Advantages of Authentication usage:
Increases security of network
Prevents unauthorised network access
Implementation of Authentication is an Operator decision
Can be implemented internally to network and/or for
roaming subscribers
Excessive Authentication increases signalling capacity
requirements
Use of Authentication should be uniform across network
Education Division
Parameter Analysis Parameter Analysis Authentication Parameters Authentication Parameters
N Mobile Terminating Call
N Mobile Originating SMS
N Mobile Originating Call
N Periodic Location Update COUNT
N Location Update AUTHENTICATION
Y IMSI Attach
REC
PARAMETER
Education Division
Parameter Analysis Parameter Analysis IMEI Check Parameters IMEI Check Parameters
BLOCK Unknown IMEI Effect
TRACE Grey List Effect
BLOCK Black List Effect
N Mobile Originating SMS
N Mobile Terminating Call IMEI CHECKING ON.
N Mobile Originating Call
N Periodic Location Update
Y(10) Location Update
Y IMSI Attach
REC PARAMETER
Education Division
Reduces signalling load towards the BSS by
reducing unnecessary paging
Deregisters MS after period of inactivity
Compares timestamp value in subscriber profile
with inactivity timer.
Timer is reset when MSC-MS interaction takes
place
If timer value > timestamp MS is deregistered
and paging ceases
Generally at least 2x periodic update timer value
NSS NSS - - Implicit Deregistration Implicit Deregistration
Education Division
Database Usage Database Usage VLR Cleaning VLR Cleaning
Procedure to reduce loading within VLRs by
removing redundant subscriber records
Each subscriber record has an associated timer.
If timer expired before next contact with subscriber,
the subscriber record in marked for deletion
VLR cleaning takes place at periodic intervals
normally during periods of reduced activity.
All subscriber records marked for deletion a removed
during VLR cleaning process
Education Division
MSRN allocated by VLR for incoming call routing
Temporary Assignment
After call is connected, MRSN released back to
VLR for future use.
MSRN Lifetime refers to the time between MSRN
allocation and being releasing back to the network
Typical life times are:
Local Subscribers: 20 secs
Visiting Roaming subscribers: 90 secs
NSS NSS MSRN Lifetime MSRN Lifetime
Education Division
Time between call offered and call accepted
Network resources assigned during call waiting
time
If call not accepted within call waiting period,
network resources are released
Typical call waiting time: 25 secs
NSS NSS Call Waiting Time Call Waiting Time
Education Division
Database Usage Database Usage - - VLR VLR
VLR capacity should be evenly spread where possible
75% capacity limit is recommended to allow for peaks
and future expansion
Options for reducing VLR loading include:
Redistribution of BSCs amongst MSCs
Redesigning LA boundaries
Too large - excessive paging requirement
Too small - excessive location update requirement
Modify parameter settings to reduce signalling occurrences
VLR Cleaning
Implicit Deregistration
Education Division
NSS Databases NSS Databases VLR Utilisation VLR Utilisation
VLR utilisation refers to subscriber record
capacity and usage within a VLR
Utilisation problems include:
Over-capacity in certain VLRs
Uneven spread of records across the
network
Over-capacity can result in:
Call function failures (e.g. setup/handover,
authentication etc)
Lack of growth capability
7% MSC-10 10
62% MSC-9 9
100% MSC-8 8
71% MSC-7 7
66% MSC-6 6
86% MSC-5 5
100% MSC-4 4
26% MSC-3 3
82% MSC-2 2
65% MSC-1 1
VLR
Utilisation MSC Sr. No.
Education Division
Reduces capacity loading within a VLR (Nokia proprietary)
Deletes subscriber records that are inactive for a specified
period of time
Compares timestamp value in subscriber profile with
inactivity timer.
Timer is reset when sub scriber record is accessed
If timer value > timestamp, subscriber record is marked for
deletion
Cleaning takes place during non-busy periods
Common settings are typically 24hrs to 3 days
NSS NSS - - VLR Cleaning VLR Cleaning
Education Division
