Sunteți pe pagina 1din 52

In Confidence

O2 'Do Once' Radio Design


Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 1 of 52

Compressed Mode Scenarios and Strategy


Document Type: Report

Summary:
Inter-system and inter-frequency handovers typically require a set of compressed mode neighbour cell
measurements. Handover may be triggered for reasons of coverage, load or traffic. This report focuses upon
coverage reason handover. Handover scenarios have been categorised using four dimensions – rate of
change of propagation conditions, service, triggering mechanism and measurement purpose. These
dimensions have been used as a basis for balancing trade-offs and generating recommended compressed
mode configurations. The 3GPP compressed mode parameter set has been identified and compared with the
Nokia compressed mode parameter set. The O2 preferred compressed mode configuration has been
identified based upon a set of primarily theoretical arguments. Field experience has been presented and used
to a limited extent. The impact of using the Nokia default configuration rather than the O2 preferred
configuration has been evaluated. Recommendations have been made in terms of adjusting configurable
RNC parameters and requesting changes to non-configurable RNC parameters.

Author(s): Chris Johnson


Author(s) Operating Nokia UK
Business:
Date: 22nd April 2004
Version: Issue 1
Reference: O2RD/04/021
Approved by:
Name: Oscar Clop
Role / Authority:
Signature:

Date:

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 2 of 52

DISTRIBUTION LIST

O2 Nokia
Ian Miller Aleksi Toikkanen
Brendan O'Reilly Mike Lawrence
Tony Conlan George Grayland
Nick Outram Chris Johnson
Oscar Clop Colm Jones
Jude Saldanha Filippo Belloni
Lara Pazienza

CHANGE HISTORY

Tool used Microsoft Word 2000


File location
Create, save & print dates 08/02/2002 00:00 22/04/2004 18:04:00 27/04/2011 09:59:00
Version Date Changed by Changes
Draft A 08/02/04 Chris Johnson Document created
Issue 1 22/04/04 Chris Johnson Raised to Issue 1 status

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 3 of 52

EXECUTIVE SUMMARY:

Compressed mode forms an essential part of inter-system and inter-frequency handover for the majority of
dual mode UE. It may be avoided in some cases if UE are equipped with a dual receiver. Inter-system
handover will be more important than inter-frequency handover until O2 starts to deploy dual carrier Node
B. Inter-system handover will be required at coverage boundaries between the WCDMA and GSM systems.
These boundaries may or may not be planned. It is desirable to maximise the coverage of the WCDMA
system and minimise the quantity of inter-system handovers. The extent of WCDMA connected mode
coverage is defined by the set of thresholds which are used to trigger inter-system handover. The extent of
WCDMA idle mode coverage is defined by the equivalent set of thresholds for cell re-selection.

Compressed mode scenarios can be categorised using four dimensions. These dimensions are the
measurement purpose (GSM RSSI, BSIC identification, FDD measurement); the service being used (AMR
speech, RT data, NRT data); the triggering mechanism (uplink transmit power, downlink transmit power,
CPICH Ec/Io, CPICH RSCP, uplink quality); and the rate of change of propagation conditions (high, low). A
compressed mode strategy may be developed for each combination of the variables associated with these
four dimensions.

Compressed mode functionality and signalling has been standardised by 3GPP. The standardisation of
compressed mode allows both inter-operability and flexibility. The flexibility means that the implementation
of compressed mode involves making a set of design decisions. Nokia has implemented compressed mode
according to the 3GPP specifications using a combination of configurable and non-configurable RNC
parameters. These parameters provide some but not complete flexibility. Some aspects of compressed mode
have been implemented without parameterisation.

This report has identified the trade-offs associated with each compressed mode parameter. The Nokia default
configuration and the O2 preferred configuration have been identified. The impact of applying the Nokia
default configuration rather than the O2 preferred configuration has been evaluated. This evaluation has been
based primarily upon theoretical expectations and arguments. A relatively limited quantity of field trial
experience has been made available for this report. The quantity of experience will increase throughout 2004
as O2’s WCDMA network performance is evaluated and refined.

A number of configurable parameter changes have been proposed within the recommendations section of
this report. These parameter changes can be made by O2 but should be tested during a localised parameter
trial prior to deploying across a large area. The majority of arguments for these proposed parameter changes
are theoretical rather than being based upon field trial experience.

A number of non-configurable parameter changes have been proposed within the recommendations section
of this report. These parameter changes cannot be made by O2 and require requests to be made to Nokia
product line for future implementation. Requests for these parameter changes should be prioritised with other
outstanding requests to Nokia product line. The majority of arguments for these proposed parameter changes
are theoretical rather than being based upon field trial experience.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 4 of 52

TABLE OF CONTENTS

1. Scope and Objectives ............................................................................................................ 6

2. Introduction .......................................................................................................................... 7

3. Scenarios ............................................................................................................................... 9
3.1. Coverage Reason ................................................................................................................................. 9
3.2. Load Reason ...................................................................................................................................... 11
3.3. Traffic Reason.................................................................................................................................... 12
4. Compressed Mode Methods and Patterns ......................................................................... 13
4.1. Method ............................................................................................................................................... 14
4.2. Timing................................................................................................................................................ 18
4.3. Direction ............................................................................................................................................ 21
4.4. Power Control .................................................................................................................................... 21
4.5. Scrambling Code................................................................................................................................ 23
4.6. Downlink Frame Type ....................................................................................................................... 24
4.7. Measurement Reporting ..................................................................................................................... 24
4.8. Other Configuration Parameters ........................................................................................................ 25
5. Preferred Compressed Mode Methods and Patterns ........................................................ 26
5.1. Method ............................................................................................................................................... 26
5.2. Timing................................................................................................................................................ 29
5.3. Direction ............................................................................................................................................ 35
5.4. Power Control .................................................................................................................................... 35
5.5. Scrambling Code................................................................................................................................ 37
5.6. Downlink Frame Type ....................................................................................................................... 37
5.7. Measurement Reporting ..................................................................................................................... 38
5.8. Other Configuration Parameters ........................................................................................................ 38
6. Impact of using the Nokia Default Parameter Set............................................................. 40
6.1. Method ............................................................................................................................................... 40
6.2. Timing................................................................................................................................................ 41
6.3. Direction ............................................................................................................................................ 42
6.4. Power Control .................................................................................................................................... 42
6.5. Scrambling Code................................................................................................................................ 42
6.6. Downlink Frame Type ....................................................................................................................... 43
6.7. Measurement Reporting ..................................................................................................................... 43
6.8. Other Configuration Parameters ........................................................................................................ 43
7. Field Experience ................................................................................................................. 44
7.1. Spreading Factor Division by 2 ......................................................................................................... 44
7.2. Higher Layer Scheduling ................................................................................................................... 45
7.3. Puncturing .......................................................................................................................................... 45
8. Recommendations .............................................................................................................. 46

9. Conclusions ......................................................................................................................... 49

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 5 of 52

10. References ........................................................................................................................... 50

11. Abbreviations ..................................................................................................................... 51

Appendix A. Example UE capability message ...................................................................... 52

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 6 of 52

1. Scope and Objectives


The scope of this report is based upon the timescales of Nokia RNC software release RAN1.5.2.ED2 and the
contents of the 3GPP R99 specifications. These timescales limit the use of compressed mode to coverage
reason handovers. Load and traffic reason handover scenarios are referenced although they are not addressed
in detail. Compressed mode is treated as one part of the inter-frequency and inter-system handover
procedures. Thresholds defining the initial triggering of these handover procedures are outside the scope of
this report although the triggering mechanisms themselves are accounted for when defining the compressed
mode strategies. The handover and associated service break subsequent to the use of compressed mode are
also outside the scope of this report. Although the scope includes both inter-frequency and inter-system
handovers, the focus has been directed towards inter-system handovers. This is aligned with the expectation
that a second 3G carrier is unlikely to be deployed within the timescales of RAN1.5.2.ED2. The report is
primarily theoretical although one section is included to present indicative compressed mode performance
observed during Nokia field trials.

The objectives of this report are to provide:

• a description of scenarios where the use of compressed mode will be required


• a strategy defining how compressed mode should be customised for each scenario
• a specification of the compressed mode parameter sets used by 3GPP and Nokia
• recommendations for how the 3GPP compressed mode parameter set can be customised for each
measurement purpose and for each service type
• An evaluation of the impact of using Nokia’s compressed mode implementation with Nokia’s
recommended compressed mode parameter set
• Recommendations for refining Nokia’s compressed mode parameter set and if appropriate, for
requesting changes to Nokia’s compressed mode implementation

The report has been generated with the intention of making it applicable to O2 business operations in the
UK, Ireland and Germany.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 7 of 52

2. Introduction
Compressed mode forms an essential part of inter-system and inter-frequency handover for the majority of
dual mode UE. It may be avoided in some cases if UE are equipped with a dual receiver. Inter-system
handover will be more important than inter-frequency handover until O2 starts to deploy dual carrier Node
B. Inter-system handover will be required at coverage boundaries between the WCDMA and GSM systems.
These boundaries may or may not be planned. It is desirable to maximise the coverage of the WCDMA
system and minimise the quantity of inter-system handovers. The extent of WCDMA connected mode
coverage is defined by the set of thresholds which are used to trigger inter-system handover. The extent of
WCDMA idle mode coverage is defined by the equivalent set of thresholds for cell re-selection.

There is a delay associated with completing the compressed mode measurements procedure. This means that
the thresholds used to trigger inter-system handover must incorporate a back-off that allows sufficient time
for the compressed mode procedure and for the handover to occur prior to the WCDMA radio links failing.
The back-off should be minimised to avoid a significant reduction in WCDMA coverage. Large back-offs
will also result in more frequent inter-system handovers. It is possible to define a strategy that tunes the
back-off factor on a per scenario basis. For example, low mobility scenarios may have a smaller back-off as
a result of UE having more time to complete the compressed mode measurements. Alternatively, the same
back-off could be applied to all scenarios and greater time could be allowed for compressed mode
measurements in low mobility scenarios. This strategy would have the effect of making compressed mode
measurements more reliable in low mobility scenarios.

The delay associated with compressed mode measurements depends to a large extent upon the service and
whether the associated handover is inter-system or inter-frequency. Inter-system handover of real time
services have a relatively large delay associated with them due to the requirement to complete GSM RSSI
measurements and GSM BSIC identification. BSIC identification is not required for the inter-system
handover of non-real time services. Figure 1 illustrates these measurement purposes.

Inter-System Inter-Frequency
Handover Handover
Real Time • GSM RSSI Measurement • FDD Measurement
Services • BSIC Identification

Non-Real Time • GSM RSSI Measurement • FDD Measurement


Services

Figure 1 – Compressed mode measurement purposes as a function of service and handover type
The compressed mode delay also depends upon the configuration of compressed mode in terms of its
transmission gaps and transmission gap patterns. Smaller transmission gaps and longer transmission gap
patterns tend to reduce the impact of compressed mode but also increase the compressed mode delay. The
current strategy is to complete compressed mode with a short delay such that handover can occur rapidly and
the back-offs associated with the inter-system handover triggering mechanisms can be reduced. Compressed
mode should not however be completed so rapidly that its reliability becomes poor.

Transmission gaps may be generated using a set of standardised compressed mode methods. These methods
include puncturing, spreading factor reduction, and higher layer scheduling. Puncturing is only applicable to
the downlink. The R99 3GPP specifications limit the use of puncturing and higher layer scheduling
according to whether the transport channel starting position is fixed or flexible. For example, higher layer
scheduling is only applicable to transport channels using flexible starting positions. The AMR speech service
uses fixed starting positions in the downlink direction to allow UE to complete blind AMR bit rate detection.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 8 of 52

This means that higher layer scheduling can not be applied to the AMR speech service in the downlink. It
can however be applied to the AMR speech service in the uplink.

Each compressed mode method has a different impact upon system performance. Puncturing is applied on a
per TTI basis and requires an increased SIR target to achieve the Eb/No. Increasing the SIR target has an
impact upon both the coverage and capacity of the system. Spreading factor reduction is applied on a per
frame basis and also requires an increased SIR target. Higher layer scheduling is applied on a per TTI basis
and results in a reduced service throughput. In this case the MAC layer is responsible for selecting a lower
transport format combination. All compressed mode methods have a similar impact upon the performance of
inner loop power control, i.e. transmit power control commands are stopped during the transmission gaps
and the average transmit power tends to increase. There is a requirement to select and customise each
compressed mode method on a per scenario basis.

Compressed mode functionality and signalling has been standardised by 3GPP. The relevant specifications
are listed as references [1], [2], [3], [4], [5], [6], [7], [8]. The standardisation of compressed mode allows
both inter-operability and flexibility. The flexibility means that the implementation of compressed mode
involves making a set of design decisions. Nokia has implemented compressed mode according to the 3GPP
specifications using a combination of configurable and non-configurable RNC parameters. These parameters
provide some but not complete flexibility. Some aspects of compressed mode have been implemented
without parameterisation.

The contents of this report are equally applicable to Omni-Transmit Sector Receive (OTSR) and Sector
Transmit Sector Receive (STSR) Node B. Separate OTSR and STSR compressed mode strategies are not
required. It is more likely that separate OTSR and STSR strategies are required for inter-system and inter-
frequency handovers. For example, the CPICH RSCP inter-system handover triggering threshold is likely to
be different for OTSR and STSR Node B. The term OTSR is equivalent to Nokia’s term Roll-Out
Configuration (ROC) whereas the term STSR is equivalent to Nokia’s term Capacity Enhanced
Configuration (CEC).

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 9 of 52

3. Scenarios
Dual mode UE may be required to handover from the WCDMA system to the GSM system as a result of
moving into an area of relatively poor WCDMA coverage. Handover may also be required if the WCDMA
system becomes heavily loaded or if the UE requests a service which is provided by the GSM system. These
reasons for inter-system handover provide the basis for defining a set of compressed mode scenarios:

 coverage reason compressed mode  traffic reason compressed mode


 load reason compressed mode

These inter-system handover reasons may also be applied to inter-frequency handovers. Inter-system
handovers require compressed mode transmission gaps for RSSI measurements and BSIC identification.
Inter-frequency handovers require compressed mode transmission gaps for FDD measurements.

3.1. Coverage Reason

Coverage reason compressed mode may be used for either inter-frequency or inter-system handover. Inter-
frequency handover will be required at a coverage boundary between a group of Node B which have two
carriers assigned and a group of Node B which have only a single carrier assigned. UE using the second
carrier belonging to a dual carrier Node B will be required to handover onto the first carrier as they move
toward a single carrier Node B. Inter-frequency handover is less likely to be required for coverage reasons
when a UE is located within an area of contiguous dual carrier Node B coverage. The coverage footprint of
the two carriers will be approximately equal and there will be little or no coverage gain when moving from
one carrier to the other. Inter-system handovers will be required at coverage boundaries between the
WCDMA and GSM systems. Coverage boundaries may be either planned or a result of a gap in planned
coverage. Coverage reason compressed mode scenarios in terms of their location are illustrated in Figure 2.

Coverage Based

Edge of Planned
WCDMA Coverage

• Coverage boundary at the edge of an urban/suburban area


• Coverage boundary within an urban/suburban area
• Coverage boundary on an open stretch of road
• Coverage boundary on an open stretch of rail
• Coverage boundary when leaving a railway station
• Coverage boundary when leaving an airport
• Coverage boundary when leaving a motorway service station
• Coverage boundary when leaving a special events venue
• Coverage boundary when entering a building

Gap within Planned


WCDMA Coverage

• Coverage hole within an urban/suburban area


• Coverage hole within a building
• Coverage hole on an open stretch of road
• Coverage hole on an open stretch of rail

Figure 2 – Coverage reason compressed mode scenarios (location)

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 10 of 52

It is equally important for compressed mode and inter-system handover to be successful in all of these
location based scenarios. In all cases there is a requirement to maximise the reliability of the compressed
mode and inter-system handover procedures. Some scenarios also have a requirement to complete the
compressed mode and inter-system handover procedures more rapidly than other scenarios. This requirement
is generated by the rate of change of propagation conditions associated with each scenario. If coverage is
changing relatively slowly for a scenario then more time can be spent completing inter-frequency and inter-
system measurements to help ensure that the subsequent handover is reliable.

