Maximizing Network Performance

Ultima™ Forté™
Training Session
Appendix

Ultima Forté
Guidelines
for
Optimization Area Selection,
Measurement Plan & Neighbor List Optimization

1 Introduction 116
2 Optimization Project Workflow 117
3 Area Selection for Optimization 119
3.1 Selected Area – Definitions 119
3.1.1 Optimization Area 119
3.1.2 Affected Area 120
3.1.3 Rest of the Network 120
3.2 Optimization and Affected Area Selections 121
3.2.1 Prerequisites 121
3.2.2 Selection Guidelines 124
4 Measurement Plan 127
4.1 Recording Duration 127
4.2 Recording Goals 128

Slide 2

Confidential & Proprietary

Appendix

1. Introduction
The goal of the document is to present the methodology to be used for selection of the guard zone
for handling optimization activity on the network with the Ultima™ Forté product.
The objective is to implement a complete network optimization cycle including:
 Preliminary Analysis
 Recording Set Up
 Neighbor List Optimization
 Frequency Planning
This document does not aim to replace the existing documentation but to offer a general approach
to the optimization process.

Slide 3

Confidential & Proprietary

Appendix
2. Optimization Project Workflow
The following block diagram outlines a typical optimization workflow with Ultima™ Forté:

Pre-Project KPI Collection

Neighbor List Deletion/BSIC Check

Initial GSM 900 Recording Initial DCS 1800 Recording

Initial GSM 900 Retune Initial DCS 1800 Retune (Optional)

Final GSM 900 Recording Final DCS 1800 Recording

Neighbor List Optimization

Final Retune (Both Bands)

Figure 1: Typical GSM Optimization Workflow Slide 4

Confidential & Proprietary

Appendix
1. Pre-Project KPI Collection – a week or two before the project begins, network KPI data will be collected.
The KPI to be collected is described in Schema’s Data Requirements document for the relevant vendor.
2. Neighbor List Deletion/BSIC Check – prior to the collection of mobile measurements data two necessary
procedures should apply: neighbor list deletion (to allow free places in the BAL for recording non-neighbor BCCH
frequencies in the network) and BSIC check (the BCCH-BSIC reuse in the network should be as dispersed as
possible to allow a clear distinction between the recorded sectors). Neighbor list deletion can be calculated by
Ultima™ Forté™ based on handoff statistics only i.e. without a model. Use Ultima Forté’s ‘BCCH-BSIC Reuse’
report, if the BSIC plan is not optimal. A BSIC optimization should be applied to the network before recording
starts.
3. Initial Recordings – mobile measurements recordings of the network. If the network does not suffer from
shadowing[1] this cycle of recordings may be the Final Recordings stage. This phase should be run on each band
in a dual-band network and can run simultaneously on both bands.
4. Initial Retune – this phase should be performed if there is a shadowing effect in the network. It should
consist of BCCH frequencies only (no retune for TCH frequencies required at this stage) and is intended to reveal
shadowing for the next step of recordings (see below).
5. Final Recordings – after a BCCH retune is executed a second phase of mobile measurements recordings
should take place. This phase is intended to provide an additional set of recordings which will be used for the
creation of a shadow-free model, when combined with the first stage of recordings.
6. Neighbor List Optimization – this phase is aimed at reducing BCCH reuse constraints due to unnecessary
adjacencies. Neighbor List Optimization will be calculated based on the complete model which was generated
after the Final Recordings phase.
7. Final Retune – this is the final phase of the Frequency Optimization which will be executed on BCCH and/or
TCH layers, according to operator requirements.

Shadowing effect is likely to exist in networks that use 20 BCCH frequencies or less in highly dense environments and in challenging geographical
[1]

environments. Slide 5

Confidential & Proprietary

Appendix
3. Area Selection for Optimization

3.1 Selected Area - Definitions

While optimizing a network it is necessary to define the following three entities:
1. Optimization Area
2. Affected Area
3. Rest of the Network

3.1.1 Optimization Area

The optimization area is a cluster of sectors whose parameters may be modified as part of the optimization
process. See figure 2 below.

Figure 2: Optimization Area (Dark Blue Sectors) Slide 6

Confidential & Proprietary

Appendix
3.1.2 Affected Area
The affected area contains sectors that might interfere with or are subject to interference by the
optimization area. Figure 3 below illustrates the distinction between sectors in the optimization area (light
blue) and those in the affected area (dark blue).

Optimization Area

Figure 3: Affected Area (Dark Blue Sectors)

3.1.3 Rest of the Network

The rest of the network is composed of the sectors that will
not affect or be affected by the optimization area during
the optimization process. No restrictions apply to the rest
of the network during the optimization process.

Slide 7

Confidential & Proprietary

Appendix
3.2 Optimization and Affected Area Selections
The following paragraph defines guidelines and provides recommendations for optimization and affected
area selections in order to achieve optimal network performance results within the shortest timeframe.