NSS Databases NSS Databases HLR Utilisation HLR Utilisation
HLR utilisation refers to
subscriber record capacity and
usage within a HLR
Utilisation problems include:
Over-capacity in certain HLRs
Uneven spread of records across
the network
Over-capacity can result in:
Call function failures (e.g.
setup/handover, authentication
etc)
Lack of growth capability
Signalling bottlenecks
0
10
20
30
40
50
60
70
80
90
100
HLR1 HLR2 HLR3 HLR4 HLR5
%AGE USAGE OF HLRs
%AGE USAGE
Education Division
Section 11
Section 11
Optimising Networks
Optimising Networks
for
for
New Services
New Services
Education Division
Section 11
Section 11
11.3 GPRS Performance Monitoring 11.3 GPRS Performance Monitoring
11.2 Dimensioning Networks For New Services 11.2 Dimensioning Networks For New Services
Education Division
Dimensioning a Multi Dimensioning a Multi- -Service System Service System
The Erlang B formula relies on the variance of the
demand equalling the mean (a Poisson distribution).
If a particular service requires more than one trunk
per connection, the demand is effectively linearly
scaled and the variance no longer equals the mean.
Methods to investigate:
Equivalent Erlangs
Post Erlang-B
Campbells Theorem
Education Division
Equivalent Erlangs Equivalent Erlangs
+
Low
Bandwidth
Equivalent
High
Bandwidth
Equivalent
Combine the two traffic sources together by
converting one to the bandwidth of the other
The trunking efficiency will VARY with the
bandwidth of equivalent Erlang that you choose!
Not suitable for use due to this property
2 Erlangs
of Low
Bandwidth
1 Erlang of
High
Bandwidth
Difference in
capacity
required for
same GoS
Education Division
Equivalent Erlangs Example Equivalent Erlangs Example
Consider 2 services sharing the same resource:
Service 1: uses 1 trunk per connection. 12 Erlangs of traffic.
Service 2, uses 3 trunks per connection. 6 Erlangs of traffic.
We could regard the above as equivalent to 30 Erlangs of
service 1:
30 Erlangs require 39 trunks for a 2% Blocking Probability
Alternatively, we could regard the above as equivalent to 10
Erlangs of service 2.
10 Erlangs require 17 trunks, (equivalent to 51 service 1 trunks) for a
2% blocking probability
Prediction varies depending on what approach you choose.
Education Division
Post Erlang Post Erlang- -B B
Combine the two traffic
sources together after
calculating required
capacity
The trunking efficiency
variation with magnitude
is not considered -
pessimistic about offered
traffic supported to the
same GoS
Not suitable for use due
to this property
1 Erlang of
Service A
1 Erlang of
Service B
+
1 Erlang and 1
Erlang of of
Service B
Illustration using 2 services of
same bandwidth
Difference
in capacity
required for
same GoS
Accepted correct method
Education Division
Consider 2 services sharing the same resource:
Service 2: uses 3 trunks per connection. 6 Erlangs of traffic.
We could calculate the requirement separately
Service 1: 12 Erlangs require 19 trunks for a 2% Blocking Probability
Service 2: 6 Erlangs require 12 trunks (equivalent to 36 service 1
trunks).
Adding these together gives 55 trunks.
This method is known to over-estimate the number of trunks
required as can be demonstrated by considering services
requiring an equal number of trunks.
Education Division
Consider 2 services requiring equal resource:
We could calculate the requirement separately
Service 1: 12 Erlangs require 19 trunks for a 2% Blocking Probability
Service 2: 6 Erlangs require 12 trunks.
Adding these together gives 31 trunks.
The accepted method of treating the above would be to regard
it as a total of 18 Erlangs that would require 26 trunks.
Post Erlang-B overestimates the requirement.
Education Division
Campbell Campbell s Theorem s Theorem
Campbells theorem creates a composite distribution where:
The amplitude (a
i
) used in the capacity is the amplitude of the target service
Once the equivalent offered traffic and capacity are derived, GoS can be
derived with Erlang-B
Required capacity can be calculated if offered traffic and GoS target is
known
( )
c
a C
Capacity
i i