Figure 2 can be expanded by considering that dual mode UE may encounter any of these scenarios while
using any of the supported services. Figure 3 expands Figure 2 to include a service dependence.

Coverage Based

Edge of Planned Service 3


WCDMA Coverage
Service 2
Service 1
• Coverage boundary at the edge of an urban/suburban area
• Coverage boundary within an urban/suburban area
• Coverage boundary on an open stretch of road
• Coverage boundary on an open stretch of rail
• Coverage boundary when leaving a railway station
• Coverage boundary when leaving an airport
• Coverage boundary when leaving a motorway service station
• Coverage boundary when leaving a special events venue
• Coverage boundary when entering a building

Gap within Planned Service 3


WCDMA Coverage
Service 2
Service 1
• Coverage hole within an urban/suburban area
• Coverage hole within a building
• Coverage hole on an open stretch of road
• Coverage hole on an open stretch of rail

Figure 3 – Coverage reason inter-system handover scenarios (location and service)


The set of services supported by the WCDMA system can be categorised according to how sensitive they are
to the effects of compressed mode. The selection of the various compressed mode methods and parameter
sets should account for the service requirements. Real time data services should be configured such that
compressed mode does not cause a throughput reduction. The AMR speech service is able to incur a
reduction in throughput if a lower AMR bit rate is available. Table 1 presents a set of example services
which shall be considered when defining compressed mode strategies on a per service basis.

Application CN Domain Traffic Class


AMR Speech CS Conversational
Video Call CS Conversational
Video Streaming CS Streaming
Audio Streaming CS Streaming
Internet Browsing PS Interactive
WAP Browsing PS Interactive
Interactive Gaming PS Interactive
File Transfer PS Background
Email PS Background

Table 1 – Example set of WCDMA services

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 11 of 52

Conversational and streaming services are treated as real time service and assumed to be throughput sensitive
(with the exception of the speech AMR service which has the ability to operate at a reduced throughput).
Interactive and background services are treated as non-real time services and assumed to be less throughput
sensitive.

Figure 3 can be expanded by considering that a range of coverage related mechanisms may be used to trigger
compressed mode measurements and the subsequent handover. Figure 4 expands Figure 3 to include a set of
coverage related triggering mechanisms.

Coverage Based

Edge of Planned Service 3


WCDMA Coverage
Service 2
Triggering
Service 1 Mechanisms
• Coverage boundary at the edge of an urban/suburban area
• Coverage boundary within an urban/suburban area
• Coverage boundary on an open stretch of road
• Coverage boundary on an open stretch of rail

• High downlink transmit power


• Coverage boundary when leaving a railway station

• High uplink transmit power


• Coverage boundary when leaving an airport
• Coverage boundary when leaving a motorway service station

• Low CPICH RSCP

• Poor uplink quality


• Low CPICH Ec/Io
• Coverage boundary when leaving a special events venue
• Coverage boundary when entering a building

Gap within Planned Service 3


WCDMA Coverage
Service 2
Service 1
• Coverage hole within an urban/suburban area
• Coverage hole within a building
• Coverage hole on an open stretch of road
• Coverage hole on an open stretch of rail

Figure 4 – Coverage reason inter-system handover scenarios (location, service and triggering mechanism)
Five triggering mechanisms have been specified. These are aligned with the capability of Nokia’s
RAN1.5.2.ED2 RNC software. They are not necessarily aligned with the capability of specific UE. For
example, a specific model of UE may not support measurement reports triggered by uplink transmit power. It
is important to consider these triggering mechanisms when defining a compressed mode strategy because
they provide an indication of whether or not a particular compressed mode configuration is likely to be
feasible. One technique used to generate compressed mode transmission gaps is to reduce the spreading
factor. This increases the air-interface bit rate and allows the same amount of data to be transmitted in less
time. It also requires a transient increase in transmit power. If compressed mode has been triggered by either
the uplink or downlink transmit power mechanism then it may not be possible to increase the transmit power
to support a reduction in spreading factor.

3.2. Load Reason

Load reason compressed mode may be used for either inter-frequency or inter-system handover. Inter-
frequency handover may be applicable if a Node B is configured with multiple carriers and there is an
unequal distribution of load between those carriers. Directed RRC establishment can be used to help achieve
a relatively even load between carriers. Unequal loads are likely to occur at coverage boundaries between

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 12 of 52

dual carrier Node B and single carrier Node B. All UE which use soft handover to move from a single carrier
Node B to a dual carrier Node B will remain on the first carrier unless instructed to complete a inter-
frequency handover. Load reason compressed mode for inter-system handover is applicable to any Node B
which is incurring high quantities of traffic.

Nokia’s RAN1.5.2.ED2 and RAN04 RNC software releases do not support load reason inter-frequency nor
inter-system handover. The required functionality is planned for the RAN05 RNC software release. These
timescales mean that this report places little focus upon load reason inter-frequency and inter-system
handover.

3.3. Traffic Reason

Traffic reason compressed mode may be used for either inter-frequency or inter-system handover. Inter-
frequency handover may be applicable if a strategy has been defined to link specific services with specific
carriers. When a UE initiates a capacity request for a specific service then it may be desirable to move the
UE to a different carrier. Inter-system handover may be applicable if a strategy has been defined to link
specific services with specific systems. Similar to the inter-frequency scenario, when a UE initiates a
capacity request for a specific service then it may be desirable to move the UE to the GSM system.

Nokia’s RAN1.5.2.ED2 and RAN04 RNC software releases do not support traffic reason inter-frequency nor
inter-system handover. The required functionality is planned for the RAN05 RNC software release. These
timescales mean that this report places little focus upon traffic reason inter-frequency and inter-system
handover.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 13 of 52

4. Compressed Mode Methods and Patterns


Compressed mode methods and the associated compressed mode patterns are controlled by the RNC. The
Node B is informed of the configuration using the dedicated NBAP: RADIO LINK RECONFIGURATION
PREPARE and the dedicated NBAP: COMPRESSED MODE COMMAND messages. The UE is informed
of the configuration using either the RRC: PHYSICAL CHANNEL RECONFIGURATION and RRC
MEASUREMENT CONTROL messages, or the RRC: TRANSPORT CHANNEL RECONFIGURATION
and RRC: MEASUREMENT CONTROL messages. Selection between the RRC: PHYSICAL CHANNEL
RECONFIGURATION message and the RRC: TRANSPORT CHANNEL RECONFIGURATION message
depends upon the specific parameters affected by compressed mode. Figure 5 illustrates an example of the
NBAP and RRC signalling for an inter-system handover of a CS service.

UE Node B RNC

RRC: Measurement Control


ISHO
RRC: Measurement Report triggering
NBAP: Radio Link Reconfiguration Prepare
NBAP: Radio Link Reconfiguration Ready
Initial
NBAP: Radio Link Reconfiguration Commit Compressed
RRC: Physical Channel Reconfiguration Mode
Configuration
RRC: Physical Channel Reconfiguration Complete
NBAP: Compressed Mode Command
RRC: Measurement Control
GSM RSSI
RRC: Measurement Report Measurement
NBAP: Compressed Mode Command
RRC: Measurement Control GSM BSIC
Identification
RRC: Measurement Report
RRC: Handover from UTRAN Command

Figure 5 – Example NBAP and RRC signalling used to configure and activate compressed mode
The signalling flow for a PS service inter-system handover makes use of the RRC: CELL CHANGE
ORDER FROM UTRAN message rather than the RRC: HANDOVER FROM UTRAN COMMAND
message. In addition, a PS service inter-system handover does not require GSM BSIC identification. The
signalling flow for an inter-frequency handover makes use of FDD carrier measurement rather than GSM
RSSI measurement and GSM BSIC identification.

The remainder of this section provides a description of 3GPP specified and Nokia implemented compressed
mode methods and patterns. This section is intended to be a statement of implementation rather than an
evaluation of the associated trade-offs. Section 5 evaluates the trade-offs associated with specific compressed
mode configurations.

Nokia’s RAN1.5.2.ED2 implementation makes use of two separate compressed mode strategies – an
advanced strategy and a normal strategy. The normal strategy includes fewer combinations of compressed
mode configurations. The non-configurable RNC databuild parameter CompressedModeStrategy is used to
define which strategy is used. This parameter is currently configured such that the RNC uses the normal
compressed mode strategy. This section describes both the normal and advanced strategies. Inclusion of the
advanced strategy is for information only.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 14 of 52

4.1. Method

The RNC is able to configure the method used to generate the compressed mode transmission gaps. This is
done independently for the uplink and downlink.

4.1.1. Uplink

3GPP TS25.331 and TS25.433 specify that uplink compressed mode transmission gaps may be achieved
using either spreading factor division by 2 (SF/2) or higher layer scheduling (HLS). These methods are
presented in Table 2.
3GPP Name Description 3GPP
Range
Uplink Compressed Mode Defines the uplink compressed mode method SF/2, HLS
Method

Table 2 – 3GPP uplink compressed mode method parameter


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message and to the UE within the RRC: PHYSICAL/TRANSPORT
CHANNEL RECONFIGURATION message.

Nokia’s normal compressed mode strategy configures the compressed mode method based upon service
type. No account is taken of the mechanism that has triggered compressed mode. Table 3 presents the
compressed mode methods used by the normal strategy.
AMR Speech RT Data NRT Data
Uplink Compressed Mode SF/2 SF/2 HLS
Method

Table 3 – Nokia uplink compressed mode methods (normal strategy)


Nokia’s advanced compressed mode strategy configures the compressed mode method based upon service
type and triggering mechanism. Nokia’s RAN1.5.2.ED2 supports five coverage reason compressed mode
triggering mechanisms – high uplink transmit power, high downlink transmit power, low CPICH Ec/Io, low
CPICH RSCP and low uplink quality.

In the case of the AMR speech service, the SF/2 method is used independent of the compressed mode
triggering mechanism. Selection of the SF/2 method is presented in Table 4.

Triggered by 1st Choice for CM Method 2nd choice for CM Method 3rd Choice for CM Method
UL Tx Power SF/2 and adding of lower TF - -
DL Tx Power SF/2 and adding of lower TF - -
CPICH Ec/Io SF/2 and adding of lower TF - -
CPICH RSCP SF/2 and adding of lower TF - -
Uplink Quality SF/2 and adding of lower TF - -

Table 4 – Nokia AMR speech service uplink compressed mode methods (advanced strategy)
The addition of a lower AMR speech transport format corresponds to a lower AMR bit rate. This helps the
UE maintain its connection in case it starts to approach its maximum transmit power capability. The lower
AMR speech transport format is limited by the RNC databuild parameter LowerULAMRmodeCM. This
parameter is presented in Table 5.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 15 of 52

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable LowerULAMRmodeCM RNC No 1 (12.2 kbps), 2 (10.2 kbps), 1
3 (7.95 kbps), 4 (7.4 kbps),
5 (6.7 kbps), 6 (5.9 kbps),
7 (5.15 kbps), 8 (4.75 kbps)

Table 5 – AMR speech service uplink compressed mode lower transport format bit rate
The LowerULAMRmodeCM parameter is currently configured with a value of 1, i.e. an AMR bit rate of 12.2
kbps. This means that a lower bit rate transport format is not added.

In the case of CS data services, the SF/2 method is used independent of the compressed mode triggering
mechanism. Selection of the SF/2 method is presented in Table 6.

Triggered by 1st Choice for CM Method 2nd choice for CM Method 3rd Choice for CM Method
UL Tx Power SF/2 - -
DL RL Power SF/2 - -
CPICH Ec/Io SF/2 - -
CPICH RSCP SF/2 - -
Uplink Quality SF/2 - -

Table 6 – Nokia CS data service uplink compressed mode methods (advanced strategy)
Unlike the AMR speech service, CS data services are assumed to be sensitive to throughput and a lower
transport format is not added.

In the case of PS data services, the SF/2 and HLS methods are used. These methods are presented in Table 7.

Triggered by 1st Choice for CM Method 2nd choice for CM Method 3rd Choice for CM Method
UL Tx Power SF/2 and adding of lower TF ½ rate HLS ¾ rate HLS
DL RL Power SF/2 and adding of lower TF ½ rate HLS ¾ rate HLS
CPICH Ec/Io SF/2 ½ rate HLS ¾ rate HLS
CPICH RSCP SF/2 ½ rate HLS ¾ rate HLS
Uplink Quality SF/2 ½ rate HLS ¾ rate HLS

Table 7 – Nokia PS data service uplink compressed mode methods (advanced strategy)
The preferred method is SF/2. If the triggering mechanism is uplink or downlink transmit power then a lower
transport format may also be added to help reduce the transmit power requirement. If the spreading factor
being used prior to entering compressed mode is equal to 4 then the SF/2 method is not possible and HLS
must be used. ½ rate HLS corresponds to transmitting 8 slots out of the 15 slots (single frame method when
using a transmission gap length of 7 slots) and ¾ rate HLS corresponds to transmitting 11 slots out of the 15
(double frame method when using a transmission gap length of 7 slots). Nokia’s RAN1.5.2.ED2
implementation includes the RNC databuild parameter HLSModeSelection, to define whether or not ½ data
rate HLS can be used. The HLSModeSelection parameter is presented in Table 8.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable HLSModeSelection RNC Yes 0 (1/2 HLS allowed), 1 0
(1/2 HLS not allowed)

Table 8 – Use of ½ data rate higher layer scheduling


If the HLSModeSelection parameter specifies that ½ data rate HLS is not permitted then ¾ rate HLS must be
used.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 16 of 52

4.1.2. Downlink

3GPP TS25.331 and TS25.433 specify that downlink compressed mode transmission gaps may be achieved
using either puncturing, spreading factor division by 2 (SF/2) or higher layer scheduling (HLS). These
options are presented in Table 9.
3GPP Name Description 3GPP
Range
Downlink Compressed Defines the Downlink compressed mode method Puncturing,
Mode Method SF/2, HLS

Table 9 – 3GPP downlink compressed mode method parameter


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message and to the UE within the RRC: PHYSICAL/TRANSPORT
CHANNEL RECONFIGURATION message.

Nokia’s normal compressed mode strategy configures the compressed mode method based upon service
type. No account is taken of the mechanism that has triggered compressed mode. Table 10 presents the
compressed mode methods used by the normal strategy.
AMR Speech RT Data NRT Data
Downlink Compressed SF/2 SF/2 HLS
Mode Method

Table 10 – Nokia downlink compressed mode methods (normal strategy)


Nokia’s advanced compressed mode strategy configures the compressed mode method based upon service
and triggering mechanism. Nokia’s RAN1.5.2.ED2 supports five coverage reason compressed mode
triggering mechanisms – high uplink transmit power, high downlink transmit power, low CPICH Ec/Io, low
CPICH RSCP and low uplink quality.

In the case of the AMR speech service, the preferred compressed mode method is SF/2 for all triggering
mechanisms except the downlink transmit power mechanism. In the case of the downlink transmit power
mechanism, the preferred compressed mode method is puncturing. Puncturing has been selected to help
avoid the increased downlink transmit power requirement associated with a reduction in spreading factor.
Table 11 presents the compressed mode methods for the AMR speech service.

