-
This document is the property of Astellia and can not be
reproduced in any form or conveyed to a third party without the
prior written permission of Astellia. All trademarks, service
marks, registered trademarks or registered service marks mentioned
in this document are the property of their respective owners.
Astellia Audit December 2011
Preliminary Voice performance report
A and IuCS interfaces: focus on call drop
Order acknowledgement: AR20001119
-
This document is the property of Astellia and can not be
reproduced in any form or conveyed to a third party without the
prior written permission of Astellia. All trademarks, service
marks, registered trademarks or registered service marks mentioned
in this document are the property of their respective owners.
Written by Reviewed by
Name Yann LE GAL / Jerme JAMIN Eric GATEL
Company ASTELLIA ASTELLIA
Job title Telecom Engineer Consulting Project Manager
Date 13/12/2011 13/12/2011
Signature
Recipients
Telenor
- Peter FEKETE
- Lszl TOBIAS
- Bela NAGY
Astellia :
- Jerme JAMIN
- Yann LE GAL
- Jean-Pierre THOMAS
- Enrico BASSO
- Eric GATEL
Releases
Release Date Author Purpose
A 13/12/2011 Yann LE GAL Jrme JAMIN
Creation
-
Astellia Voice audit report Page: 3 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
1. TABLE OF CONTENTS
1. TABLE OF CONTENTS
.............................................................................................................................
3 2.
INTRODUCTION.........................................................................................................................................
4 3. VOICE SERVICE PERFORMANCE
...........................................................................................................
5
3.1. 3G VOICE CALL DROP (IUCS)
.......................................................................................................
5 3.1.1. VOICE CALL DROP METRICS (IUCS)
.....................................................................................
5 3.1.2. VOICE CALL DROP NETWORK (IUCS)
...................................................................................
6 3.1.3. VOICE CALL DROP DETAIL OF CAUSES (IUCS)
......................................................................
7
3.2. 2G VOICE CALL DROP
.................................................................................................................
15 3.2.1. VOICE CALL DROP METRICS (A)
.......................................................................................
15 3.2.2. VOICE CALL DROP NETWORK (A)
......................................................................................
16 3.2.3. VOICE CALL DROP DETAIL OF CAUSES (A)
.........................................................................
17
3.3. HANDOVERS
.................................................................................................................................
24 3.3.1. 3G-2G HO EFFICIENCY NETWORK (IUCS)
...........................................................................
24
-
Astellia Voice audit report Page: 4 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
2. INTRODUCTION
TELENOR HUNGARY has selected Astellia to perform an E2E audit in
order to evaluate the quality of performances of its both 2G
&nd 3G networks. Objective of this audit is to assist TELENOR
HUNGARY in improving its 3G, 2G and 2.5G network after the swap to
ZTE. Despite good results providing by ZTE OSS counters, Telenor is
now facing an increase of the customer complaints This first
preliminary report focuses on the call drop part, based on A and
IuCS interfaces,. An additional analysis for handovers is also
included in this report. This audit is based on the analysis of raw
data captured through Astellia probe on A and IuCS interfaces, for
the BSC ZBSC51 and RNC ZRNC11.
Schedule for the capture phase was the following:
Site Node name Capture Begin Capture End Duration Comments
Budapest ZBSC51 ZRNC11
25/11/2011 17:00
27/11/2011 23:59
55H
Capture of A and IuCS 76GB of signalling raw data was
retrieved
-
Astellia Voice audit report Page: 5 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3. VOICE SERVICE PERFORMANCE
3.1. 3G VOICE CALL DROP (IUCS)
3.1.1. VOICE CALL DROP METRICS (IUCS)
0.44% of the established calls are dropped due to NSS reason.
Around 1.6% of the subscribers have a call drop rate higher than 1%
due to NSS inefficiency.
0.23% of the established calls are dropped due to BSS reason.
0.91 % of the subscribers have a call drop rate higher than 1% due
to BSS inefficiency.
More details about the inefficiency causes are provided below in
the report.
-
Astellia Voice audit report Page: 6 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.1.2. VOICE CALL DROP NETWORK (IUCS) This graph displays the
call drop rate. On the left Y-axis you have the call drop rate and
on the right Y-axis you can see the number of CS voice calls.
The call drop efficiency is good. During the busiest periods of
the day the call drop rate is around 0.8% if we dont take User
causes into account. The main source of drop inefficiency is due to
NSS/CN causes and more details about the causes are provided on the
next pages.