3.2.1 Prerequisites
A small optimization area is less sensitive to modeling mistakes. Since the sectors in the affected area are
not participating in the optimization process and their frequencies are not being changed, modeling
mistakes that might occur in such an area can be tolerated. However, such mistakes in the model cannot
be tolerated in the optimization area. An example of a modeling mistake can be the wrong association of
interference with sectors.
The cluster selected for optimization should be frozen during the optimization process. The following
changes should be avoided:
- BSC split (new BSC activation)
- Site re-homing (BSC/site relations change)
- Changes of LAC and CI
- Site additions
The implementation process should be executed during low-traffic activity of the network (weekends,
nights) in order to ensure regular network performance during busy hours.
- Mechanical changes such as tilt, azimuth, height and antenna type
- Frequency changes (BCCH and BSIC)
- Neighbor addition or deletion
- Transmission power
- UL/OL definitions
- UL/OL thresholds Slide 8

Confidential & Proprietary

Appendix
The following changes should be performed only during the maintenance window (not while recording is taking
place):
- TRX additions
- Frequency changes (BCCH and BSIC). In this case – a new Measurements Plan might have to be regenerated
(with some vendors)
 The data collection process should not affect daily network performance.
 The data between the different data sources i.e. coordinates file, traffic file, network configuration file,
mobile measurements data etc. should be consistent and relate to the same timeframe.
The size of the optimization area selection depends on the data collection process, which may vary between
different vendors.
The following table outlines Schema’s recommendation for the relationships between the vendor, the size of
the optimization area and the data collection period:

Vendor Typical Data Recommended Optimization Maximum Recommended

Collection Period Area Size Optimization Area Size
(Days) (#Sectors) (#Sectors)
Ericsson 2-5 200 – 1000 10000
Nokia 2-7 200 – 1000 10000
Alcatel 3-10 200 – 1000 5000
Siemens 3-10 200 – 1000 2500
Huawei 3-10 200 – 1000 1000
Nortel 5-15 100 – 200 500
Motorola 5-15 100 – 200 300
Slide 9
Table 1: Schema Optimization Parameter Recommendations*
Confidential & Proprietary
Appendix

3.2.2 Selection Guidelines

As a general rule the affected area zone should consist of the sectors that belong to the BSCs surrounding
the optimization area. This selection should be used for mobile measurements collection. The following
images describe the optimization area (figure 4), the recommended affected area (figure 5) and the
relationship between them (figure 6).

Optimization Area

Figure 4: Optimization Area

Slide 10

Confidential & Proprietary

Appendix

Optimization and
Affected Areas

Optimization Area

Affected Area

Figure 5: Optimization and Affected Areas (1)

Figure 6: Optimization and Affected Areas (2)

Slide 11

Confidential & Proprietary

Appendix

Ultima™ Forté™ also calculates the affected area, based on the mobile measurements model. Sectors in
the affected area have outgoing/incoming interferences with sectors in the optimization area. The user’s
affected area selection should be equal to or greater than Ultima Forté’s calculation of the affected area.
If Ultima Forté’s affected area is significantly larger than the user’s selection it is recommended to
reconsider enhancement of the latter and record additional mobile measurements for those sectors.
In order to effectively implement a measurements plan, the following rules should be used as guidelines:
 When the optimization area fits the recommended size the measurements plan can be applied to
both the optimization area and the affected area.
 When the optimization area exceeds the recommended size, the measurements plan for Alcatel,
Siemens, Huawei, Nortel and Motorola infrastructures should not include sectors that belong to the
affected area.
 Neighbor list optimization requires cross-band (900-1800) measurements in dual band networks.
 When frequency planning is the only optimization objective (no neighbor list optimization is
required) a measurements plan should be calculated without the ‘cross-band’ option (each band on
its own).

Slide 12

Confidential & Proprietary

Appendix
4. Measurements Plan
The measurements plan functionality has been developed for mobile measurements data collection. It should
be properly configured in order to provide correct data collection within the shortest time frame. Different
optimization tasks may require different configurations of the measurements plan.
The following processes can distinguish the measurements plan functionality configuration:
 Single band frequency plan
 Dual band frequency plan
 Neighbor list optimization for single band network
 Neighbor list optimization for dual band network

4.1 Recording Duration

The recording duration depends on the number of BCCH frequencies to be measured. This number depends
on the optimization task.
For cells using the Common BCCH feature (CBCCH), it is necessary to measure both bands.
All DCS (1800/1900) cells will measure GSM (850/900) BCCH frequencies used by the common BCCH cell, as
well as their own band, resulting in additional recording days.
The recommended duration of the mobile measurements time frame should not exceed 10 days. If mobile
measurements cannot be completed within this time frame, the optimization targets should be re-evaluated.
In addition, some changes in network configuration may be required in order to fit into the recommended
time frame. Such changes, which may result in a shorter recording time frame, include but are not limited to
the trimming of neighbor lists especially for sectors with very long neighbor lists (more than 25).
If the recording time frame still cannot fit into 10 days for all sectors in the optimization area, Ultima Forté is
able to automatically provide an efficient measurements plan that will fit into the desired recording time
frame. This feature will eliminate less important frequencies for some of the sectors from the measurement
process. Slide 13

Confidential & Proprietary

 Maximizing Network Performance

Thank you
www.schema.com

Schema retains the sole proprietary rights to all information contained in this document. No part of this publication may be reproduced, stored in a retrieval
system, transmitted in any form or by any means, including but not limited to: electronic, mechanical, magnetic, photocopy, recording, or otherwise, in use now
or in the future, without prior written consent from Schema

Copyright © 2006, Schema