Triggered by 1st Choice for CM Method 2nd choice for CM Method 3rd Choice for CM Method
UL Tx Power SF/2 SF/2 with alternative SC -
DL Tx Power Puncturing for lower AMR mode SF/2 SF/2 with alternative SC
CPICH Ec/Io SF/2 SF/2 with alternative SC -
CPICH RSCP SF/2 SF/2 with alternative SC -
Uplink Quality SF/2 SF/2 with alternative SC -

Table 11 – Nokia AMR speech service downlink compressed mode methods (advanced strategy)
If puncturing is used then sufficient bits must be removed to allow the use of a lower AMR mode. The lower
AMR mode is limited by the RNC databuild parameter presented in Table 12.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable MinDLAMRmodeCM RNC No 1(12.2 kbps), 2(10.2 kbps), 1
4(7.95 kbps), 8(7.4 kbps),
16(6.7 kbps), 32(5.9 kbps),
64(5.15 kbps), 128(4.75 kbps)

Table 12 – AMR speech service downlink compressed mode lower transport format bit rate

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 17 of 52

The LowerDLAMRmodeCM parameter is currently configured with a value of 1, i.e. an AMR bit rate of 12.2
kbps. This means that a lower AMR mode is not used and the second preferred compressed mode method is
considered. If puncturing is used as a downlink compressed mode method then the RNC databuild parameter
DlpuncturingLimit is used to limit the number of bits that can be removed. This parameter is presented in
Table 13.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable DLpuncturingLimit RNC Yes 0 (0.4), 1 (0.44), 2 (0.48), 7
3 (0.52), 4 (0.56), 5 (0.6),
6 (0.64), 7 (0.68. default),
8 (0.72), 9 (0.76), 10 (0.8),
11 (0.84), 12 (0.88), 13 (0.92),
14 (0.96), 15 (1)

Table 13 – Downlink puncturing limit


If the puncturing method is not used for the downlink transmit power triggering mechanism then the SF/2
method is applied for all triggering mechanisms. If possible, the same scrambling code is used to ensure that
downlink orthogonality is maintained. If the required channelisation code is not available then an alternative
scrambling code may be used. Downlink orthogonality is lost for the compressed radio link if an alternative
scrambling code is used. The RNC databuild parameter AltScramblingCodeCM defines whether or not an
alternative scrambling code can be used. This parameter is presented in Table 14.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable AltScramblingCodeCM WCEL Yes 0 (allowed), 0
1 (not allowed)

Table 14 – Use of an alternative scrambling code for downlink compressed mode


In the case of CS data services, the puncturing method is preferred for the uplink and downlink transmit
power triggering mechanisms. The SF/2 method is preferred for the CPICH and uplink quality triggering
mechanisms. It is assumed that CS data services are throughput sensitive and are unable to incur the
throughput reduction associated with HLS. Table 15 presents the compressed mode methods for CS data
services.

Triggered by 1st Choice for CM Method 2nd choice for CM Method 3rd Choice for CM Method
UL Tx Power Puncturing SF/2 SF/2 with alternative SC
DL RL Power Puncturing SF/2 SF/2 with alternative SC
CPICH Ec/Io SF/2 SF/2 with alternative SC -
CPICH RSCP SF/2 SF/2 with alternative SC -
Uplink Quality SF/2 SF/2 with alternative SC -

Table 15 – CS data service downlink compressed mode methods (advanced strategy)


If puncturing is applied, the RNC databuild parameter, DlpuncturingLimit is used to limit the number of bits
that can be removed. This parameter is presented in Table 13. If SF/2 is applied, the use of an alternative
scrambling code is controlled using the RNC databuild parameter, AltScramblingCodeCM. This parameter is
presented in Table 14.

In the case of PS data services, the puncturing method is preferred for the uplink transmit power triggering
mechanism. ½ rate HLS is preferred for the downlink transmit power triggering mechanism. SF/2 is
preferred for the remaining triggering mechanisms. Table 16 presents the compressed mode methods for PS
data services.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 18 of 52

Triggered by 1st Choice for CM 2nd choice for CM 3rd Choice for 4th Choice for CM 5th Choice for CM
Method Method CM Method Method Method
UL Tx Power Puncturing SF/2 SF/2 with ½ HLS ¾ HLS
alternative SC
DL RL Power ½ HLS ¾ HLS - - -
CPICH Ec/Io SF/2 SF/2 with ½ HLS ¾ HLS -
alternative SC
CPICH RSCP SF/2 SF/2 with ½ HLS ¾ HLS -
alternative SC
Uplink Quality SF/2 SF/2 with ½ HLS ¾ HLS -
alternative SC

Table 16 – PS data service downlink compressed mode methods (advanced strategy)


If SF/2 is applied, the use of an alternative scrambling code is controlled using the RNC databuild parameter,
AltScramblingCodeCM. This parameter is presented in Table 14. When the preference is to use SF/2 and
there are other options available, the RNC databuild parameter PowerMarginCMcodeHalving, is used to
determine whether or not SF/2 is used. This parameter is presented in Table 17.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable PowerMarginCMcodeHalving RNC Yes 0..10 dB, step 0.25 dB 3 dB

Table 17 – Downlink transmit power margin required to half the spreading factor
Based upon the default value of this parameter, if the downlink transmit power is within 3 dB of the
maximum, HLS is used. Nokia’s RAN1.5.2.ED2 implementation includes the HLSModeSelection RNC
databuild parameter to define whether or not ½ data rate HLS can be used. This parameter is presented in
Table 8.

4.2. Timing

3GPP TS25.331 and TS25.433 specify that the RNC is able to configure a maximum of 6 compressed mode
transmission gap pattern sequences. These pattern sequences are identified by their Transmission Gap
Pattern Sequence Identifier (TGPSI). A different pattern sequence can be defined for each Transmission Gap
Measurement Purpose (TGMP). Nokia’s RAN1.5.2.ED2 implementation makes use of three compressed
mode pattern sequences to support three measurement purposes. TGPSI 1 is used for FDD inter-frequency
measurements, TGPSI 2 is used for GSM RSSI measurements and TGPSI 3 is used for GSM BSIC
identification. Nokia’s RAN1.5.2.ED2 does not support GSM BSIC re-confirmation.

3GPP TS25.215 specifies that each transmission gap pattern sequence includes two transmission gap
patterns. Each transmission gap pattern includes one or two transmission gaps. The transmission gap patterns
are repeated according to the value of the Transmission Gap Pattern Repetition Count (TGPRC). Figure 6
illustrates the transmission gap patterns within a transmission gap pattern sequence.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 19 of 52

#1 #2 #3 #4
TG Pattern 1 TG Pattern 2 TG Pattern 1 TG Pattern 2

TGSN TGL1 TGL2 TGSN TGL1 TGL2

TGD TGD

TGPL1 TGPL2

Figure 6 – 3GPP Transmission gap patterns within a transmission gap pattern sequence
The 3GPP specified parameters which define the timing of the transmission gap patterns and the
transmission gap pattern sequences are presented in Table 18.

3GPP 3GPP Name Description 3GPP


Acronym Range
TGSN Transmission Gap Starting Slot number of the first transmission gap slot within the first radio frame of Integer
Slot Number the transmission gap pattern (0 to 14)
TGL1 Transmission Gap Length 1 Duration of the first transmission gap expressed in terms of the number of Integer
slots. (1 to 14)
TGL2 Transmission Gap Length 2 Duration of the second transmission gap expressed in terms of the number Integer
of slots. If not configured by higher layers then assumed to equal TGL1. (1 to 14)
TGD Transmission Gap start Duration between starting slots of two consecutive transmission gaps within Integer
Distance a transmission gap pattern expressed in terms of the number of slots. A (0, 15 to
value of 0 indicates only 1 transmission gap in the transmission gap pattern 269)
TGPL1 Transmission Gap Pattern Duration of transmission gap pattern 1 expressed in terms of the number of Integer
Length 1 frames. (1 to 144,…)
TGPL2 Transmission Gap Pattern Duration of transmission gap pattern 2 expressed in terms of the number of Integer
Length 2 frames. If not configured by higher layers then TGPL2 = TGPL1. (1 to 144,…)
TGPRC Transmission Gap Pattern Number of transmission gap patterns within the transmission gap Integer
Repetition Count sequence. A value of 0 corresponds to an infinite number until requested to (0 to 511)
stop.
TGCFN Transmission Gap CFN of the first radio frame of the first pattern 1 within the transmission gap Integer
Connection Frame Number pattern sequence. (0 to 255)

Table 18 – 3GPP transmission gap pattern and transmission gap pattern sequence timing parameters
The TGSN, TGL1, TGL2, TGD, TGPL1 and TGPL2 are provided to the Node B within the Dedicated
NBAP: RADIO LINK RECONFIGURATION PREPARE message. The TGPRC and TGCFN are provided
to the Node B within the Dedicated NBAP: COMPRESSED MODE COMMAND message.

The TGSN, TGL1, TGL2, TGD, TGPL1, TGPL2 and TGPRC are provided to the UE within the RRC:
PHYSICAL/TRANSPORT CHANNEL RECONFIGURATION message. The TGCFN is provided to the UE
within the RRC: MEASUREMENT CONTROL message.

The transmission gaps may be located such that they do or do not span two consecutive radio frames. The
former is known as the double frame approach whereas the latter is known as the single frame approach. In
each case the transmission gap must be defined such that at least 8 slots are transmitted in each radio frame.

Nokia’s RAN1.5.2.ED2 implementation of these timing parameters includes a set of RNC databuild
parameters which are configurable and a set which are non-configurable. These RNC databuild parameters
are presented in Table 19.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 20 of 52

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
TGSN (for single GapPositionSingleFrame RNC Yes 0..8 , step 1 slot 4
frame method)
TGSN (for double No parameter name. RNC No - slot 11
frame method)
TGL1 (for single & No parameter name. RNC No - 7 slots
double frame method)
TGL2 (for single & Not applicable because TGD defined - - - -
double frame method) as 0.
TGD (for single & No parameter name. RNC No - 0 slots
double frame method)
TGPL1 (for single TGPLsingleframeAMRgsm RNC Yes 2..144 , step 1 4 frames
frame method) TGPLsingleframeAMRinterFreq RNC Yes 2..18 , step 1 4 frames
TGPLsingleframeNRTPSgsm RNC Yes 2..144 , step 1 4 frames
TGPLsingleframeNRTPSinterFreq RNC Yes 2..18 , step 1 4 frames
TGPLsingleframeCSgsm RNC Yes 2..144 , step 1 4 frames
TGPLsingleframeCSinterFreq RNC Yes 2..18 , step 1 4 frames
TGPLsingleframeRTPSgsm RNC Yes 2..144 , step 1 4 frames
TGPLsingleframeRTPSinterFreq RNC Yes 2..18 , step 1 4 frames
TGPL1 (for double TGPLdoubleframeAMRgsm RNC Yes 3..144 , step 1 4 frames
frame method) TGPLdoubleframeAMRinterFreq RNC Yes 3..18 , step 1 4 frames
TGPLdoubleframeNRTPSgsm RNC Yes 3..144 , step 1 4 frames
TGPLdoubleframeNRTPSinterFreq RNC Yes 3..18 , step 1 4 frames
TGPL2 (for single Not configured and so TGPL2 = - - - -
frame method) TGPL1

TGPL2 (for double Not configured and so TGPL2 = - - - -


frame method) TGPL1
TGPRC (for single & No parameter name. RNC No - 0
double frame method) (infinite)

Table 19 – Nokia transmission gap pattern and transmission gap pattern sequence timing parameters
Nokia’s use of the single and double frame transmission gap approaches is a function of the compressed
mode method. Compressed mode methods are described in Section 4.1. If spreading factor division by 2 or ½
data rate higher layer scheduling is used then the single frame approach is adopted. If puncturing or ¾ data
rate higher layer scheduling is used then the double frame approach is adopted.

Figure 7 illustrates the timing of the single and double frame methods based upon the default Nokia
parameter set.

TGPL1 = 4 frames

Single
frame
method
TGL1=
7 slots

Double
frame
method
TGL1=
7 slots

Figure 7 – Nokia default transmission gap patterns within a transmission gap pattern sequence
3GPP TS25.331 specifies that when the transmission gap pattern sequence has been configured for GSM
BSIC identification, a limit is placed upon the maximum number of pattern repetitions that the UE is allowed
to decode the unknown BSIC. 3GPP TS25.331 also specifies that when the transmission gap has been
configured for GSM BSIC re-confirmation, a limit is placed upon the maximum time that the UE is allowed
to re-confirm the BSIC. These limits are defined by the parameters presented in Table 20.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 21 of 52

3GPP Name Description 3GPP


Range
N Identify abort Indicates the maximum number of repeats of patterns that the UE shall use Integer
to attempt to decode the unknown BSIC of the GSM cell in the initial BSIC (1 to 128)
identification procedure.
T Reconfirm abort Indicates the maximum time allowed for the re-confirmation of the BSIC of Real
one GSM cell in the BSIC re-confirmation procedure. The time is given in (0.5 to 10.0,
steps of 0.5 seconds. step 0.5)

Table 20 – 3GPP parameters used to limit BSIC identification and BSIC re-confirmation
This information is provided to the UE within the RRC: PHYSICAL/TRANSPORT CHANNEL
RECONFIGURATION message.

Nokia’s RAN1.5.2.ED2 implementation supports the BSIC identification measurement purpose but not the
BSIC re-confirmation measurement purpose. The Nokia implementation of the associated parameters is
presented in Table 21.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
N Identify abort No parameter name. RNC No - 128
patterns
T Reconfirm abort Not applicable because BSIC re- - - - -
confirmation is not supported.

Table 21 – Nokia parameters used to limit BSIC identification and BSIC re-confirmation times
The UE is permitted a maximum of 128 pattern repetitions to decode an unknown BSIC. Based upon a 4
frame TGPL, this corresponds to 5.12 s.

4.3. Direction

3GPP TS25.331 and TS25.433 specify that the RNC is able to configure compressed mode such that it is
activated in the uplink alone, in the downlink alone or in both the uplink and downlink. The parameter used
to define the compressed mode direction is defined in Table 22.
3GPP Name Description 3GPP Range
UL/DL mode Defines whether only DL, only UL, or combined UL/DL compressed mode is Enumerated(Code
used. change, No code
change)

Table 22 – 3GPP direction of compressed mode parameter


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message and to the UE within the RRC: PHYSICAL/TRANSPORT
CHANNEL RECONFIGURATION message.

Nokia’s RAN1.5.2.ED2 implementation configures uplink and downlink compressed mode according to the
UE capability information. This information is provided to the RNC as part of the RRC: RRC
CONNECTION SETUP COMPLETE message.

4.4. Power Control

TS25.331, TS25.433 and TS25.214 specify that the RNC is able to configure increases in the uplink and
downlink SIR targets during compressed mode. Table 23 presents the 3GPP parameters which specify the
increase in uplink SIR target.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 22 of 52

3GPP Name Description 3GPP


Range
DeltaSIR1 SIR target increase to be set in the Node B during the frame containing the Integer
start of the first transmission gap in the transmission gap pattern (without (0 to 3 dB,
including the effect of the bit-rate increase) step 0.1)
DeltaSIRafter1 SIR target increase to be set in the Node B one frame after the frame Integer
containing the start of the first transmission gap in the transmission gap (0 to 3 dB,
pattern. step 0.1)
DeltaSIR2 SIR target increase to be set in the Node B during the frame containing the Integer
start of the second transmission gap in the transmission gap pattern (0 to 3 dB,
(without including the effect of the bit-rate increase) When omitted, step 0.1)
DeltaSIR2 = DeltaSIR1.
DeltaSIRafter2 SIR target increase to be set in the Node B one frame after the frame Integer
containing the start of the second transmission gap in the transmission gap (0 to 3 dB,
pattern. When omitted, DeltaSIRafter2 = DeltaSIRafter1. step 0.1)

Table 23 – 3GPP increase in uplink SIR target parameters


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message.