-
Astellia Voice audit report Page: 7 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.1.3. VOICE CALL DROP DETAIL OF CAUSES (IUCS) This graph
displays the call drop causes. On the left Y-axis you have the
percentage of RNS call drop causes and on the right Y-axis you can
see the number of established calls.
The Call Drop due to RNS causes is lower than 0.3%. This value
is constant during the whole day. The two main causes are Failure
in the Radio Interface Procedure and Radio Connection With UE
Lost.
-
Astellia Voice audit report Page: 8 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
This graph displays the call drop causes. On the left Y-axis you
have the percentage of CN call drop causes and on the right Y-axis
you can see the number of established calls.
The Call Drop due to NSS/CN causes is around 0.4%. This value is
constant during the whole day. The two main causes are Destination
Out Of order
-
Astellia Voice audit report Page: 9 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
This graph displays the top 40 cells with highest call drop
rates. On the left Y-axis you have the percentage of call drop
causes and on the right Y-axis you can see the number of
established calls (filter > 50 occurrences).
The worst cell is the SAI 2902 with around 15.5% of call drop
rate. Its interesting to investigate each cell in order to
understand high call drop rates.
-
Astellia Voice audit report Page: 10 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
Radio Connection with UE lost This graph displays the 40 cells
having highest cause radio connection with UE lost drop rate. On
the left Y-axis you have the call drop rate and on the right Y-axis
you can see the number of established calls (>50
occurrences).
When the mobile goes out of radio coverage the connection is
released with this cause. No specific cell is concerned.
-
Astellia Voice audit report Page: 11 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
CLREQ Radio Interface Failure (RIF) This graph displays the 40
cells having highest cause radio interface failure drop rate (RIF).
On the left Y-axis you have the call drop rate and on the right
Y-axis you can see the number of established calls (>50
occurrences).
If we focus on the Failure in the Radio interface procedure, we
can see the following call flow showing the loss of the radio link
during the call.
-
Astellia Voice audit report Page: 12 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
When the radio connection is not good enough to maintain the
communication, the node B when one single radio link is available
will release the connection with the Mobile. No specific cell is
concerned.
-
Astellia Voice audit report Page: 13 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
DISC_F Destination Out Of Order (DOOO) o In the first 2 lines,
the calling party (A: +362093021) calls the number B: 204981185
with a failure end.
The customer A is in the cell 4852 and the Customer B in the
cell 3961
The call drops at 18h:26min 28s 529ms with the loss of the radio
link for A (Failure in the Radio interface, cause investigated in
previous chapter) and the cause Destination out of order for B
Call flow for A Call flow for B
-
Astellia Voice audit report Page: 14 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
o In the last 2 lines, the calling party A has changed of cell;
the customer A is now in cell 4853 and the calls is a success. So
bad radio quality is responsible of such a message for the mobile
A. For B this cause is the consequence of the loss of the radio
link for the side party.
This cause is generated when a radio problem on the other side
during the call.
-
Astellia Voice audit report Page: 15 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.2. 2G VOICE CALL DROP
3.2.1. VOICE CALL DROP METRICS (A)
0.3% of the established calls are dropped due to BSS reason.
Around 1% of the subscribers have a call drop rate higher than 1%
due to BSS inefficiency.
0.57% of the established calls are dropped due to NSS reason.
1.87% of the subscribers have a call drop rate higher than 1% due
to NSS inefficiency.
The call drop rate is very low and more details about the
inefficiency causes are provided below in the report.
-
Astellia Voice audit report Page: 16 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.2.2. VOICE CALL DROP NETWORK (A) This graph displays the call
drop rate. On the left Y-axis you have the call drop rate and on
the right Y-axis you can see the number of established calls.
The call drop efficiency is good. During the busiest periods of
the day the call drop rate is lower than 1.5%. The main source of
drop inefficiency is due to NSS causes and more details about the
causes are provided on the next page.
-
Astellia Voice audit report Page: 17 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.2.3. VOICE CALL DROP DETAIL OF CAUSES (A) This graph displays
the call drop causes. On the left Y-axis you have the percentage of
call drop causes and on the right Y-axis you can see the number of
established calls.