=
c
ffic OfferedTra

=
= =
i
i i i
i
i i i
b a
b a
c
2
C
i
= available capacity
= mean
= variance
i
= arrival rate
a
i
= amplitude of service
b
i
= mean holding time
c = capacity factor
i i
b = Traffic Offered Service
Education Division
Campbell Campbell s Theorem Example s Theorem Example
Consider the same 2 services sharing the same
resource:
Service 2, uses 3 trunks per connection. 6 Erlangs of traffic.
In this case the mean is:
The variance is:
Note: = traffic in Erlangs

= + = = = 30 ) 6 3 ( ) 12 1 ( Erlangs
i i i i
a a b

= + = = = 66 ) 3 6 ( ) 1 12 ( Erlangs
2 2 2 2
i i i i
a a b
i i
b
Education Division
Capacity Factor c (average trunks per connection):
Equivalent offered traffic:
Trunk capacity for equivalent traffic at 2% GoS = 21
2 . 2
30
66
= = =
c
63 . 13
2 . 2
30
Traffic Offered Equivalent = = =
c

Education Division
Actual trunk requirement is trunk capacity x capacity
factor:
= 21 x 2.2 = 46.2 or 46 trunks
Required Capacity is modified depending upon target
service for GoS (in service 1 Erlangs):
Target is Service 1 C
1
=(2.2 x 21) + 1 = 47
Target is Service 2, C
2
=(2.2 x 21) + 3 = 49
Different services will require a different capacity for the
same GoS depending on the amplitude
Education Division
Traffic Analysis Methods Compared Traffic Analysis Methods Compared
Equivalent Erlangs
Optimistic if you use the smallest amplitude of trunk (39)
Pessimistic if you use the largest amplitude of trunk (51)
Post Erlang-B
Pessimistic (55)
Trunking efficiency improvement with magnitude ignored
Campbells theorem
Middle band (47 - 49)
Different capacities required for different services - realistic
Preferred solution for dimensioning, but not ideal...
Education Division
Capacity Dimensioning with Campbell Capacity Dimensioning with Campbell s s
Theorem Theorem
Consider the following service definition and traffic
forecast.
Service Amplitude Forecast
Voice 1 250 E
HSCSD 2 63 E

Based on a theoretical availability of 15 voice trunks per
cell and using voice as the benchmark service and with
2% blocking, determine the number of cells required to
serve the above traffic levels and the traffic offered per
cell for each service
Education Division
Theorem Theorem
Assuming we have n cells, we can determine the loading
per cell.
n n c
c
n n n
n n n
282
335 . 1
376 mean
traffic offered
335 . 1
376
502
mean
variance
502 2 63 250
variance
376 2 63 250
mean
2
=
= =
= = =
=
+ =
=
+ =
Education Division
Theorem Theorem
Considering the equation:
C
i
(available capacity) is predefined as 15. a
i
(amplitude)
depends on the service we use as our benchmark or priority
service. Choosing the voice service as the benchmark
service make a
i
equal to 1.
10.5 (rounded to 10) trunks will service 5.08 Erlangs.
c
a C
i i