Nokia’s RAN1.5.2.ED2 implementation makes use of only a single transmission gap within each
transmission gap pattern and so the DeltaSIR2 and DeltaSIRafter2 parameters are not applicable. The
relevant Nokia RNC databuild parameters are presented in Table 24.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia Default
DeltaSIR1 ULdeltaSIR1 RNC No 0..3 dB, step 0.1 dB 1.5 dB
DeltaSIRafter1 ULdeltaSIRafter1 RNC No 0..3 dB, step 0.1 dB 0.7 dB for 0 slot
gap, 1.5 dB for 3
slots gap

Table 24 – Nokia increase in uplink SIR target parameters


Table 25 presents the 3GPP parameters which specify the increase in downlink SIR target.
3GPP Name Description 3GPP
Range
DeltaSIR1 SIR target increase to be set in the Node B during the frame containing the Integer
start of the first transmission gap in the transmission gap pattern (without (0 to 3 dB,
including the effect of the bit-rate increase) step 0.1)
DeltaSIRafter1 SIR target increase to be set in the Node B one frame after the frame Integer
containing the start of the first transmission gap in the transmission gap (0 to 3 dB,
pattern. step 0.1)
DeltaSIR2 SIR target increase to be set in the Node B during the frame containing the Integer
start of the second transmission gap in the transmission gap pattern (0 to 3 dB,
(without including the effect of the bit-rate increase) When omitted, step 0.1)
DeltaSIR2 = DeltaSIR1.
DeltaSIRafter2 SIR target increase to be set in the Node B one frame after the frame Integer
containing the start of the second transmission gap in the transmission gap (0 to 3 dB,
pattern. When omitted, DeltaSIRafter2 = DeltaSIRafter1. step 0.1)

Table 25 – 3GPP increase in downlink SIR target parameters


This information is provided to the UE within the RRC: PHYSICAL/TRANSPORT CHANNEL
RECONFIGURATION message.

Nokia’s RAN1.5.2.ED2 implementation makes use of only a single transmission gap within each
transmission gap pattern and so the DeltaSIR2 and DeltaSIRafter2 parameters are not applicable. The
relevant Nokia RNC databuild parameters are presented in Table 26.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 23 of 52

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia Default
DeltaSIR1 DLdeltaSIR1 RNC No 0..3 dB, step 0.1 dB 1.5 dB
DeltaSIRafter1 DLdeltaSIRafter1 RNC No 0..3 dB, step 0.1 dB 0.7 dB for 0 slots
gap, 1.5 dB for 3
slots gap

Table 26 – Nokia increase in downlink SIR target change parameters


TS25.331 and TS25.214 specify that a UE should increase its transmit power at the start of the first slot after
an uplink or downlink transmission gap. The Initial Transmit Power (ITP) mode has an impact upon this
increase. TS25.331 and TS25.214 also specify that a UE should use a modified power control algorithm
during the transmission gap recovery period. The modified power control algorithm is known as the
Recovery Period Power (RPP) control algorithm. The 3GPP ITP and RPP parameters are presented in Table
27.

3GPP 3GPP Name Description Range


Acronym
ITP Initial Transmit Power Initial Transmit Power is the uplink power control method to be used Enumerated
to compute the initial transmit power after the compressed mode (0, 1)
gap.
RPP Recovery Period Power Recovery Period Power control mode during the frame after the Enumerated
transmission gap within the compressed frame. Indicates whether (0, 1)
normal PC mode or compressed PC mode is applied

Table 27 – 3GPP ITP and RPP power control parameters


This information is provided to the UE within the RRC: PHYSICAL/TRANSPORT CHANNEL
RECONFIGURATION message.

Table 28 presents the equivalent Nokia RNC databuild parameters.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
ITP UpLinkInitialTransmitPowerMode RNC Yes 0, 1 0
RPP UpLinkRecoveryPeriodPowerMode RNC Yes 0, 1 0

Table 28 – Nokia ITP and RPP power control parameters

4.5. Scrambling Code

3GPP TS25.213 specifies that when the spreading factor division by 2 method is used for downlink
compressed mode then it is possible for the RNC to configure the use of an alternative scrambling code. The
parameter used to do this is presented in Table 29.
3GPP Name Description 3GPP Range
Scrambling Code Change Indicates whether or not an alternative scrambling code is used for Enumerated
compressed mode method 'SF/2'. (code change,
no code change)

Table 29 – 3GPP alternative scrambling code parameter


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message and to the UE within the RRC: PHYSICAL/TRANSPORT
CHANNEL RECONFIGURATION message.

Nokia’s RAN1.5.2.ED2 implementation includes the RNC databuild parameter, AltScramblingCodeCM to


control whether or not an alternative scrambling code may be used. This parameter is presented in Table 14.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 24 of 52

4.6. Downlink Frame Type

The RNC is able to configure whether downlink frame type ‘A’ or ‘B’ shall be used for downlink
compressed mode. Frame type ‘A’ is intended to maximise the transmission gap length. Frame type ‘B’ is
optimised for power control. Frame types ‘A’ and ‘B’ are illustrated in Figure 8.

Figure 8 – Downlink frame types ‘A’ and ‘B’


The 3GPP frame type configuration parameter is presented in Table 30.
3GPP Name Description 3GPP Range
Downlink Frame Type Defines the downlink frame structure type A, B

Table 30 – 3GPP downlink frame type parameter


This information is provided to the Node B within the Dedicated NBAP: RADIO LINK
RECONFIGURATION PREPARE message and to the UE within the RRC: PHYSICAL/TRANSPORT
CHANNEL RECONFIGURATION message.

Table 31 presents the equivalent Nokia RNC databuild parameter.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Downlink Frame Type No parameter name RNC No - A

Table 31 – Nokia downlink frame type parameter


The downlink frame type is configured independently of the downlink slot format. Downlink slot structure
‘A’ is used for HLS and puncturing whereas downlink slot structure ‘B’ is used for SF/2. In the uplink
direction, DPDCH slot structures are not defined specifically for compressed mode. Uplink DPCCH slot
structures are defined specifically for compressed mode. The choice of uplink DPCCH slot structure depends
upon the number of slots transmitted and whether or not TFCI bits are transmitted.

4.7. Measurement Reporting

3GPP TS25.331 specifies the parameter to define the interval between periodic measurement reports. This
parameter is presented in Table 32.
3GPP Name Description 3GPP Range
Periodic Reporting Interval Defines the interval between periodic measurement 250, 500, 1000, 2000, 3000, 4000,
reports 6000, 8000, 12000, 16000, 20000,
24000, 28000, 32000, 64000

Table 32 – 3GPP periodic measurement reporting interval

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 25 of 52

This information is provided to the UE within the RRC: MEASUREMENT CONTROL message.

Nokia’s RAN1.5.2.ED2 implementation includes RNC databuild parameters that can be defined
independently for inter-frequency and inter-system measurements. In addition to the periodic reporting
interval there are also parameters that define, the minimum number of measurement reports that must be
received by the RNC prior to making a handover decision and the maximum number of measurement reports
than can be received prior to the RNC cancelling the measurement procedure. These parameters are
presented in Table 33.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Periodic Reporting InterFreqMeasRepInterval FMCI Yes 2 (0.5 s), 3 (1 s), 4 2
Interval (2 s), 5 (3 s), 6 (4 s)
Not applicable InterFreqNcellSearchPeriod FMCI Yes 0..20 MeasReport, 0
step 1 MeasReport
Not applicable InterFreqMaxMeasPeriod FMCI Yes 1..120 MeasReport, 20
step 1 MeasReport
Periodic Reporting GsmMeasRepInterval FMCG Yes 2 (0.5 s), 3 (1 s), 4 2
Interval (2 s), 5 (3 s), 6 (4 s)
Not applicable GsmNcellSearchPeriod FMCG Yes 0..20 MeasReport, 0
step 1 MeasReport
Not applicable GsmMaxMeasPeriod FMCG Yes 1..120 MeasReport, 20
step 1 MeasReport

Table 33 – Nokia periodic measurement reporting parameters


The default value for the minimum number of reports required prior to making a handover decision is 0. This
means that the RNC can proceed immediately after receiving the first measurement report. The default value
for the maximum number of periodic measurement reports is 20. The RNC will cancel the measurement
procedure if handover hasn’t been completed after receiving 20 measurement reports.

4.8. Other Configuration Parameters

Nokia’s RAN1.5.2.ED2 implementation includes a number of additional RNC databuild parameters which
are applicable to compressed mode. These are presented in Table 34.

3GPP Name Nokia Name Scope Configurable Nokia Range Nokia


Default
Not applicable CMmasterSwitch RNC Yes 0 (false), 1 (true) 1
Not applicable MaxNumbUECMcoverHO WCEL Yes 0..255 , step 1 16
Not applicable PrxTarget WCEL Yes 0..30 dB step 0.1 4 dB
Not applicable PrxOffset WCEL Yes 0..6 dB step 0.1 1 dB
Not applicable PtxTarget WCEL Yes -10..50 dBm step 0.1 40 dBm
Not applicable PtxOffset WCEL Yes 0..6 dB step 0.1 1 dB

Table 34 – Additional Nokia RNC databuild parameters applicable to compressed mode


The CMmasterSwitch parameter defines whether or not compressed mode is used within any cell connected
to the RNC. The MaxNumbUECMcoverHO parameter defines the maximum number of UE which are
permitted to simultaneously use compressed mode within a single cell. If a UE is in soft handover then this
limit applies to all cells to which the UE is connected.

If a Node B uplink RSSI measurement exceeds PrxTarget while MaxNumbUECMcoverHO has not been
reached, then one more UE can have compressed mode activated within that radio resource indication period.
If ]a Node B uplink RSSI measurement exceeds PrxTarget + PrxOffset then no more UE can have
compressed mode activated within that radio resource indication period. Similar rules are applied in the
downlink direction using PtxTarget and PtxOffset.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 26 of 52

5. Preferred Compressed Mode Methods and Patterns


This section identifies the trade-offs associated with adopting various strategies in terms of compressed
mode methods and compressed mode patterns. The scenarios identified in section 3 are used as a basis for
balancing each trade-off and generating a set of preferred compressed mode configurations. These scenarios
have four dimensions, i.e.

 rate of change of propagation conditions: high or low


 service: AMR speech, RT data or NRT data
 triggering mechanism: uplink transmit power, downlink transmit power, CPICH Ec/Io, CPICH
RSCP or uplink quality
 measurement purpose: GSM RSSI measurement, GSM BSIC identification or FDD measurement

Recommendations are made specifically for each of these four dimensions. The discussion is primarily
theoretical and is not limited by Nokia’s implementation. The field experience presented in section 7 is used
to a limited extent.

5.1. Method

Section 4.1 introduced the three methods that may be used to generate compressed mode transmission gaps,
i.e. puncturing, spreading factor division by 2 (SF/2) and higher layer scheduling (HLS). The choice of
which method to adopt is one of the most important compressed mode design decisions. The choice cannot
be made in isolation but requires consideration of other compressed mode design decisions. For example,
selecting puncturing as a compressed mode method has implications upon the maximum transmission gap
length and whether the transmission gaps need to span single or multiple radio frames. The choice of
compressed mode method must also account for limitations imposed by other areas of system design. For
example, the AMR speech service must use fixed downlink transport channel starting positions if the UE is
required to apply blind detection of the AMR bit rate. The R99 3GPP specifications do not support HLS with
fixed transport channel starting positions and so this method must be discounted from the selection process.

5.1.1. Uplink

Uplink compressed mode methods are limited to spreading factor division by 2 (SF/2) and high layer
scheduling (HLS).

SF/2 compressed mode does not necessitate the use of a lower transport format combination. The physical
layer is reconfigured to convey the same quantity of data within a shorter period of time. The reduction in
spreading factor is applied on a per radio frame basis rather than on a per transmission time interval (TTI)
basis. For example, if the 20 ms TTI AMR speech service were to use SF/2 compressed mode configured
using the single frame approach, a single 7 slot transmission gap and a 4 frame transmission gap pattern
length, then every 60 slots, 8 slots would be transmitted with a halved spreading factor, 7 slots would not be
transmitted and 45 slots would be transmitted using the non-compressed mode format. The 8 slots
transmitted with a halved spreading factor would require approximately 3 dB greater transmit power than the
slots transmitted using the non-compressed mode format. This short term increase in UE transmit power
means that SF/2 is not suitable for scenarios in which the UE is transmitting close to its maximum transmit
power capability. It may be possible to combine SF/2 with a reduction in transport format combination to
decrease the UE transmit power requirement. The transport format combination can be reduced for services
which are not sensitive to changes in throughput, i.e. AMR speech and NRT data services. SF/2 compressed

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 27 of 52

mode is applicable to all services except those that use a spreading factor of 4. The field experience
presented in Section 7 illustrates that the average UE transmit power increases by 1.3 dB when SF/2 is used
with the single frame approach and a 7 slot transmission gap. This increase in average transmit power is
generated by the compressed mode inner loop power control behaviour rather than the short term increase
caused by the spreading factor reduction. SF/2 compressed mode may use either the single or double frame
approach although it is usually associated with the single frame approach to limit the number of slots which
are transmitted with a reduced spreading factor. Relatively large transmission gaps can be generated if the
double frame approach is used.

HLS necessitates the use of a lower transport format combination while the physical layer remains
unchanged, i.e. less data is transmitted within a shorter period of time. The MAC layer within the UE is
responsible for selecting an appropriate lower transport format combination. Layer 3 of the RNC is not
required to remove any higher transport format combinations. The result of using a lower transport format
combination is a reduction in service throughput. The R99 3GPP specifications do not support HLS for
transport channels using fixed starting positions. The AMR speech, RT data and NRT data services all use
flexible transport channel starting positions in the uplink direction. HLS is not appropriate for RT data
services because of their sensitivity to variations in throughput. HLS can be applied to the AMR speech
service if a lower AMR bit rate can be used. The cost of using a lower AMR bit rate is a reduction in speech
quality. HLS can be applied to NRT data services because of their ability to incur variations in throughput.
The field experience presented in Section 7 illustrates that the average UE transmit power increases by
approximately 1.6 dB when HLS is used with the single frame approach and a 7 slot transmission gap. This
increase in average transmit power is generated by the compressed mode inner loop power control
behaviour. HLS can use either the single or double frame approach. Relatively large transmission gaps can
be generated if the double frame approach is used.

Table 35 presents a summary of the uplink compressed mode methods which may be used for each service.

Compressed Mode Method


Service HLS SF/2
AMR Speech Available* Available
RT Data Not Available Available
NRT Data Available Available
* only available if it is possible to use a lower AMR bit rate

Table 35 – Summary of uplink compressed mode methods


Table 36 presents the preferred methods for generating uplink compressed mode transmission gaps These are
independent of the compressed mode measurement purpose and the rate of change of propagation conditions.

Compressed Mode Triggering Mechanism


Service UL Transmit DL Transmit CPICH Ec/Io CPICH RSCP UL Quality
Power Power
AMR Speech HLS* HLS* HLS* HLS* HLS*
RT Data SF/2 SF/2 SF/2 SF/2 SF/2
NRT Data HLS HLS HLS HLS HLS
* only if it is possible to use a lower AMR bit rate. Otherwise SF/2 should be used.

Table 36 – Preferred uplink compressed mode methods


SF/2 is the only compressed mode method that is available for RT data services. HLS has been chosen for
the AMR speech service on the basis that it is acceptable to incur a temporary speech quality reduction. If
AMR bit rates below 12.2 kbps are not supported by either the network or the UE then SF/2 must be used by
the AMR speech service. HLS has been chosen for NRT data services on the basis that it is acceptable to
incur a temporary reduction in throughput.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 28 of 52

5.1.2. Downlink

Downlink compressed mode methods are puncturing, spreading factor division by 2 (SF/2) and high layer
scheduling (HLS).

