3GPP TR 36.922 V13.0.

0 (2016-01)
Technical Report

3rd Generation Partnership Project;

Technical Specification Group Radio Access Network;
Evolved Universal Terrestrial Radio Access (E-UTRA);
TDD Home eNode B (HeNB) Radio Frequency (RF)
requirements analysis
(Release 13)

The present document has been developed within the 3rd Generation Partnership Project (3GPP TM) and may be further elaborated for the purposes of 3GPP.

The present document has not been subject to any approval process by the 3GPP Organizational Partners and shall not be implemented.
This Specification is provided for future development work within 3GPP only. The Organizational Partners accept no liability for any use of this Specification.
Specifications and reports for implementation of the 3GPP TM system should be obtained via the 3GPP Organizational Partners' Publications Offices.
Release 13 2 3GPP TR 36.922 V13.0.0 (2016-01)

Keywords
LTE, Radio

3GPP

Postal address

3GPP support office address

650 Route des Lucioles – Sophia Antipolis
Valbonne – FRANCE
Tel.: +33 4 92 94 42 00 Fax: +33 4 93 65 47 16

Internet
http://www.3gpp.org

Copyright Notification

No part may be reproduced except as authorized by written permission.

The copyright and the foregoing restriction extend to reproduction in all media.

© 2016, 3GPP Organizational Partners (ARIB, ATIS, CCSA, ETSI, TSDSI, TTA, TTC).
All rights reserved.

UMTS™ is a Trade Mark of ETSI registered for the benefit of its members
3GPP™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
LTE™ is a Trade Mark of ETSI registered for the benefit of its Members and of the 3GPP Organizational Partners
GSM® and the GSM logo are registered and owned by the GSM Association

3GPP
Release 13 3 3GPP TR 36.922 V13.0.0 (2016-01)

Contents
Foreword............................................................................................................................................................. 6
1 Scope ........................................................................................................................................................ 7
2 References ................................................................................................................................................ 7
3 Definitions, symbols and abbreviations ................................................................................................... 7
3.1 Definitions ......................................................................................................................................................... 7
3.2 Symbols ............................................................................................................................................................. 7
3.3 Abbreviations ..................................................................................................................................................... 8
4 General ..................................................................................................................................................... 8
5 Radio scenarios ........................................................................................................................................ 8
5.1 Deployment configurations ................................................................................................................................ 8
5.2 Interference scenarios ........................................................................................................................................ 9
6 RF Aspects ............................................................................................................................................. 10
6.1 Transmitter characteristics ............................................................................................................................... 10
6.1.1 HeNB output power ................................................................................................................................... 10
6.1.1.1 HeNB maximum output power............................................................................................................. 10
6.1.1.1.1 Analysis .......................................................................................................................................... 10
6.1.1.1.2 Minimum requirement .................................................................................................................... 11
6.1.1.2 HeNB output power for adjacent UTRA channel protection................................................................ 11
6.1.1.3 HeNB output power for adjacent E-UTRA channel protection ............................................................ 12
6.1.2 Frequency error .......................................................................................................................................... 13
6.1.2.1 Handover performance ......................................................................................................................... 13
6.1.2.2 Cell capacity ......................................................................................................................................... 13
6.1.2.3 Timing .................................................................................................................................................. 16
6.1.2.4 Minimum requirement .......................................................................................................................... 17
6.1.3 Adjacent Channel Leakage power Ratio (ACLR) ...................................................................................... 17
6.1.3.1 Minimum requirement .......................................................................................................................... 17
6.1.4 Operating band unwanted emissions .......................................................................................................... 18
6.1.4.1 Minimum requirements ........................................................................................................................ 19
6.1.5 Spurious emissions ..................................................................................................................................... 21
6.1.5.1 Mandatory requirements ....................................................................................................................... 21
6.1.5.2 Co-existence with HNB/HeNB operating in other bands ..................................................................... 21
6.1.5.2.1 Minimum requirement .................................................................................................................... 21
6.1.6 Transmitter intermodulation ....................................................................................................................... 22
6.1.6.1 Minimum requirement .......................................................................................................................... 23
6.2 Receiver characteristics ................................................................................................................................... 23
6.2.1 Reference sensitivity level ......................................................................................................................... 23
6.2.1.1 Uplink performance degradation of macrocell ..................................................................................... 23
6.2.1.1.1 Simulation setup ............................................................................................................................. 23
6.2.1.1.2 Simulation results............................................................................................................................ 24
6.2.1.2 HeNB desensitization ........................................................................................................................... 25
6.2.1.3 Minimum requirement .......................................................................................................................... 27
6.2.2 Dynamic range ........................................................................................................................................... 27
6.2.2.1 Deterministic analysis........................................................................................................................... 27
6.2.2.2 System-level simulations ...................................................................................................................... 28
6.2.2.3 Minimum requirement .......................................................................................................................... 29
6.2.3 Adjacent channel selectivity (ACS) and narrow-band blocking ................................................................ 29
6.2.3.1 Simulation assumptions ........................................................................................................................ 29
6.2.3.2 Simulation results ................................................................................................................................. 30
6.2.3.3 Minimum requirements ........................................................................................................................ 31
6.2.4 Blocking characteristics ............................................................................................................................. 32
6.2.4.1 General blocking requirement .............................................................................................................. 32
6.2.4.1.1 Minimum requirement .................................................................................................................... 33
6.2.4.2 Co-location with other HNB/HeNB ..................................................................................................... 34