= Capacity
( )
5 . 10
335 . 1
1 15
=
=
i
C
Education Division
Theorem Theorem
Each cell requires 10 trunks to service 5.08 Erlangs.
Therefore:
Number of cells required = Equivalent Traffic
Traffic per Cell
=
Cell requirement is established at 56 cells.
Each of the cells will service:
4.46 Erlangs of voice
1.13 Erlangs of HSCSD.
5 . 55
08 . 5
282
=
Education Division
Assessing Cell Loading using Campbell Assessing Cell Loading using Campbell s s
Theorem Theorem
After placing sites on the coverage map and
spreading the traffic, the next stage is to assess the
cell loading (timeslots required).
If mixed services are used, it is necessary to use
Campbells Theorem to assess the required number
of timeslots to satisfy the likely demand.
Consider the case where a particular cell captures 7
Erlangs of voice and 2 Erlangs of HSCSD traffic that
requires 2 timeslots per connection.
Education Division
Assessing Cell Loading using Campbell Assessing Cell Loading using Campbell s s
Theorem Theorem
Using Campbells Theorem:
Hence 20 timeslots required.
( ) 20 1 36 . 1 14 : benchmark as voice Taking
required. trunks 14 B, Erlang From
Erlangs 09 . 8
36 . 1
11
traffic offered Equivalent
36 . 1
11
15
) ( Factor Capacity
15 ) 2 2 ( ) 1 7 ( variance
11 ) 2 2 ( ) 1 7 ( mean
2 2
= +
= =
= =
= + =
= + =
c
Education Division
Evaluating Traffic Requirements Evaluating Traffic Requirements
Real time non-controllable load
Spare capacity for which can
be allocated to non real time
applications
Peak traffic
Load
Time
Average circuit
switched traffic
Data may be packet switched, in which case it can be made to fill the
gaps in the demand for voice services.
Education Division
Evaluating Traffic Requirements Evaluating Traffic Requirements
To evaluate the required cell capacity:
First assume that the packet data can be scheduled to fill the spare real time
capacity.
When all the spare real time capacity has been exhausted we must convert the
remaining capacity to an equivalent data capacity.
One GSM timeslot can carry 13 kb/s of data.
Education Division
Traffic Requirement Example Traffic Requirement Example
A cell captures 2 Erlangs of voice traffic and is assigned a GSM carrier
with 7 timeslots. The grade of service is 2%.
Estimate the amount of data traffic that can be handled.
How much of this must be packet data and how much can be circuit
switched?
Solution:
Timeslots available on average = 7 - 2 = 5
This represents a total data rate of 5 x 13 = 65 kb/s
2 Erlangs voice traffic requires 6 trunks (timeslots) for 2% blocking.
1 timeslot can be dedicated (circuit switched) for data, i.e. 13 kb/s
Remaining data must be packet switched = 65 - 13 = 52 kb/s.
Education Division
Examples of GPRS Examples of GPRS- -Specific Parameters Specific Parameters
Cell Reselection
Service Exceptions
Average Response Times
Suspend/Resume Procedures
Attach/Detach Procedures
Routing Area (RA) Update Procedures
PDP Context Activations
Data Volumes and Rates
Education Division
GPRS Service Exceptions GPRS Service Exceptions
Radio Link Signal Strength falls below
threshold (link failure)
Link quality falls below threshold (link failure)
Cell Reselection initiated
Education Division
GPRS Transmission Connections GPRS Transmission Connections
SGSN SGSN
MS MS
GGS GGS
N N
BSS BSS
G G
b b
G G
n n
U U
m m
PDP Context PDP Context
Logical Link Control Logical Link Control
Connection Connection
Virtual Tunnel Virtual Tunnel
Data Link Connection Data Link Connection
IMSI/NSAPI IMSI/NSAPI
TLLI TLLI
DLCI DLCI
TID TID
CONNECTION TYPE CONNECTION TYPE
CONNECTION CONNECTION
IDENTIFIER IDENTIFIER
Education Division
GPRS Cell Reselection GPRS Cell Reselection
Cell Reselection per source cell
Cell Reselection Subscribers in READY state
Number of cell
reselections
Number of
Octets
Average Number of Octets Affected
Average Number of Octets Deleted
Average Number of Frames Deleted
Number of Cell Reselections
| 0000100002 | 0000100003 | CI
Education Division
Frames Lost Through Service Exceptions Frames Lost Through Service Exceptions
Number of frames deleted per TLLI
Number of Frames
Deleted
Education Division
GPRS Average Response Times GPRS Average Response Times
Average response time after paging, per BVCI
Time in millsecs
BVCI
Education Division
Suspend / Resume Procedures Suspend / Resume Procedures
Suspend Procedure:
Enables a GPRS Class B MS to suspend GPRS mode in
order to make a circuit-switched connection
Resume Procedure:
Enables a GPRS Class B MS to resume GPRS mode having
made a circuit switched connection
Monitoring:
Procedure can be monitored and failures can be displayed
by TLLI, cell or cause
Education Division
GPRS Attach/Detach Procedures GPRS Attach/Detach Procedures
Number of Attach Procedure Messages
Attach Request Attach Accept Attach Complete Attach Reject
Number of
Procedures
Attach Procedures
Education Division
Number of Attach Reject Messages per Cell
Attach Rejects per Cell Number of
Attach Rejects
Education Division
Number of Attach Reject Messages per Subscriber (TLLI)
Number of
Attach Rejects
Attach Rejects per TLLI
Education Division
GPRS RA Update Procedures GPRS RA Update Procedures
Number of RA Update Messages
RA Updates
RA Request RA Accept RA Reject RA Request without response
Number
Education Division
GPRS RA Update Procedures GPRS RA Update Procedures
Number of RA Update Rejects by Cause
Number of
RA Rejects
RA Rejects per Cause
Unknown Causes Network Failures
Education Division
GPRS PDP Contexts GPRS PDP Contexts
PDP Context Activation by MS
PDP Contexts Activated PDP Contexts Accepted PDP Contexts Rejected
Number PDP
Context
Activations
Number of PDP Context Activation Messages
Education Division
GPRS Data Volumes and Rates GPRS Data Volumes and Rates
Downlink Data Rate of a Cell
DL LLC Rate
Bytes
Time (minutes)
Education Division
GPRS Data Volumes and Rates GPRS Data Volumes and Rates
Downlink Data Rate of a Mobile Station
Time (minutes)
Volume (bytes)
DL LLC Global Size
Education Division
End of Course
End of Course
Any Questions?
Any Questions?
Education Division
Thank You Thank You Thank You Thank You
Brian
0811800402
radio_planner@yahoo.com

RF Parameter & Optimization Document

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

RF Parameter & Optimization Document

Uploaded by

Copyright:

Available Formats

Brian - CommServ Radio Frequency Management & Optimization

Brian - CommServ Radio Frequency Management & Optimization

You might also like