The R99 3GPP specifications do not support puncturing for transport channels using flexible starting
positions. The RT data and NRT data services use flexible transport channel starting positions in the
downlink. The AMR speech service uses fixed transport channel starting positions in the downlink.
Puncturing removes layer 1 bits after channel coding and causes an increase in the downlink SIR target.
Puncturing is applied on a per TTI basis. This means that the double and single frame approaches have
similar impacts if the double frame approach transmission gap spans two frames belonging to the same TTI.
If the double frame approach is synchronised with the TTI such that the transmission gap spans two frames
belonging to different TTI then the puncturing ratio is reduced. However, in this case two TTI rather than
one are affected. The precise extent to which coded transport blocks can be punctured for compressed mode
depends upon the non-compressed mode rate matching attributes and the maximum acceptable increase in
SIR target. If the non-compressed mode rate matching attributes indicate repetition rather than puncturing
then there is increased scope for puncturing for compressed mode. Typically, the maximum total puncturing
applied to a transport channel is such that approximately 70 % of the coded bits remain after puncturing.
Puncturing for the AMR speech service may necessitate the use of a lower AMR bit rate.

Similar to the uplink, downlink SF/2 compressed mode requires the physical layer to be reconfigured to
convey the same quantity of data within a shorter period of time. The reduction in spreading factor is applied
on a per radio frame basis rather than on a per transmission time interval (TTI) basis. For example, if the
AMR speech service were to use SF/2 compressed mode configured using the single frame approach, a
single 7 slot transmission gap and a 4 frame transmission gap pattern length, then every 60 slots, 8 slots
would be transmitted with a halved spreading factor, 7 slots would not be transmitted and 45 slots would be
transmitted using the non-compressed mode format. The 8 slots transmitted with a halved spreading factor
would require approximately 3 dB greater transmit power than the slots transmitted using the non-
compressed mode format. This short term increase in downlink transmit power means that SF/2 is not
suitable for scenarios in which the cell is transmitting close to its maximum radio link transmit power. It may
be possible to combine SF/2 with a reduction in transport format combination to decrease the cell transmit
power requirement. The transport format combination can be reduced for services which are not sensitive to
changes in throughput, i.e. AMR speech and NRT data services. SF/2 compressed mode is applicable to all
services except those that use a spreading factor of 4. The increase in average transmit power during SF/2
compressed mode is generated by the inner loop power control behaviour rather than the short term increase
caused by the spreading factor reduction. SF/2 compressed mode may use either the single or double frame
approaches although it is usually associated with the single frame format to limit the number of slots which
are transmitted with a reduced spreading factor. SF/2 in the downlink direction may require the use of an
alternative scrambling code. There is a one-to-one relationship between the original channelisation code and
the channelisation code with half the spreading factor. If the channelisation code with half the spreading
factor is blocked by another connection then it must be applied with an alternative scrambling code. The use
of an alternative scrambling code means that downlink orthogonality is lost for that radio link and downlink
transmit powers for all radio links will increase.

Similar to the uplink, downlink HLS necessitates the use of a lower transport format combination while the
physical layer remains unchanged, i.e. less data is transmitted within a shorter period of time. The MAC
layer within the RNC is responsible for selecting an appropriate lower transport format combination. Layer 3
of the RNC is not required to remove the higher transport format combinations. The result of using a lower
transport format combination is a reduction in service throughput. The R99 3GPP specifications do not
support HLS for transport channels using fixed transport channel starting positions. The RT data and NRT

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 29 of 52

data services make use of flexible transport channel starting positions in the downlink. The AMR speech
services makes use of fixed transport channel starting positions in the downlink. Fixed transport channel
positions are used to allow blind detection of the AMR bit rate at the UE. HLS is not appropriate for RT data
services because of their sensitivity to variations in throughput. HLS can be applied to NRT data services
because of their ability to incur variations in throughput. The field experience presented in Section 7
illustrates that the average cell transmit power increases by approximately 6.5 dB when HLS is used with the
single frame approach and a 7 slot transmission gap. This increase is considerably greater than expected and
warrants further investigation. The increase in average transmit power is generated by the compressed mode
inner loop power control behaviour. HLS can use either the single or double frame approach. Relatively
large transmission gaps can be generated if the double frame approach is used

Table 37 presents a summary of the compressed mode methods which may be used with each service type.

Compressed Mode Method


Service Puncturing HLS SF/2
AMR Speech Available Not Available Available
RT Data Not Available Not Available Available
NRT Data Not Available Available Available

Table 37 – Summary of downlink compressed mode methods


Table 38 presents the preferred methods for generating downlink compressed mode transmission gaps These
are independent of the compressed mode measurement purpose and the rate of change of propagation
conditions.

Compressed Mode Triggering Mechanism


Service UL Transmit DL Transmit CPICH Ec/Io CPICH RSCP UL Quality
Power Power
AMR Speech Puncturing Puncturing Puncturing Puncturing Puncturing
RT Data SF/2 SF/2 SF/2 SF/2 SF/2
NRT Data HLS HLS HLS HLS HLS

Table 38 – Preferred downlink compressed mode methods


SF/2 is the only compressed mode method that is available for RT data services. Puncturing has been chosen
for the AMR speech service because it avoids the transient increase in downlink transmit power requirement
associated with SF/2. HLS has been chosen for NRT data services on the basis that it is acceptable to incur a
temporary reduction in throughput.

5.2. Timing

Section 4.2 introduced the parameter set associated with the timing of compressed mode patterns. This
parameter set is summarised below:

 TGSN
 TGL1
 TGL2
 TGD
 TGPL1
 TGPL2
 TGPRC
 TGCFN
 N Identify Abort
 T Reconfirm Abort

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 30 of 52

The most important of these are the TGL, TGPL and TGSN.

Selecting an appropriate TGL involves balancing the efficiency with which measurements can be made with
the impact upon layer 1 and layer 2 performance. Large TGL are more efficient in terms of the number of
measurements that can be completed per unit time but have a greater impact upon layer 1 and layer 2. Layer
1 performance is affected in terms of inner loop power control independent of the compressed mode method.
Larger TGL result in less correlation between the propagation conditions before and after the transmission
gap. This makes it more difficult for the inner loop power control to recover. In the case of compressed mode
by puncturing, larger TGL require larger quantities of puncturing and result in larger SIR target increases. In
the case of compressed mode by HLS, larger TGL require the use of lower transport format combinations
and result in lower layer 2 throughput. In the case of compressed mode by SF/2, larger TGL require the use
of the double frame approach meaning that two radio frames rather than a single radio frame have their
spreading factor reduced. The fundamental requirement for a specific TGL originates from the compressed
mode measurement purpose.

Table 39 presents the relationship between the TGL and the minimum requirement for the UE’s ability to
sample GSM RSSI. These figures have been extracted from 3GPP TS25.133. The third column represents
the efficiency with which measurements are made. Also included in the table is the equivalent time required
to complete 16 and 32 GSM RSSI measurements based upon 3 samples per measurement and a TGPL of 4
radio frames.

TGPL = 4 Frames
Transmission Number of GSM carrier Number of GSM carrier Time to complete 16 Time to complete 32
Gap Length RSSI samples in each RSSI samples in each GSM RSSI meas. (3 GSM RSSI meas. (3
(slots) gap gap per slot samples / meas.) samples / meas.)
3 1 0.33 1920 ms 3840 ms
4 2 0.50 960 ms 1920 ms
5 3 0.60 640 ms 1280 ms
7 6 0.86 320 ms 640 ms
10 10 1.00 200 ms 400 ms
14 15 1.07 160 ms 280 ms

Table 39 – TGL impact upon GSM RSSI measurement


GSM RSSI measurements are made without acquiring GSM synchronisation and do not require the
compressed mode transmission gap to coincide with a particular section of the GSM radio frame. The
measurement efficiency becomes relatively poor for transmission gap lengths of less than 7 slots.
Quantifying the impact upon inner loop power control requires either link level simulations or field
measurements. The field measurements presented in Section 7 indicate that a TGL of 7 slots with a TGPL of
4 frames degrades inner loop power control to the extent that the average uplink transmit power increases by
between 1.3 to 1.6 dB.

Table 40 presents the preferred GSM RSSI TGL as a function of the service and compressed mode method.
These are independent of the compressed mode triggering mechanism and the rate of change of propagation
conditions.

Uplink Downlink
Service HLS SF/2 Puncturing HLS SF/2
AMR Speech 7 7 7 Not applicable 7
RT Data Not applicable 7 Not applicable Not applicable 7
NRT Data 10 10 Not applicable 10 10

Table 40 – Preferred TGL for GSM RSSI measurements

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 31 of 52

A TGL of 7 slots has been selected for the AMR speech and RT data services. This value has been selected
on the basis of balancing the efficiency with which measurements can be made with the impact upon inner
loop power control. A TGL of 7 slots can be generated relatively comfortably by each compressed mode
method and may use either the single or double frame approach. A TGL of 10 slots has been selected for
NRT services. This value has been selected on the basis that NRT services are less sensitive to L1 and L2
performance. NRT services can make use of layer 2 re-transmissions if the performance of inner loop power
becomes poor. NRT services are also able to incur a decreased layer 2 throughput if in the case of HLS, a 10
slot transmission gap requires a lower transport format combination than a 7 slot transmission gap.

In the case of the BSIC identification compressed mode measurement purpose, the frame structure and
timing of the GSM system has a more significant impact upon the required transmission gap length. The
GSM system is based upon an 8 slot radio frame structure with a duration of 4.615 ms. The first slot of each
frame is dedicated to the BCCH. The BSIC is broadcast periodically within the SCH of the BCCH. There are
51 GSM frames within a GSM multi-frame and the SCH is transmitted 5 times within each multi-frame.
Assuming that frame numbering starts at 0 then the frames occupied by the SCH are numbers 1, 11, 21, 31
and 41. The UE has no knowledge of the timing of the GSM system and must capture 9 slot’s worth of GSM
data to be sure of capturing the BCCH. A compressed mode TGL of 7 slots is equivalent to 4.667 ms and
provides a high probability of capturing the BCCH. Shorter transmission gap lengths could be used but doing
so would reduce the probability of capturing the BCCH. The fact that the BSIC is broadcast 5 times per 51
frames means that multiple transmission gaps are likely to be required.

Table 41 presents the relationship between the TGL and the BSIC identification time that guarantees the UE
at least two attempts at decoding the BSIC. These figures have been extracted from 3GPP TS25.133.
TGL Guaranteed attempts TGPL Number of Equivalent time
(slots) at decoding the BSIC (frames) transmission
gap patterns
7 2 3 51 1.5 s
7 2 8 65 5.2 s
10 2 12 23 2.7 s
14 2 8 22 1.8 s
14 2 24 21 5.0 s

Table 41 – TGL impact upon GSM BSIC identification


In practise the BSIC identification times may be less than those presented in Table 41. It is possible that the
UE manages to identify the BSIC within the first transmission gap. Longer TGL and shorter TGPL result in
more rapid BSIC identification times. The drawback of using long TGL and short TGPL is the associated
impact upon layer 1 and layer 2 performance.

Table 42 presents the preferred BSIC identification TGL as a function of the service and compressed mode
method. These are independent of the compressed mode triggering mechanism and the rate of change of
propagation conditions.

Uplink Downlink
Service HLS SF/2 Puncturing HLS SF/2
AMR Speech 7 7 7 Not applicable 7
RT Data Not applicable 7 Not applicable Not applicable 7
NRT Data 10 10 Not applicable 10 10

Table 42 – Preferred TGL for GSM BSIC identification


A TGL of 7 slots has been selected for the AMR speech and RT data services. This value has been selected
on the basis of balancing the efficiency with which BSIC can be identified with the impact upon inner loop
power control. A TGL of 7 slots can be generated relatively comfortably by each compressed mode method
and may use either the single or double frame approach. A TGL of 10 slots has been selected for NRT

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 32 of 52

services. This value has been selected on the basis that NRT services are less sensitive to L1 and L2
performance. NRT services can make use of layer 2 re-transmissions if the performance of inner loop power
becomes poor. NRT services are also able to incur a decreased layer 2 throughput if in the case of HLS, a 10
slot transmission gap requires a lower transport format combination than a 7 slot transmission gap.

In the case of the FDD inter-frequency compressed mode measurement purpose, 3GPP TS25.133 specifies
that the UE must be capable of completing 6 inter-frequency CPICH measurements within Tmeasurement_inter ms.

 480ms 
Tmeasurement _ int er = MAX 480ms ,50ms × × N Freq ms
 Tint er 

Where,
TInter = the minimum time available for inter-frequency measurements within a 480 ms period.
NFreq = the number of FDD carriers to be measured.

Consider the example of the AMR service which has been configured with fixed transport channel starting
positions to allow blind detection of the AMR bit rate. If blind detection is used then TFCI bits are not
required and downlink slot format 8B can be used during SF/2 compressed mode. This slot format has a
spreading factor of 64. Downlink slot format 8B includes 8 pilot bits. If the compressed mode transmission
gap includes 7 slots then actual gap is equal to (7 * 0.667) – (8 * 0.008) = 4.605 ms. 3GPP TS25.133
specifies that a 0.5 ms implementation margin should be introduced at each end of the transmission gap. This
reduces the gap duration to 3.605 ms. 3GPP TS25.133 also specifies that subsequent to the inclusion of the
implementation margin only full slots should be included in the calculation. This reduces the duration to 5
slots (1 slot discounted from the start of the transmission gap and 1 slot discounted from the end of the
transmission gap). If a TGPL of 4 frames is assumed then there are 12 gaps within the 480 ms period. The
time permitted to complete 6 inter-frequency CPICH measurements then becomes:

 480ms 
Tmeasurement _ int er = MAX 480ms ,50ms × × 1 = 600ms
 12 × 3.333 

Table 43 presents the results of the same calculation for a range of TGL and TGPL
TGL TGPL Time to measure 6 inter-
(slots) (frames) frequency neighbours
7 4 600 ms
10 4 480 ms
14 4 480 ms
7 8 1200 ms
10 8 750 ms
14 8 500 ms

Table 43 – TGL impact upon FDD inter-frequency measurement (AMR speech service)
Longer TGL and shorter TGPL result in more rapid inter-frequency measurement times. The drawback of
using long TGL and short TGPL is the associated impact upon layer 1 and layer 2 performance.

Table 44 presents the preferred GSM BSIC identification TGL as a function of the service and compressed
mode method. These are independent of the compressed mode triggering mechanism and the rate of change
of propagation conditions.

Uplink Downlink
Service HLS SF/2 Puncturing HLS SF/2
AMR Speech 7 7 7 Not applicable 7

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 33 of 52

RT Data Not applicable 7 Not applicable Not applicable 7


NRT Data 10 10 Not applicable 10 10

Table 44 – Preferred TGL for FDD inter-frequency measurements


A TGL of 7 slots has been selected for the AMR speech and RT data services. This value has been selected
on the basis of balancing the efficiency with which inter-frequency measurements can be completed with the
impact upon inner loop power control. A TGL of 7 slots can be generated relatively comfortably by each
compressed mode method and may use either the single or double frame approach. A TGL of 10 slots has
been selected for NRT services. This value has been selected on the basis that NRT services are less sensitive
to L1 and L2 performance. NRT services can make use of layer 2 re-transmissions if the performance of
inner loop power becomes poor. NRT services are also able to incur a decreased layer 2 throughput if in the
case of HLS, a 10 slot transmission gap requires a lower transport format combination than a 7 slot
transmission gap.

3GPP TS25.331 and TS25.433 allow the configuration of TGL1 and TGL2, i.e. having two transmission
gaps per transmission gap pattern separated by TGD. This increases the flexibility with which transmission
gap patterns can be designed. This additional flexibility is not required at this point in time and it is
recommended that the design includes a single transmission gap per transmission gap pattern. This means
that the TGD should be configured as 0 slots and that TGL1 = TGL and TGL2 does not require
configuration.

The TGPL provides a trade-off between the time spent in compressed mode and the potential impact upon
layer 1 and layer 2 performance. Long TGPL increase the time spent in compressed mode. This means that
compressed mode must be triggered relatively early to prevent radio link failure occurring prior to
completing a successful handover. Triggering compressed mode relatively early means that compressed
mode will also be triggered more frequently. The TGPL should be defined such that compressed mode can
be triggered relatively late and less frequently. The benefit of using a long TGPL is that inner loop power
control has more time to recover between transmission gaps. Throughput reductions caused by higher layer
scheduling and L2 re-transmissions will also be less frequent and thus will have lower average impact. The
TGPL should also account for the service specific transmission time interval (TTI). In general, the TGPL
should not be defined such that multiple transmission gaps occur within the same TTI. The maximum TTI
specified by 3GPP within TS25.331 is 80 ms. The maximum TTI used by Nokia’s RAN1.5.2.ED2
implementation is 40 ms. As long as the TGPL is greater than 4 radio frames then multiple transmission gaps
will never occur within the same TTI (based upon the assumption of a single transmission gap per
transmission gap pattern).