Around 0.3% of the inefficiency is due to the causes Radio
Interface Failure (RIF) and Radio Interface Message Failure (RIMF),
in other words due to bad radio condition (classified as BSS
inefficiency). It is also important to note than 0.3% of the
established calls are dropped due to the cause Destination Out Of
Order (classified as NSS inefficiency) and 0.5% of the established
calls are dropped due to the cause Normal Unspecified (classified
as a NSS inefficiency).
-
Astellia Voice audit report Page: 18 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
CLREQ Radio Interface Failure (RIF) and Radio Interface Message
Failure (RIMF) This graph displays the 20 cells having highest
radio call drop rate (RIF & RIMF). On the left Y-axis you have
the call drop rate and on the right Y-axis you can see the number
of established calls.
The highlighted cells above have an important call drop rate and
we recommend checking the global radio condition for these cells.
Deeper Abis investigations could be useful to understand the issue.
Analysing Outgoing handover causes, Astellia KPI could help to
focus on the radio problem.
-
Astellia Voice audit report Page: 19 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
Here are more details about radio condition based on Outgoing
handover causes:
Causes of outgoing handover
Cell Nb of established calls Drops RIF+RIMF(%) UQ US DQ DS BC
Possible radio
problem
dujvatanq8_3512_15498 2139 2,9453 2,69% 2,78% 0,15% 0,38% 94,00%
/
kvolwest3_3511_7823 721 2,91262 2,52% 9,63% 5,45% 6,85% 75,54%
Radio signal strength
in both ways
dujvatan3_3512_15493 882 2,60771 1,46% 1,01% 1,34% 2,10% 94,09%
/
kvolwest1_3511_7821 603 2,32172 5,94% 20,18% 8,66% 4,85% 60,37%
Uplink strength
problem
biatorbag2_3511_21012 1249 2,08167 1,34% 2,77% 1,07% 6,66%
88,16% Downlink strength
problem
ormezo2_3511_1882 518 1,5444 1,73% 1,15% 0,93% 2,18% 94,00%
/
rokales1_3511_7471 1050 1,2381 7,16% 8,84% 4,27% 10,24% 69,49%
Radio Quality and Strength problem
dujvaros2_3512_15112 725 1,10345 5,23% 7,53% 7,81% 6,13% 73,28%
Quality problem in
both ways
kikoto7_3512_7947 646 0,928793 7,40% 10,91% 4,46% 1,44% 75,80%
Uplink way problem
bfokwest3_3511_5793 761 0,919842 4,45% 7,80% 3,67% 8,74% 75,31%
Radio strength
problem in both ways
soroksmil2_3512_7322 563 0,888099 1,49% 2,38% 11,38% 6,60%
78,12% Downlink problem
erdeszak1_3511_61181 571 0,875657 3,16% 2,03% 4,97% 14,15%
75,69% Downlink strength
problem
menyecsk2_3511_6812 916 0,873362 1,35% 0,40% 2,04% 0,78% 95,42%
/
borstesc1_3511_41471 581 0,860585 1,34% 0,93% 1,48% 2,58% 93,67%
/
szhalwat3_3512_61233 2114 0,851466 4,37% 2,47% 4,52% 7,82%
80,43% Quality problem in
both ways
szoborpar1_3511_3381 707 0,848656 8,36% 2,30% 8,72% 8,83% 71,74%
Quality problem in
both ways
budaors3_3511_21153 1071 0,840336 3,32% 2,06% 1,62% 4,26% 88,71%
/
highwayq8_3511_1828 601 0,831947 1,04% 3,17% 0,86% 3,10% 91,78%
/
m7m02_3511_41972 970 0,824742 2,64% 2,80% 4,89% 11,38% 78,29%
Downlink strength
problem
harbor3_3511_7563 981 0,815494 1,48% 10,47% 5,55% 9,06% 73,44%
Radio strength
problem
-
Astellia Voice audit report Page: 20 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
DISC_F Normal Unspecified (Nu) During the conversation phase a
disconnect message associated with the cause Normal Unspecified is
sent from the MSC site. The following pie show the state when the
disconnect message is sent:
Here are some established calls CDRs which have been dropped
with the cause Normal Unspecified:
Most of the case this cause appears for terminating established
calls (82%).
It clearly appears most of the failed procedure are established
calls towards T-Mobile subscribers and more particularly for
terminating call procedure.
