Professional Documents
Culture Documents
Use case
Tool:
Specification number:
Title:
ANATOM
3d
Related request Nr:
Detection of unnecessary neighbour relation
Author:
Date:
Description Version:
Relevant Specification
Marek Lipski
10.12.2004
3.0
Department:
Phone:
Description State:
Com SDC NA B1
+48 71 7992362
IUS
Title
Date
Number
Version
Author
Page
Error: Reference
source not found
Error: Reference
source not found
Error: Reference
source not found
1 of 17
SIEMENS
Table of contents
1
General...........................................................................................................................3
1.1
Contributions................................................................................................................. 3
1.2
Document history.......................................................................................................... 3
1.3
Abbreviations................................................................................................................. 3
1.4
References..................................................................................................................... 3
1.5
Hints................................................................................................................................ 4
Objective........................................................................................................................5
3.1
3.2
3.3
3.4
3.5
3.6
3.7
Problem detection.........................................................................................................9
4.1
Problem description......................................................................................................9
4.2
Used thresholds...........................................................................................................10
4.3
Problem reporting........................................................................................................11
4.3.1
Title
Number
Author
Geographical analysis............................................................................................................... 16
Date
Version
Page
2 of 17
SIEMENS
1 General
1.1 Contributions
In addition to the authors named on the cover page, the following persons have collaborated on
this document:
D. Blechschmidt
A. Rocchetti
K. Majchrowicz
M. Chrostek
A. Sikora-Godlewska
S. Lasek
Com MN PG NT NE 5
Com MN PG NT NE 5
Com SDC NA B1
Com SDC NA B1
Com SDC NA B1
Com SDC NA B1
Date
15.07.2003
15.03.2004
2.1
31.05.2004
2.2
2.2
3.0
23.06.2004
06.07.2004
10.12.2004
Change
Initial version
Second version, upgraded to ANATOM 3.7 using its new
functionalities
Main formula "DEL_NB" changed, minor changes in text of
document
Small changes according to MS P
Set to IUS
IUS version:
use case analysis including geographical reporting is based on
ANATOM 3.9.3
1.3 Abbreviations
ADJCN
BSC
BSCN
BTS
BTSN
CELLGLID
CM
HO
KPI
NB
1.4 References
[1]
[2]
[3]
[4]
Title
Number
Author
Date
Version
Page
3 of 17
SIEMENS
1.5 Hints
Formulas as denoted in chapter 3 can be copied directly into ANATOM Formula Editor to be used
for analysis. All necessary files to perform the analysis are in the table below:
Exported KPIs:
AnatomKPIExport_us
er_nb_del.sql
Title
Number
Author
Report template:
3d_Detection_of_un
necessary_NB_relation.arc
Date
Version
Page
4 of 17
SIEMENS
2 Objective
The adjacency plan is one of the major contributors to network quality. Without the definition of the
optimum adjacencies, handovers may occur to cells, which are not optimal, and subsequently
a call drop might not be avoided. The goal of adjacency management is to have the right cell in the
neighbour list, to provide always the best target cell to perform a handover.
Based on performance measurement data provided through the network and the configuration of
the network the operator wants to analyse if specific cells have unnecessary neighbour relations.
To keep the best HO results some performance metrics such as total number of HO attempts,
relative number of HO attempts per neighbour and number of successful handovers should be
analysed.
NOTE: This document does not describe the whole process of neighbour optimisation. Its main
purpose is to help the user while looking for possible database inconsistencies in
neighbour definition. The use case is meant to be only the rough user guide so the user
can easily find problematic adjacencies in the network. The output of the use case should
only attract user's attention to possible improvements. The user must be aware of the fact
that the analysis should be checked by drive-tests performed before and after deleting the
potentially unnecessary neighbour.
Title
Number
Author
Date
Version
Page
5 of 17
SIEMENS
Formula:
Remarks:
KPIs used in
formula:
Anatom category:
Aggregation:
Remarks:
KPIs used in
formula:
Anatom category:
Aggregation:
Title
Number
Author
6 of 17
SIEMENS
Remarks:
KPIs used in
formula:
Anatom
category:
Aggregation:
7 of 17
SIEMENS
Description:
Formula:
Remarks:
KPIs used in
formula:
Anatom category:
Aggregation:
8 of 17
SIEMENS
4 Problem detection
4.1 Problem description
The important matter of the deletion process is to identify the required criteria. Any existing
neighbours that do not meet these conditions should be deleted from the neighbour list.
Handovers to most suitable cells are possible only with the definition of correct adjacencies. The
basic reasons for triggering a handover is better power budget provided through coverage from
neighbouring cells. To maintain a call also handovers due to other reasons might be invoked. Socalled emergency handovers are triggered by low receive level, bad quality or high distance from
the serving BTS. To perform HO, the target cells need to be evaluated using adjacent cell
measurements. Therefore, adjacent cell measurements are carried out for all the registered target
cells and a HO should be initiated to the best-suited one.
The analysis of neighbour relations' validity should cover certain period. The operator should be
aware of the fact, that long averaging windows tend to reduce the number of unnecessary
handovers due to traffic fluctuations. That is why the optimum period should be about a week,
because only then it is possible to detect badly or under-performing neighbours.