Table 45 presents the preferred TGPL as a function of the service, compressed mode measurement purpose
and rate of change of propagation conditions. These are independent of the compressed mode triggering
mechanism.

High rate of change of propagation Low rate of change of propagation


conditions conditions
Service GSM RSSI GSM BSIC Inter-Freq. GSM RSSI GSM BSIC Inter-Freq.
AMR Speech 4 4 4 6 6 6
RT Data 4 4 4 6 6 6
NRT Data 4 4 4 6 6 6

Table 45 – Preferred TGPL


A relatively short TGPL has been selected on the basis of minimising the time spent in compressed mode
and thus allowing compressed mode to be triggered relatively late and less frequently. A TGPL of 4 has been
selected for scenarios associated with a high rate of change of propagation conditions. A TGPL of 6 has been
selected for scenarios associated with a low rate of changed of propagation conditions. The low rate of
change of propagation conditions means that more time is available to complete the compressed mode

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 34 of 52

measurements (assuming that the same inter-system and inter-frequency handover triggering threshold is
used for both high and low rate of change of propagation conditions).

3GPP TS25.331 and TS25.433 allow the configuration of TGPL1 and TGPL2. This increases the flexibility
with which transmission gap patterns can be designed. This additional flexibility is not required at this point
in time and it is recommended that TGPL1 = TGPL2 = TGPL. Configuring TGPL1 and TGPL2 in this way
simplifies the design and any subsequent field trial measurement analysis.

The TGSN is dictated to some extent by the TGL. 3GPP TS25.212 specifies that at least 8 slots must be
transmitted within each 15 slot radio frame. This means that if the TGL is greater than 7 then the TGSN must
be configured to ensure that the transmission gap spans two radio frames. When the transmission gap spans
two radio frames then it is preferable that the gap is equally distributed between those frames. This means
that the DPCCH and DPDCH belonging to each of the two frames has an equal allocation of bits.

Table 46 presents the preference in terms of adopting the single or double frame approach. The preference is
defined as a function of the TGL and compressed mode method. The preference is independent of the
service, compressed mode triggering mechanism and the rate of change of propagation conditions.

Uplink Downlink
TGL HLS SF/2 Puncturing HLS SF/2
7 Double Frame Single Frame Double Frame Double Frame Single Frame
10 Double frame Double frame Double frame Double frame Double frame
14 Double frame Double frame Double frame Double frame Double frame

Table 46 – Preferred single or double frame approach


The double frame approach is required for any TGL greater than 7 slots. The double frame approach is
preferred for a TGL of 7 slots when the HLS or puncturing compressed mode methods are used. These
compressed mode methods are completed on a per TTI basis and it is preferable for the transmission gap to
span two frames belonging to separate TTI. The single frame approach is preferred for a TGL of 7 slots
when the SF/2 compressed mode method is used. This is because the SF/2 method is applied on a per frame
basis and adopting the single frame approach limits the spreading factor reduction to a single frame.

Table 47 presents the preference for TGSN based upon the single and double frame approach preferences
specified in Table 46.

TGL TGSN TGSN


(single frame approach) (double frame approach)
7 4 11
10 Not applicable 10
14 Not applicable 8

Table 47 – Preferred TGSN


The TGSN has been selected such that the transmission gap is positioned at the center of the radio frame
when the single frame approach is adopted. This provides a small improvement in layer 1 interleaving
performance. The TGSN has been selected such that the transmission gap is equally distributed between two
frames when the double frame approach is adopted.

The TGPRC should be sufficiently large to allow the UE to complete its measurements. If it is known that a
UE can complete its measurements within a specific period of time then the TGPRC can be used to limit the
time spent in compressed mode. Otherwise the TGPRC can be disabled to allow the UE to remain in
compressed mode until instructed otherwise by the network. At this point in time there is insufficient field
experience to configure a refined value of the TGPRC. The current preference is to disable the TGPRC.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 35 of 52

The importance of the TGCFN depends upon whether the single or double frame approach has been adopted.
If the single frame approach has been adopted the TGCFN has little impact upon performance. In this case
the transmission gap is always limited to a single frame and a single TTI. If the double frame approach has
been adopted then it is preferable to synchronise the transmission gap with the TTI such that the transmission
gap spans two frames belonging to different TTI rather than two frames belonging to the same TTI. In the
case of HLS this may allow a higher transport format combination to be selected by the MAC layer. In the
case of puncturing this makes it less likely that the maximum permissible level of puncturing is exceeded.
Synchronising the transmission gap with the TTI for the SF/2 compressed mode method has little impact
because the spreading factor reduction is applied on a per frame basis rather than a per TTI basis.

The value of ‘N Identify Abort’ should be sufficiently large to allow the UE to successfully decode an
unknown BSIC. If it is known that a UE can complete BSIC identification within a specific period of time
then ‘N Identify Abort’ can be used to limit the time spent in compressed mode. Otherwise ‘N Identify
Abort’ should be configured with a relatively large value. At this point in time there is insufficient field
experience to configure a refined value of the ‘N Identify Abort’. The current preference is to configure ‘N
Identify Abort’ with its maximum allowed value of 128 transmission gap patterns.

At this point in time, the value of ‘T Reconfirm Abort’ is not important because the Nokia RNC1.5.2.ED2
implementation does not support BSIC re-confirmation. Once BSIC re-confirmation is supported then ‘T
Reconfirm Abort’ should be sufficiently large to allow the UE to successfully re-confirm a BSIC. If it is
known that a UE can complete re-confirmation within a specific period of time then ‘T Reconfirm Abort’
can be used to limit the time spent in compressed mode. Otherwise ‘T Reconfirm Abort’ should be
configured with a relatively large value. At this point in time there is insufficient field experience to
configure a refined value of the ‘T Reconfirm Abort’. The current preference is to configure ‘T Reconfirm
Abort’ with its maximum allowed value of 10 s.

5.3. Direction

The requirement for uplink and downlink compressed mode depends upon the UE receiver implementation.
The RNC is informed of the relevant aspects of the UE receiver implementation in the UE capability
information included within the RRC: RRC CONNECTION SETUP COMPLETE message. An example of
this message is presented in Appendix A. The RNC should configure compressed mode according to the
information included within this message. If a UE has a single receiver with a fixed duplex spacing then it
will require both uplink and downlink compressed mode for inter-frequency measurements. If a UE has a
dual receiver then downlink compressed mode is not required and the requirement for uplink compressed
mode depends upon the frequency separation of the carrier to be measured and the carrier of the existing
uplink transmission. The existing uplink transmission is likely to interfere with any measurements if the
frequency separation is not sufficiently large.

The compressed mode design decision regarding direction, is that the RNC should configure compressed
mode in the directions indicated by the UE within its capability information. This design decision is
independent of the scenarios identified in Section 3.

5.4. Power Control

Section 4.4 introduced the parameter set associated with inner loop power control behaviour during
compressed mode. This parameter set is summarised below:

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 36 of 52

 Uplink DeltaSIR1  Downlink DeltaSIRafter1


 Uplink DeltaSIRafter1  Downlink DeltaSIR2
 Uplink DeltaSIR2  Downlink DeltaSIRafter2
 Uplink DeltaSIRafter2  ITP
 Downlink DeltaSIR1  RPP

The design decision to configure a single transmission gap per transmission gap pattern means that the
deltaSIR2 nor deltaSIRafter2 parameters are not applicable. Configuring the deltaSIR1 parameters with
relatively high values leads to increased transmit powers but also increases the ability of inner loop power
control to recover subsequent to a transmission gap. Outer loop power control is able to reduce the SIR target
back to its normal operating point subsequent to a transmission gap increase. Field experience evaluating the
impact of these SIR target increases has not been identified. It may be beneficial to use a larger SIR target
increase for scenarios associated with a high rate of change of propagation conditions. In these cases there
will be less correlation between the propagation conditions before and after a transmission gap.

The compressed mode design decision regarding the SIR target increases, is that the Nokia default
configuration should be used, i.e. 1.5 dB for uplink and downlink deltaSIR1; 0.7 dB for deltaSIRafter1 if the
transmission gap is not included in the frame; 1.5 dB for deltaSIRafter1 if the transmission gap is included in
the frame. This design decision is independent of the scenarios identified in Section 3. Once the quantity of
field experience has increased there may be a requirement to refine these values.

The ITP parameter provides a means to control the UE transmit power immediately after a transmission gap.
TS25.214 specifies that a UE should apply a change in the transmit power of its DPCCH at the start of the
first slot after an uplink or downlink transmission gap. The size of this change is given by:

∆ DPCCH = ∆ Re sume + ∆ Pilot

Where,
∆ Re sume = ∆ TPC × TPC _ cmd gap if ITP = 0
∆ Re sume = δ last if ITP = 1
∆ Pilot = 10 × LOG( N Pilot _ Pr ev / N Pilot _ Curr )
δ last = δ i = 0.9375 × δ i −1 − 0.96875 × ∆ TPC × TPC _ cmd i × k sc

In the case of an uplink transmission gap, TPC_cmdgap is the value of the TPC_cmd derived in the first slot
of the uplink transmission gap, if a downlink TPC command is transmitted in that slot. Otherwise
TPC_cmdgap is equal to zero. ∆TPC is the power control step size. Npilot_Prev is the number of pilot bits in the
most recently transmitted slot and Npilot_Curr is the number of pilot bits in the current slot. TPC_cmdi is the
power control command derived from the current downlink slot. ksc is equal to zero if power scaling has been
applied in the current and previous slot as a result of the UE reaching its maximum power. Otherwise ksc is
equal to 1.

Field experience evaluating the impact of the ITP parameter has not been identified. The compressed mode
design decision regarding ITP, is that it should be configured with a value of 0. This design decision is
independent of the scenarios identified in Section 3. Using a value of 0 is a simpler approach and will
simplify the interpretation of field measurement results. Once the quantity of field experience has increased
there may be a requirement to change the value of ITP to 1.

The RPP parameter provides a means to control the behaviour of inner loop power control during the
recovery period. The recovery period is defined as the minimum of the TGL and 7 slots. The recovery period

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 37 of 52

starts after a transmission gap once there is simultaneous uplink and downlink transmission. Configuring
RPP as 0 means that inner loop power control behaves as normal during the recovery period. Configuring
RPP as 1 means that uplink power control algorithm 1 is used during the recovery period with a modified
power control step size. If uplink power control algorithm 1 is being used, the modified step size is given by
the minimum of 3 dB and twice the normal step size. If uplink power control algorithm 2 is being used then
the modified step size is 1 dB.

Field experience evaluating the impact of the RPP parameter has not been identified. The compressed mode
design decision regarding RPP, is that it should be configured with a value of 0. This design decision is
independent of the scenarios identified in Section 3. Using a value of 0 is a simpler approach and will
simplify the interpretation of field measurement results. Once the quantity of field experience has increased
there may be a requirement to change the value of RPP to 1.

5.5. Scrambling Code

SF/2 in the downlink direction may require the use of an alternative scrambling code. There is a one-to-one
relationship between the original channelisation code and the channelisation code with half the spreading
factor. If the channelisation code with half the spreading factor is blocked by another connection then it must
be applied with an alternative scrambling code. The use of an alternative scrambling code means that
downlink orthogonality is lost for that radio link and downlink transmit powers for all radio links will
increase. The trade-off associated with the use of an alternative scrambling code is thus to allow the use of an
alternative scrambling code and incur short term increases in downlink transmit power or to forbid the use of
an alternative scrambling code and deny compressed mode requests when the required SF/2 channelisation
code is not available. The probability of the required SF/2 channelisation code not being available is
dependent upon the traffic loading the cell, the existing spreading factor and the effectiveness of the routines
used to de-fragment the channelisation code tree. As the level of traffic increases there will be more frequent
instances of the required SF/2 channelisation code not being available. This is especially true for high data
rate services which require channelisation codes that block large parts of the code tree.

The compressed mode design decision regarding scrambling code is that the RNC should allow the use of an
alternative scrambling code when the required SF/2 channelisation code is not available from the primary
scrambling code. This design decision is independent of the scenarios identified in Section 3.

5.6. Downlink Frame Type

The RNC is able to select between downlink frame type ‘A’ and downlink frame type ‘B’. These frame types
are illustrated in Figure 8.

Frame type ‘A’ maximises the duration of the transmission gap. The transmission gap duration for frame
type ‘A’ is equal to the duration of the number of transmission gap slots minus the duration of the DPCCH
pilot field. As an example consider a 7 slot transmission gap applied to the AMR speech service using
downlink slot format 8. The duration of 7 slots is 4.67 ms. The duration of the 4 pilot bits associated with
downlink slot format 8 is 0.07 ms. The resulting transmission gap is thus 4.60 ms.

Frame type ‘B’ reduces the duration of the transmission gap in return for an additional uplink power control
command. The reduction in transmission gap length is defined by the duration of the DPDCH data1 field
plus the duration of the DPCCH TPC field. In the case of the AMR speech service using downlink slot
format 8, the transmission gap is reduced from 4.60 ms to 4.47 ms. The benefit of the additional uplink
power control command depends upon the transmission gap length. As the transmission gap length increases

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 38 of 52

the relative benefit of the additional command decreases. The benefit decreases because there is less
correlation between the propagation conditions before and after the transmission gap. The relative reduction
in transmission gap length also decreases as transmission gap length increases.

Based upon the assumption that transmission gaps will always have a duration of at least 7 slots, the
additional TPC command provided by frame type ‘B’ is not expected to provide a significant benefit. The
compressed mode design decision regarding downlink frame type is that the RNC should configure the use
of frame type ‘A’. This design decision is independent of the scenarios identified in Section 3.

5.7. Measurement Reporting

The RNC is able to configure the interval between periodic compressed mode measurement reports. A short
interval increases the responsiveness of the system but also increases the quantity of signalling traffic. While
the level of network traffic is relatively low, this trade-off can be balanced in the direction of making the
system more responsive. The minimum value permitted by 3GPP is 250 ms. The compressed mode design
decision regarding the reporting interval is that an interval of 250 ms should be used. This design decision is
independent of the scenarios identified in Section 3.
The RNC is able to configure the minimum number of measurement reports that are required prior to being
able to make a handover decision. Increasing the number of measurement reports increases the probability of
the UE identifying the best cell but also increases the handover delay. The current compressed mode strategy
is to minimise the handover delay. The design decision regarding the minimum number of measurement
reports is that the RNC should be able to make a handover decision as soon as it receives the first
measurement report. This design decision is independent of the scenarios identified in Section 3.

The RNC is also able to configure the maximum number of measurement reports after which the UE is
instructed to stop measurements. Decreasing the maximum number of measurement reports limits the
probability of the UE identifying the best cell. The design decision regarding the maximum number of
measurement reports is that the RNC should be allow up to 20 measurement reports. This design decision is
independent of the scenarios identified in Section 3.

5.8. Other Configuration Parameters

Nokia’s RAN1.5.2.ED2 implementation allows the RNC to enable or disable compressed mode
measurements on a per RNC basis. This is done using the RNC databuild parameter CMmasterSwitch. It is
not possible to enable or disable compressed mode measurements on a per cell nor per Node B basis. Based
upon the assumption that compressed mode measurements form an essential part of inter-system and inter-
frequency handover, the design decision regarding the CMmasterSwitch parameter is that it should be
configured to enable compressed mode. This design decision is independent of the scenarios identified in
Section 3.