To investigate this cause some Core CS investigations could be
useful and Astellia recommends checking the ISUP efficiency by
switches couples between T-Mobile and Telenor.
-
Astellia Voice audit report Page: 21 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
Here is the associated flow chart:
-
Astellia Voice audit report Page: 22 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
DISC_F Destination Out Of Order (DOOO) During the conversation
phase a disconnect message associated with the cause Destination
Out Of Order is sent from the MSC site. Here is an end to end
procedure based on the A interfaces capture:
It was possible to match the outgoing and terminating procedures
thanks to a specific feature implemented on your network
illustrated below:
For terminating call procedures, the setup message contains the
MSISDN of the calling and the called parties.
-
Astellia Voice audit report Page: 23 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
Here are the flow charts of the analyzed end to end case:
Calling party: Called party:
In this example, during the conversation phase the calling party
lost the radio coverage (failure with the cause RIF). The MSC
released to procedure on the called site with the associated cause
Destination Out Of Order. This cause is classified in the NSS
inefficiency and based on standard the associated cause should be
Normal Clearing.
The call is connected and the
communication phase begins
The call is dropped with the cause RIF on the calling party
and DOOO on the called party
-
Astellia Voice audit report Page: 24 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.3. HANDOVERS
3.3.1. 3G-2G HO EFFICIENCY NETWORK (IUCS) The 3G/2G HO
efficiency can be divided into 2 phases.
This graph displays the 3G-2G HO failure rate.
-
Astellia Voice audit report Page: 25 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
3.3.1.1. During the preparation phase With a decreasing filter
on the Preparation Failure rate per 3G cell and for a number of
relocation procedures> 10. We find a lot of cells with high rate
of Preparation Failure
-
Astellia Voice audit report Page: 26 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
The first column shows us rates of failure higher than 30%. With
the associated Call Flow, we can find that the first time this
procedure of preparation is not successfully achieved but this
procedure can be repeated and the second time, the mobile ends its
communication in 2G. An interruption of around 2s during the
communication can be estimated.
-
Astellia Voice audit report Page: 27 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
The most concerned couple SAI 3G/Cell 2G are the following
-
Astellia Voice audit report Page: 28 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
So some couples have a high failure rate. The main cause is
RANAP ReLOCation Preparation FAILure , cause: RADio
RelocationFailureInTargetCNRNCorTargetSystem for 90%.
This cause is mainly due to a lack of resources on the target 2G
Radio system
-
Astellia Voice audit report Page: 29 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
With a multi interface analysis Iucs/A (see next page), we can
find that with a focus on preparation phase with problem of
RelocationFailureInTargetCNRNCorTargetSystem
During the 3G to 2G mobility, we can see a handover failure on
the A interface with the cause NO RADIO RESOURCE AVAILABLE o GSM
resources are not available for the 3G call
So the mobile cant go to the 2G, a handover failure is sent to
the CN with the cause nor resource available The mobile must
continue its communication on 3G but the Core Networks sends a
DISCONNECT with the cause Normal clearing
o This is a abnormal behaviour as the call is drop with a normal
message-cause
-
Astellia Voice audit report Page: 30 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
But Abnormal message is sent by the Core Networks to end
the Call
After the HO failure due to a lack of GSM resources, the
relocation preparation failure is sent
-
Astellia Voice audit report Page: 31 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
The number of HO is quite important in Telenors networks. An
interruption during the call can be suspected (An interruption of
around 2s during the communication can be estimated) and this high
number needs explanation, bad radio threshold? Bad radio coverage?
(Iub analysis could help for better understanding) We have around
0.9% of failure during the preparation phase. Most of the problems
during the preparation Phase are due to a lack of GSM resources.
The Call drop rate must be impacted by the couple Disconnect-
Normal Clearing sends to drop the call when mobility fails.
Investigation is needed with the vendor to understand such
behaviour.
3.3.1.2. During the completion phase
-
Astellia Voice audit report Page: 32 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
The most concerned couple of cells are the following
The main cause of failure for the Completion phase is RANAP
ReLOCation CANcel , cause: MISCellaneous UnspecifiedFailure
-
Astellia Voice audit report Page: 33 / 33
Model: AuditReport - Release: B AR20001119_Astellia_Preliminary
Voice auditreport_RA
We observe around 0.7 % of failure during the Completion phase.
Most of the problem is due to the cause unspecified. Additional
investigation on UTRAN/IUB is needed.