3GPP
Release 13 4 3GPP TR 36.922 V13.0.0 (2016-01)

6.2.5 Receiver Intermodulation ........................................................................................................................... 35

6.2.5.1 Analysis ................................................................................................................................................ 35
6.2.5.2 Minimum requirement .......................................................................................................................... 35
6.2.6 In-channel selectivity ................................................................................................................................. 37
6.2.6.1 Analysis ................................................................................................................................................ 37
6.2.6.2 Minimum requirement .......................................................................................................................... 37
6.3 Performance requirement ................................................................................................................................. 37
6.4 Synchronization requirement ........................................................................................................................... 38
6.4.1 Synchronization Accuracy ......................................................................................................................... 38
6.4.1.1 Synchronization error analysis ............................................................................................................. 38
6.4.1.2 Synchronization requirement ................................................................................................................ 39
6.4.2 Techniques for Synchronization................................................................................................................. 40
6.4.2.1 Synchronization using Network Listening ........................................................................................... 40
6.4.2.1.1 Interference Problems with Network Listening and Solutions........................................................ 40
6.4.2.1.2 MBSFN Subframe based Network Listening.................................................................................. 41
6.4.2.1.3 TDD Special Subframe based Network Listening .......................................................................... 42
6.4.2.1.4 Indication of Stratum Level and Synchronization Status ................................................................ 45
6.4.2.1.5 Scheme Comparison ....................................................................................................................... 48
7 Interference control ................................................................................................................................ 48
7.1 HeNB measurements ....................................................................................................................................... 48
7.1.1 Measurements from all cells ...................................................................................................................... 49
7.1.2 Measurements to identify surrounding cell layers ...................................................................................... 49
7.1.3 Measurements from macro cell layer ......................................................................................................... 49
7.1.4 Measurements of other HeNB cells............................................................................................................ 50
7.2 HeNB self-configuration.................................................................................................................................. 51
7.2.1 Information Exchange between eNBs and HeNBs..................................................................................... 51
7.3 Uplink interference control .............................................................................................................................. 53
7.3.1 Control Channel Protection ........................................................................................................................ 53
7.3.1.1 HeNB Uplink Control Channel Protection ........................................................................................... 53
7.3.1.2 Signalling offset over the backhaul ...................................................................................................... 54
7.3.2 Smart Power Control based on Path Loss to Worst Victim Macro eNodeB .............................................. 54
7.3.2.1 Power Cap Method ............................................................................................................................... 54
7.3.2.1.1 Simulation Assumptions ................................................................................................................. 54
7.3.2.1.2 Simulation Results .......................................................................................................................... 55
7.3.2.1.3 Discussion of Results ...................................................................................................................... 58
7.3.2.2 Power Control based on PL from HUE to its serving HeNB and PL from HUE to its worst victim
MeNB ................................................................................................................................................... 58
7.3.2.2.1 Simulation Assumptions ................................................................................................................. 58
7.3.2.2.2 Simulation Results .......................................................................................................................... 59
7.3.2.2.3 Discussion of Results ...................................................................................................................... 60
7.3.2.3 For Future Releases .............................................................................................................................. 60
7.4 Downlink interference control ......................................................................................................................... 60
7.4.1 Control Channel Protection ........................................................................................................................ 60
7.4.1.1 Control of HeNB downlink interference towards macro eNB control channels by frequency
partitioning with per-subband interference estimation ......................................................................... 61
7.4.1.2 Control of HeNB downlink interference among neighboring HeNBs control channels by
frequency partitioning........................................................................................................................... 62
7.4.2 Data Channel Protection............................................................................................................................. 62
7.4.2.1 Control of HeNB Downlink Interference towards macro eNB data channels by frequency
partition ................................................................................................................................................ 62
7.4.2.2 Control of HeNB Downlink Interference among neighboring HeNBs ................................................. 63
7.4.2.2.1 Centralized coordination ................................................................................................................. 63
7.4.2.2.2 Distributed Dynamic Frequency Partitioning ................................................................................. 64
7.4.2.3 Adaptive Frequency Selection .............................................................................................................. 64
7.4.2.4 Downlink interference management based on mapping between PCIs and transmission patterns ....... 65
7.4.2.5 Control of HeNB Downlink Interference by dynamically changing HeNB CSG ID ........................... 65
7.4.3 Power Control ............................................................................................................................................ 66
7.4.3.1 HeNB power control based on HeNB-MUE path loss ......................................................................... 66
7.4.3.2 Smart power control based on interference measurement from macro BS ........................................... 67
7.5 Hybrid Cells ..................................................................................................................................................... 68
7.5.1 Hybrid Access Level of Service ................................................................................................................. 68