Nokia’s RAN1.5.2.ED2 implementation allows the RNC to limit the number of simultaneous compressed
mode UE. This is done using the RNC databuild parameter MaxNumbUECMcoverHO. At this point in time
there is insufficient field experience to indicate the impact upon system performance when large number of
UE are simultaneously in compressed mode. The design decision regarding the MaxNumbUECMcoverHO
parameter is that it should be configured to allow a maximum of 16 UE to simultaneously use compressed
mode. This figure represents approximately one third of the cell’s capacity when the cell is loaded by speech
users, i.e. the maximum capacity of a cell is expected to be approximately 50 speech users. There may be a
requirement to reduce this figure on a per cell basis once the quantity of field experience has increased. This
design decision is independent of the scenarios identified in Section 3.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 39 of 52

Nokia’s RAN1.5.2.ED2 implementation makes use of the PrxTarget and PrxOffset RNC databuild
parameters for uplink admission control. At this point in time there is little field experience indicating the
effectiveness of uplink admission control. The design decision regarding the PrxTarget parameter is that it
should be configured with its maximum value of 30 dB. This effectively disables uplink admission control
and removes the requirement to configure PrxOffset with a refined value. There may be a requirement to
reduce PrxTarget on a per cell basis once the quantity of field experience has increased. This design decision
is independent of the scenarios identified in Section 3.

Nokia’s RAN1.5.2.ED2 implementation makes use of the PtxTarget and PtxOffset RNC databuild
parameters for downlink admission control purposes. The design decision regarding the PtxTarget and
PtxOffset parameters is that they should be configured with their default values of 40 dBm and 1 dB
respectively. There may be a requirement to increase PtxTarget once the quantity of field experience has
increased. This design decision is independent of the scenarios identified in Section 3.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 40 of 52

6. Impact of using the Nokia Default Parameter Set


This section evaluates the impact of using the Nokia default configuration presented in Section 4 rather than
the O2 preferred configuration presented in Section 5.

6.1. Method

Compressed mode methods are specified separately for the uplink and downlink.

6.1.1. Uplink

Table 3 presents the Nokia default configuration in terms of uplink compressed mode method. The SF/2
method is applied to the AMR speech and RT data services whereas the HLS method is applied to the NRT
data service.

Table 36 presents the O2 preferred configuration in terms of uplink compressed mode method. The SF/2
method is applied to the RT data service whereas the HLS method is applied to the AMR speech and NRT
data services. These preferences are independent of the rate of change of propagation conditions, triggering
mechanism and measurement purpose.

The only difference between the two configurations is the compressed mode method applied to the AMR
speech service. The O2 preferred configuration of HLS is only possible if a lower AMR bit rate is supported
by both the UE and the network. If lower AMR bit rates are not supported, the SF/2 method must be applied.
At this point in time, lower AMR bit rates are not widely supported and it is acceptable to apply SF/2 rather
than HLS as a compressed mode method for the AMR speech service. Use of the SF/2 method means that
the UE must be capable of supporting the short term 3 dB increase in transmit power associated with the
halved spreading factor. The requirement for an increase in the UE transmit power increases the importance
of the UE transmit power compressed mode triggering mechanism. Additional care is required when using
UE that do not support this triggering mechanism. In this case, the threshold associated with the CPICH
RSCP triggering mechanism may be adjusted to reflect the path loss associated with a specific UE transmit
power.

6.1.2. Downlink

Table 10 presents the Nokia default configuration in terms of downlink compressed mode method. The SF/2
method is applied to the AMR speech and RT data services whereas the HLS method is applied to the NRT
data service.

Table 38 presents the O2 preferred configuration in terms of downlink compressed mode method. The
puncturing method is applied to the AMR speech service. The SF/2 method is applied to the RT data service
whereas the HLS method is applied to the NRT data service. These preferences are independent of the rate of
change of propagation conditions, triggering mechanism and measurement purpose.

The only difference between the two configurations is the compressed mode method applied to the AMR
speech service. The Nokia default configuration of using the SF/2 method means that the cell must be
capable of supporting a 3 dB short term increase in transmit power. This requirement is most significant
when the downlink transmit power is responsible for triggering compressed mode. The use of SF/2 also
places a greater requirement upon the downlink channelisation code tree and the potential requirement for

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 41 of 52

the use of an alternative scrambling code. The AMR speech service uses a relatively high downlink
spreading factor of 128 meaning that the channelisation code tree requirement is less than that of a high bit
rate service.

6.2. Timing

Table 19 presents the Nokia default configuration in terms of TGL. The Nokia default configuration applies
a TGL of 7 slots to all compressed mode measurement patterns. Table 40, Table 42 and Table 44 present the
O2 preferred configurations in terms of TGL for GSM RSSI measurements, GSM BSIC identification and
FDD inter-frequency measurements respectively. The O2 preferred configurations specify a TGL of 7 slots
for the AMR speech and RT data services and TGL of 10 slots for the NRT data service. O2 prefer a larger
transmission gap for NRT data services to increase the efficiency with which measurements can be made and
to reduce the time spent in compressed mode. Reducing the time spent in compressed mode means that
compressed mode can be triggered relatively late and less frequently. NRT services are able to incur the
throughput reduction associated with a larger transmission gap. Applying the Nokia default configuration
means that the compressed mode measurements for the NRT data service will have the same efficiency as for
the AMR speech and RT data services. It also means that the time spent in compressed mode should be
approximately the same for all services. It does not necessarily mean that the compressed mode triggering
points should be the same for all services because BSIC confirmation is not required for the inter-system
handover of NRT data services, i.e. less time is spent in compressed mode for the inter-system handover of
NRT data services.

The Nokia default and O2 preferred configurations are aligned in terms of configuring a single transmission
gap per transmission gap pattern, i.e. TGD = 0 and TGL1 = TGL2 = TGL.

Table 19 presents the Nokia default configuration in terms of TGPL. The Nokia default configuration applies
a TGPL of 4 frames to all compressed mode measurement patterns. Table 45 presents the O2 preferred
configuration in terms of TGPL for scenarios with high and low rates of change of propagation conditions.
The O2 preferred configuration specifies a TGPL of 4 frames for scenarios with a high rate of change of
propagation conditions and a TGPL of 6 frames for scenarios with a low rate of change of propagation
conditions. A TGPL of 6 frames was chosen for low rate of change of propagation conditions on the basis
that more time was available to complete the compressed mode measurements and that for these scenarios
the average impact of compressed mode could be reduced by distributing the measurements over a longer
period of time. The assumption that more time is available for measurements assumes that equal triggering
thresholds are applied to both the high and low rate of change of propagation condition scenarios. The
impact of applying a TGPL of 4 frames rather than 6 frames is that compressed mode measurements will be
completed more rapidly but potentially less reliably and with greater average impact.

The Nokia default and O2 preferred configurations are aligned in terms of configuring a single transmission
gap pattern, i.e. TGPL1 = TGLP2 = TGPL.

Table 19 presents the Nokia default configuration in terms of TGSN. The Nokia default configuration
applies a TGSN of slot 4 for the single frame approach and a TGSN of slot 11 for the double frame
approach. These figures are based upon a TGL of 7 slots. Table 47 presents the O2 preferred configuration in
terms of TGSN for TGL of 7, 10 and 14 slots. The O2 preferred configuration is aligned with the Nokia
default configuration when the TGL is 7 slots. Nokia does not have a default configuration for a TGL of 10
and 14 slots. However the basis of the O2 preferred figures is the same as that for the Nokia default figures,
i.e. position the transmission gap at the center of a frame when the single frame approach is used and
distribute the transmission gap equally between two frames when the double frame approach is used.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 42 of 52

Both the Nokia default configuration and the O2 preferred configuration define the TGPRC as 0 transmission
gap patterns. This is interpreted as an indication that the TGPRC should be disabled and that compressed
mode measurements should continue until instructed otherwise by the RNC.

The Nokia default configuration defines the TGCFN independently of the timing of the transmission time
interval (TTI), i.e. the two are not synchronised. This is aligned with the O2 preferred configuration when
using the single frame approach and when using the double frame approach with SF/2. The O2 preferred
configuration for the double frame approach when using puncturing or HLS is to synchronise the TGCFN
with the TTI such that transmission gaps always span two frames belonging to different TTI. The impact of
not synchronising the TGCFN with the TTI for these scenarios means that either more puncturing must be
applied or a lower transport format combination must be selected. The requirement for greater puncturing
will increase the associated SIR requirement and may make the puncturing method impractical. The
requirement to select a lower transport format combination reduces service throughput.

Both the Nokia default configuration and the O2 preferred configuration define ‘N Identify Abort’ as 128
transmission gap patterns. This represents the maximum allowed value.

The Nokia default configuration does not include a value for ‘T Reconfirm Abort’ because the Nokia
RNC1.5.2.ED2 implementation does not support BSIC re-confirmation. The O2 preferred configuration for
‘T Reconfirm Abort’ is 10 s. This represents the maximum allowed value and is likely to reflect the Nokia
default configuration once BSIC re-confirmation is supported.

6.3. Direction

Both the Nokia default and the O2 preferred use of uplink and downlink compressed mode is according to
the capability information provided by the UE. This information is provided within the RRC: RRC
CONNECTION SETUP COMPLETE message. Making use of this information ensures that the use of
uplink and downlink compressed mode is adjusted according to the requirements of specific UE.

6.4. Power Control

Table 24 and Table 26 present the Nokia defaults in terms of uplink and downlink SIR target increases.
These defaults are, a 1.5 dB increase for uplink and downlink deltaSIR1; a 0.7 dB increase for deltaSIRafter1
if the transmission gap is not included in the frame; and a 1.5 dB increase for deltaSIRafter1 if the
transmission gap is included in the frame. Section 5.4 presents the equivalent O2 preferred configuration.
The O2 preferred configuration is aligned with the Nokia default configuration.

Table 28 presents the Nokia defaults for the ITP and RPP parameters. The Nokia default value for ITP is 0
and the Nokia default value for RPP is 0. Section 5.4 presents the equivalent O2 preferred configuration. The
O2 preferred configuration is aligned with the Nokia default configuration.

6.5. Scrambling Code

Both the Nokia default and the O2 preferred compressed mode scrambling code strategy is to allow the use
of an alternative scrambling code when required by the SF/2 method. The alternative scrambling code is non-
orthogonal and increases downlink transmit powers but it also helps to prevent requests for compressed
mode being denied when the appropriate channelisation code is not available.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 43 of 52

6.6. Downlink Frame Type

Both the Nokia default and the O2 preferred downlink frame type is type ‘A’. This maximises the length of
the transmission gap at the expense of a potentially slightly reduced uplink inner loop power control
performance.

6.7. Measurement Reporting

Table 33 presents the Nokia default configuration in terms of inter-system and inter-frequency periodic
measurement reporting interval. The Nokia default configuration defines a 500 ms reporting interval for both
inter-system and inter-frequency measurement reports. Section 5.7 presents the O2 preferred configuration in
terms of periodic measurement reporting interval. The O2 preferred configuration specifies a reporting
interval of 250 ms. Specifying a shorter reporting interval makes the system more responsive but increases
the quantity of signalling traffic.

Table 33 presents the Nokia default configuration in terms of the minimum number of measurement reports
which are required prior to making a handover decision and the maximum number of measurement reports
after which measurements are cancelled. The minimum number of measurement reports is configured as 0
and the maximum as 20. Section 5.7 presents the O2 preferred configuration in terms of these minimum and
maximum number of measurement reports. The O2 preferred configuration is aligned with the Nokia default
configuration.

6.8. Other Configuration Parameters

Table 34 presents the Nokia default configuration for the RNC databuild parameters CMmasterSwitch and
MaxNumbUECMcoverHO. The Nokia default configuration enables compressed mode on a per RNC basis
and allows a maximum of 16 simultaneous compressed mode UE. Section 6.8 presents the O2 preferred
configuration. The O2 preferred configuration is aligned with the Nokia default configuration. It acceptable
for the CMmasterSwitch parameter to be defined on a per RNC basis rather than on a per cell or per Node B
basis because other parameters can be used to effectively disable compressed mode on a per cell or per Node
B basis, e.g. triggering thresholds may be configured such that compressed mode is never triggered. There
may be a requirement to reduce the value of MaxNumbUECMcoverHO once a greater quantity of field
experience has been obtained.

Table 34 presents the Nokia default configuration for the RNC databuild parameters PrxTarget and
PrxOffset. The Nokia default configuration defines PrxTarget as 4 dB and PrxOffset as 1 dB. Section 6.8
presents the O2 preferred configuration. The O2 preferred configuration defines PrxTarget as 30 dB and
PrxOffset as 1 dB. Increasing the value of PrxTarget to 30 dB effectively disables uplink admission control.
Disabling uplink admission control means that increases in the uplink interference floor could become
relatively large and could compromise uplink coverage. At this point in time there is an uncertainty regarding
the effectiveness of uplink admission control. This has led to the decision to disable uplink admission
control. The value of PrxTarget may be reduced as the quantity of field experience increases.

Table 34 presents the Nokia default configuration for the RNC databuild parameters PtxTarget and
PtxOffset. The Nokia default configuration defines PtxTarget as 40 dBm and PtxOffset as 1 dB. Section 6.8
presents the O2 preferred configuration. The O2 preferred configuration is aligned with the Nokia default
configuration.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 44 of 52

7. Field Experience
The majority of this report is based upon theoretical arguments. This section has been included to help
provide an indication of the performance which can be expected from the field. The measurements presented
in this section have been recorded during the second half of 2003, i.e. before the release of RAN1.5.2.ED2.
Compressed mode performance is not expected to be significantly different for RAN1.5.2.ED2. It should
however be recognised that system performance may change as new UE, Node B and RNC software is
introduced. The measurements in this section are organised in terms of the method used to generated the
compressed mode transmission gaps.

7.1. Spreading Factor Division by 2

Spreading factor division by 2 (SF/2) compressed mode has been studied for an inter-system handover of the
AMR speech service. The single frame approach was configured with a 7 slot transmission gap and a 4 frame
transmission gap pattern length. Figure 9 presents the UE transmit power characteristic recorded while using
this compressed mode configuration.

7 slot transmission
gap every 60 slots
UE transmit power

Time

Figure 9 – UE transmit power during SF/2 compressed mode (AMR speech service)
The transmission gaps are clearly visible whereas the increase in UE transmit power resulting from spreading
factor reduction is less clearly visible. The average increase in UE transmit power caused by compressed
mode was evaluated by driving the same route with and without compressed mode active. An average
increase in UE transmit power of 1.3 dB was observed. This result is reasonable when considering that 8
slots out of every 60 slots have their UE transmit power approximately doubled while 7 slots out of every 60
slots have no UE transmit power. The increase in uplink transmit power results from a combination of the
SIR target changes and the recovery period power control rather than directly from the spreading factor
reduction. Changes in the downlink transmit power were not recorded during this set of measurements.

An average compressed mode activation time of 1.0 s was observed. The activation time represents the delay
between the RNC receiving a measurement report which triggers the requirement for compressed mode and
the RNC sending the measurement control which instructs the UE to start compressed mode. The subsequent
delay between the RNC instructing the UE to start compressed mode measurements and the UE providing all
of its GSM RSSI measurements was observed to be 1.3 s. The delay associated with identifying the relevant
GSM BSIC was 2.0 s. The UE was thus in compressed mode for an average of 3.3 s with an initial triggering
delay of 1.0 s.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 45 of 52

7.2. Higher Layer Scheduling

Higher layer scheduling (HLS) compressed mode has been studied for an inter-system handover of the PS
data service. The single frame approach was configured with a 7 slot transmission gap and a 4 frame
transmission gap pattern length. A downlink file transfer was completed during the compressed mode period.
The measurements were repeated for downlink PS data rates of 64 kbps and 128 kbps and 384 kbps. Figure 9
presents the downlink throughput characteristic for the 128 kbps PS data service.