Second parameter, which should be taken into consideration, is number of handover attempts. If
there are no HO attempts in observed cell, then there is no need to analyse it.
Next important thing is to recognise potentially unnecessary neighbour by comparing HO attempts
from a source cell to a neighbour cell with all outgoing handovers to all neighbour cells. This ratio is
defined as HO_ATT_REL and if its value is below the desired threshold, it means that the observed
neighbour is not a preferred candidate for handovers and it could be considered to be deleted. On
the other hand, if HO_ATT_REL has reached the threshold such a neighbour should be kept in
the NB list. This high value indicates that the observed neighbour has been measured as a good
candidate and it should remain in the NB list.
And, last but not least, very important parameter is handover success rate. This metrics is
evaluated as a ratio of successful handovers to all attempted handovers per neighbour cell. If this
indicator is lower than the desired threshold, it means that the cell should be analysed with more
attention. The low success rate can happen in case of different factors such as interference,
wrong HO parameters or missing/unused adjacencies.
The user cannot forget to check the distances between adjacencies from geographical point of
view as well.
The process of detecting the unnecessary neighbour relation based on above metrics and rules is
shown in a diagram below:
Title
Number
Author
Date
Version
Page
9 of 17
SIEMENS
HO_ATT_BTS
>
TH1
NO
YES
HO_ATT_RELATIVE
<
TH2
NO
YES
DELETE NB
RELATION
KEEP NB
RELATION
Attention: The handover relation from cell A to cell B can be deleted only if the process
shows the same result for handover relation from cell B to cell A (A B = B A).
Threshold (TH1)
> 500
Threshold (TH2)
>
HO_ATT_RELATIVE
1
NUMBER_ADJC 50
Title
Number
Author
Date
Version
Page
10 of 17
SIEMENS
Formula
Editor
Applied
threshold
It is very important to re-calculate all modified KPIs every time the user changes the values of the
thresholds. Another way, the changes will not be visible in the Universal Report.
The report contains information about Cell Global Identifier of neighbour cell and number of
handover attempts per BTS as well as per neighbour cell. The number of handover attempts per
neighbour cell compared with HO attempts to all cells is also displayed together with handover
success rate per neighbour as additional information. The last column contains the result of the
main rule.
The lines are grouped by BSC, BTS and then Adjacent Cells.
In the above example the user can see that for all selected cells belonging to BSC0 the hint
whether delete NB relation is reported 109 times. To see the details, the user can expand the grid
to the BTS level and he will see the following details.
Title
Number
Author
Date
Version
Page
11 of 17
SIEMENS
For example, for BTSN 1.022 the DEL_NB flag was calculated 9 times and for BTSN 1.015 the flag
was calculated 18 times. Further, the user willing to see for which neighbours the rule was fulfilled,
he opens the evaluated BTS and checks every neighbour.
Title
Number
Author
Date
Version
Page
12 of 17
SIEMENS
As displayed on the report, the problem was indicated 9 times for ADJCN8 only. The user can
easily notice that, during the observation period, there was only one HO attempt outgoing to
neighbour 8. The value 9 in DEL_NB column means that the rule was fulfilled for nine days within
observation period for this neighbour, which gives true value for all days selected for evaluation.
Object
selection
panel
Time
selection
panel
To see the detailed information about number of handovers to this neighbour, their success rate or
what portion of all handovers from the source cell to the evaluated cell is attempted, the user can
expand the grid for particular adjacent cell.
Title
Number
Author
Date
Version
Page
13 of 17
SIEMENS
As the additional information, the user gets the Cell Global Identifier of the neighbour cell. In this
example, it is clearly stated that neighbour relation between BTS 1.022 (CELLGLID = 11142) and
BTS 1.015 (CELLGLID = 10652) should be considered for deletion. Now, the user must check if
the DEL_NB rule gives the same results in opposite direction. Only in this case, the relation can be
deleted.
Title
Number
Author
Date
Version
Page
14 of 17
SIEMENS
For BTS 1.015, there are two neighbours suspected to be unnecessary: ADJCN 5 and ADJCN 8.
Since ADJCN 8 is the BTS 1.022 (CELLGLID = 11142), this is the suspected neighbour, which the
user seeks.
Title
Number
Author
Date
Version
Page
15 of 17
SIEMENS
It occurs that in both handover directions (10652 11142 and 11142 10652) the result of the
rule DEL_NB gave the positive value so the neighbour relation can be removed from the system.
Except the result of the DEL_NB rule, the user gets some performance statistics as additional
information. These are:
number of handover attempts per BTS,
number of handover attempts per neighbour,
ratio of handover attempts per neighbour related to all handover attempts,
number of successful handovers per neighbour,
handover success rate per neighbour cell,
number of registered neighbours per cell.
If the user now wants to see the neighbour relations of a special cell of interest, he can click the
cell and the dedicated neighbour relations will be displayed as a line.
Title
Number
Author
Date
Version
Page
16 of 17
SIEMENS
The next step of finding the unnecessary neighbour relation is to check the same BTSs that were
pointed out to be suspected of suffering such problem:
Neighbour relations of cell 11142
As the example shows, the cell 11142 can be deleted from the neighbour list of the cell 10652 as
well as cell 10652 can be deleted from the neighbour list of cell 11142.
Title
Number
Author
Date
Version
Page
17 of 17