HLS starts

Average = 121 kbps


Throughput

Average = 71 kbps

Time

Figure 10 – Downlink throughput during ½ rate higher layer scheduling (128 kbps PS data service)
The downlink throughput reduction caused by higher layer scheduling is clearly visible. The average
throughput drops from 121 kbps to 71 kbps. The variance of the throughput is significantly greater during the
period of compressed mode. Larger reductions in throughput were observed for the 64 kbps and 384 kbps
downlink data rates, i.e. a reduction from 55 kbps to 15 kbps for the 64 kbps data service and a reduction
from 270 kbps to 30 kbps for the 384 kbps data service. It is not clear why the compressed mode throughput
associated with the 384 kbps data service is less than the compressed mode throughput associated with the
128 kbps data service. The throughput associated with the 384 kbps data service is relatively low prior to
entering compressed mode indicating that some other factors may be influencing the results for this service.

The average increase in the uplink and downlink SIR targets caused by compressed mode was evaluated by
driving the same route with and without compressed mode active. The uplink SIR target was observed to
increase by 1.6 dB and the downlink SIR target by 6.5 dB. The uplink figure is similar to the average
increase in UE transmit power observed for the AMR speech service when applying SF/2 compressed mode.
The downlink figure is considerably greater than expected.

An average compressed mode activation time of 1.4 s was observed. The activation time represents the delay
between the RNC receiving a measurement report which triggers the requirement for compressed mode and
the RNC sending the measurement control which instructs the UE to start compressed mode. The subsequent
delay between the RNC instructing the UE to start compressed mode measurements and the UE providing all
of its GSM RSSI measurements was observed to be 1.2 s. The UE was thus in compressed mode for an
average of 1.2 s with an initial triggering delay of 1.4 s.

7.3. Puncturing

Puncturing compressed mode measurement results were not available for presentation within this report.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 46 of 52

8. Recommendations
The following recommendations have resulted from the work completed.

1. The preferred set of compressed mode parameters should be identified on a per scenario basis.
Compressed mode scenarios should be categorised according to the measurement purpose (GSM
RSSI, BSIC identification, FDD measurement); the service being used (AMR speech, RT data, NRT
data); the triggering mechanism (uplink transmit power, downlink transmit power, CPICH Ec/Io,
CPICH RSCP, uplink quality); and the rate of change of propagation conditions (high, low).

2. The focus should be directed towards coverage reason compressed mode for inter-system handover.
Compressed mode for inter-frequency handover should be introduced once there are plans for Node
B to be configured with multiple carriers. Load and traffic reason compressed mode should be
evaluated prior to RAN05.

3. Nokia’s current compressed mode implementation is relatively simplistic in terms of customising the
compressed mode method to suit the compressed mode scenario, i.e. SF/2 compressed mode is
always used for AMR speech and RT data services whereas HLS is always used for NRT data
services. Consideration should be given to requesting Nokia to implement a more sophisticated set of
rules for associating specific compressed mode methods to specific compressed mode scenarios.

4. Consideration should be given to requesting Nokia to use HLS as a compressed mode method for the
uplink speech service once both the network and UE support lower AMR bit rates. It is acceptable to
use the Nokia default compressed mode methods for other uplink services. The selection of a
specific uplink compressed mode method is currently not configurable by O2.

5. Consideration should be given to requesting Nokia to use puncturing as a compressed mode method
for the downlink speech service once both the network and UE support lower AMR bit rates. It is
acceptable to use the Nokia default compressed mode methods for other downlink services. The
selection of a specific downlink compressed mode method is currently not configurable by O2.

6. Consideration should be given to requesting Nokia to use a transmission gap length (TGL) of 10
slots rather than 7 slots for NRT data services. It is acceptable to use the Nokia default of 7 slots for
other services. The selection of a specific TGL is not currently configurable by O2.

7. It is acceptable to use the Nokia default strategy of configuring single transmission gap patterns with
a single transmission gaps. The additional flexibility provided by configuring a second transmission
gap pattern and a second transmission gap is not viewed as necessary.

8. Consideration should be given to requesting Nokia to define the transmission gap pattern length
(TGPL) parameters on a per cell basis rather than on a per RNC basis. Consideration could then be
given to configuring the use of a transmission gap pattern length (TGPL) of 6 frames rather than 4
frames for scenarios associated with a low rate of change of propagation conditions. It is acceptable
to use the Nokia default of 4 frames for scenarios associated with a high rate of change of
propagation conditions. The selection of a specific TGPL is configurable by O2 but only on a per
RNC basis.

9. Consideration should be given to requesting Nokia to synchronise the transmission gap connection
frame number (TGCFN) with the transmission time interval (TTI) when the compressed mode

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 47 of 52

method is puncturing or HLS and the double frame approach has been adopted. The TGCFN and
TTI should be synchronised such that double frame transmission gaps span two frames belonging to
different TTI.

10. It is acceptable to use the Nokia default strategy of configuring the transmission gap starting slot
number (TGSN) such that the transmission gap is positioned at the center of a frame when the single
frame approach is used and to distribute the transmission gap equally between two frames when the
double frame approach is used. O2 are able to configure the TGSN for the single frame approach but
currently are not able to configure the TGSN for the double frame approach.

11. It is acceptable to use the Nokia default strategy in terms of allowing compressed mode
measurements to continue until instructed otherwise by the RNC, i.e. configuring the transmission
gap pattern repetition count (TGPRC) with a value of 0. The value of TGPRC may be refined once a
greater quantity of field experience has been gained. The selection of a specific TGPRC is not
currently configurable by O2.

12. It is acceptable to use the Nokia default of configuring ‘N Identify Abort’ with its maximum value of
128 patterns. The value of ‘N Identify Abort’ may be refined once a greater quantity of field
experience has been gained. The selection of a specific ‘N Identify Abort’ is not currently
configurable by O2.

13. It is acceptable to use the Nokia default strategy of configuring uplink and downlink compressed
mode according to the capability information provided by the UE. The use of uplink and downlink
compressed mode is currently not configurable by O2.

14. It is acceptable to use the Nokia default parameter values for the behaviour of inner loop power
control during compressed mode. The parameter values may be refined once a greater quantity of
field experience has been gained. The selection of a specific SIR target increase is not currently
configurable by O2. The selection of the initial transmit power increase mode and recovery period
power control mode are configurable by O2.

15. It is acceptable to use the Nokia default of allowing the use of an alternative scrambling code when
the SF/2 compressed mode method is applied and the required channelisation code is not available
from the primary scrambling code. The use of an alternative scrambling code is configurable by O2.

16. It is acceptable to use the Nokia default of configuring compressed mode to make use of downlink
frame type ‘A’. The selection of a specific frame type is not currently configurable by O2.

17. Consideration should be given to requesting Nokia to use a shorter compressed mode periodic
measurement reporting interval, i.e. 250 ms rather than 500 ms. The periodic reporting interval can
be configured by O2 but 500 ms is currently the lowest supported value.

18. It is acceptable to use the Nokia default values for the compressed mode measurement reporting
parameters which define the minimum required and maximum allowed number of measurement
reports. These parameters are configurable by O2.

19. It is acceptable to use the Nokia default for the maximum number of simultaneous compressed mode
UE per cell, i.e. 16. This value may be refined once a greater quantity of field experience has been
gained. The relevant parameter is configurable by O2.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 48 of 52

20. A field measurement campaign should be completed to evaluate the impact of compressed mode
upon the performance of layers 1 and 2. The measurement campaign should focus upon the transmit
power requirement, SIR target, RLC re-transmission rate and layer 2 throughput. The dominant
compressed mode triggering mechanism should also be identified.

21. A field measurement campaign should be completed to evaluate the delay associated with
compressed mode measurements. The measurement campaign should focus upon a range of services,
a range of transmission gap patterns and a range of compressed mode methods.

22. Results from the field measurement campaigns should be used to help define the thresholds that
trigger compressed mode measurements. If the delay is found to be short then the thresholds may be
configured such that compressed mode is triggered relatively late and less frequently. If the delay is
found to be long then the thresholds should be configured such that compressed mode is triggered
relatively early and more frequently.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 49 of 52

9. Conclusions
The following conclusions have resulted from the work completed:

1. Compressed mode scenarios can be categorised according to the measurement purpose (GSM RSSI,
BSIC identification, FDD measurement); the service being used (AMR speech, RT data, NRT data);
the triggering mechanism (uplink transmit power, downlink transmit power, CPICH Ec/Io, CPICH
RSCP, uplink quality); and the rate of change of propagation conditions (high, low).

2. Compressed mode will initially be limited to coverage reason inter-system handover scenarios. Inter-
frequency handover scenarios will become applicable once O2 starts to configure Node B with
multiple carriers. Load and traffic reason handovers will become applicable once RAN05 RNC
software is available.

3. Compressed mode functionality and signalling is standardised by 3GPP. There are however a set of
design decisions required in terms of selecting a compressed mode method and configuring the
associated transmission gap patterns. Nokia’s current implementation of compressed mode is
compliant with the R99 version of 3GPP specifications. Nokia’s implementation makes use of a set
of configurable RNC databuild parameters and a set on non-configurable RNC databuild parameters.

4. This report has identified the trade-offs associated with each compressed mode parameter. The
Nokia default configuration and the O2 preferred configuration have been identified. The impact of
applying the Nokia default configuration rather than the O2 preferred configuration has been
evaluated.

5. A number of configurable parameter changes have been proposed within the recommendations
section of this report. These parameter changes can be made by O2 but should be tested during a
localised parameter trial prior to deploying across a large area. The majority of arguments for
proposed parameter changes are theoretical rather than being based upon field trial experience.

6. A number of non-configurable parameter changes have been proposed within the recommendations
section of this report. These parameter changes cannot be made by O2 and require requests to be
made to Nokia product line for future implementation. Requests for these parameter changes should
be prioritised with other outstanding requests to Nokia product line. The majority of arguments for
proposed parameter changes are theoretical rather than being based upon field trial experience.

7. A relatively limited quantity of field trial experience has been made available for this report. The
quantity of experience will increase throughout 2004 as O2’s WCDMA network performance is
evaluated and refined.

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 50 of 52

10. References

[1] 3GPP TS25.133 ‘Requirements for the support of radio resource management’
[2] 3GPP TS25.211 ‘Physical channels and mapping of transport channels onto physical channels’
[3] 3GPP TS25.212 ‘Multiplexing and channel coding’
[4] 3GPP TS25.213 ‘Spreading and modulation’
[5] 3GPP TS25.214 ‘Physical layer procedures’
[6] 3GPP TS25.215 ‘Physical layer measurements’
[7] 3GPP TS25.331 ‘Radio resource control protocol specification’
[8] 3GPP TS25.433 ‘UTRAN Iub interface NBAP signalling’

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 51 of 52

11. Abbreviations
AMR Adaptive Multi-Rate
BCCH Broadcast Control Channel
BSIC Base Station Identifier Code
CEC Capacity Enhanced Configuration
CM Compressed Mode
CPICH Common Pilot Channel
CS Circuit Switched
DPCCH Dedicated Physical Control Channel
DPCH Dedicated Physical Channel
DPDCH Dedicated Physical Data Channel
FDD Frequency Division Duplex
FMCI Inter-Frequency Measurement Control
FMCG Inter-System Measurement Control
HLS Higher Layer Scheduling
IFHO Inter-Frequency Handover
ISHO Inter-System Handover
ITP Initial Transmit Power method
NBAP Node B Application Protocol
NRT Non-Real Time
OTOR Omni Transmit Omni Receive
OTSR Omni Transmit Sector Receive
PS Packet Switched
RAN Radio Access Network
RLC Radio Link Control
RNC Radio Network Controller
ROC Roll-Out Configuration
RPP Recovery Period Power control mode
RRC Radio Resource Control
RSCP Received Signal Code Power
RSSI Received Signal Strength Indicator
RT Real Time
SF Spreading Factor
SIR Signal to Interference Ratio
STSR Sector Transmit Sector Receive
TGCFN Transmission Gap Connection Frame Number
TGD Transmission Gap Distance
TGL Transmission Gap Length
TGMP Transmission Gap Measurement Purpose
TGPL Transmission Gap Pattern Length
TGPRC Transmission Gap Pattern Repetition Count
TGPSI Transmission Gap Pattern Sequence Identifier
TGSN Transmission Gap Starting slot Number
TPC Transmit Power Control
TTI Transmission Time Interval
UE User Equipment
UMTS Universal Mobile Telecommunications System
UTRAN UMTS Terrestrial Radio Access Network

Version : Issue 1 Date: 22/04/04


In Confidence
In Confidence
O2 'Do Once' Radio Design
Document Reference : O2RD/04/021
Compressed Mode Scenarios and Strategy Page : 52 of 52

Appendix A. Example UE capability message


This appendix provides an example RRC: RRC CONNECTION SETUP COMPLETE message. This
message includes the UE capability information which specifies the UE’s capability in terms of uplink and
downlink compressed mode measurements.
UL-DCCH-Message : { downlinkPhysChCapability {
message rrcConnectionSetupComplete : { maxNoDPCH-PDSCH-Codes 3,
rrc-TransactionIdentifier 0, maxNoPhysChBitsReceived b9600,
startList { supportForSF-512 FALSE,
{ supportOfPDSCH FALSE,
cn-DomainIdentity cs-domain, simultaneousSCCPCH-DPCH-Reception
start-Value '00000000000000011000'B notSupported : NULL
}, },
{ uplinkPhysChCapability {
cn-DomainIdentity ps-domain, maxNoDPDCH-BitsTransmitted b2400,
start-Value '11111111111111111111'B supportOfPCPCH FALSE
} }
}, }
ue-RadioAccessCapability { },
accessStratumReleaseIndicator r99, ue-MultiModeRAT-Capability {
pdcp-Capability { multiRAT-CapabilityList {
losslessSRNS-RelocationSupport FALSE, supportOfGSM TRUE,
supportForRfc2507 notSupported : NULL supportOfMulticarrier FALSE
}, },
rlc-Capability { multiModeCapability fdd
totalRLC-AM-BufferSize kb50, },
maximumRLC-WindowSize mws2047, securityCapability {
maximumAM-EntityNumber am6 cipheringAlgorithmCap { ciphuea0 },
}, integrityProtectionAlgorithmCap { integuia1 }
transportChannelCapability { },
dl-TransChCapability { ue-positioning-Capability {
maxNoBitsReceived b5120, standaloneLocMethodsSupported FALSE,
maxConvCodeBitsReceived b1280, ue-BasedOTDOA-Supported FALSE,
turboDecodingSupport supported: b5120, networkAssistedGPS-Supported noNetworkAssistedGPS,
maxSimultaneousTransChs e8, supportForUE-GPS-TimingOfCellFrames FALSE,
maxSimultaneousCCTrCH-Count 1, supportForIPDL FALSE
maxReceivedTransportBlocks tb16, },
maxNumberOfTFC tfc96, measurementCapability {
maxNumberOfTF tf64 downlinkCompressedMode {
}, fdd-Measurements TRUE,
ul-TransChCapability { gsm-Measurements {
maxNoBitsTransmitted b3840, gsm900 TRUE,
maxConvCodeBitsTransmitted b1280, dcs1800 TRUE,
turboEncodingSupport supported: b3840, gsm1900 FALSE
maxSimultaneousTransChs e8, },
modeSpecificInfo fdd : NULL, multiCarrierMeasurements FALSE
maxTransmittedBlocks tb8, },
maxNumberOfTFC tfc32, uplinkCompressedMode {
maxNumberOfTF tf32 fdd-Measurements TRUE,
} gsm-Measurements {
}, gsm900 TRUE,
rf-Capability { dcs1800 TRUE,
fddRF-Capability { gsm1900 FALSE
ue-PowerClass 4, },
txRxFrequencySeparation mhz190 multiCarrierMeasurements FALSE
} }
}, }
physicalChannelCapability { }
fddPhysChCapability { }

--- End of Document ---

Version : Issue 1 Date: 22/04/04


In Confidence

S-ar putea să vă placă și