3GPP
Release 13 5 3GPP TR 36.922 V13.0.0 (2016-01)

7.5.2 DL Performance Evaluation ....................................................................................................................... 68

7.5.3 Hybrid Cell RB Resource Management ..................................................................................................... 71
7.5.4 Hybrid Cell Power Management ................................................................................................................ 71

Annex A (informative): Change History .............................................................................................. 73

3GPP
Release 13 6 3GPP TR 36.922 V13.0.0 (2016-01)

Foreword
This Technical Report has been produced by the 3rd Generation Partnership Project (3GPP).

The contents of the present document are subject to continuing work within the TSG and may change following formal
TSG approval. Should the TSG modify the contents of the present document, it will be re-released by the TSG with an
identifying change of release date and an increase in version number as follows:

Version x.y.z

where:

x the first digit:

1 presented to TSG for information;

2 presented to TSG for approval;

3 or greater indicates TSG approved document under change control.

Y the second digit is incremented for all changes of substance, i.e. technical enhancements, corrections, updates,
etc.

z the third digit is incremented when editorial only changes have been incorporated in the document.

3GPP
Release 13 7 3GPP TR 36.922 V13.0.0 (2016-01)

1 Scope
The present document is the technical report for the work item on LTE TDD HeNB RF requirements, which was
approved at TSG RAN#43. The objective of the WI is to first identify the relevant scenarios and then write an RF
requirements specification that is applicable to LTE TDD HeNB.

2 References
The following documents contain provisions which, through reference in this text, constitute provisions of the present
document.

• References are either specific (identified by date of publication, edition number, version number, etc.) or
non-specific.

• For a specific reference, subsequent revisions do not apply.

• For a non-specific reference, the latest version applies. In the case of a reference to a 3GPP document (including
a GSM document), a non-specific reference implicitly refers to the latest version of that document in the same
Release as the present document.

[1] 3GPP TR 21.905: "Vocabulary for 3GPP Specifications".

[2] 3GPP TR 41.001: "GSM Release specifications".

[3] 3GPP TR 21.912: "Example 2, using fixed text".

[4] 3GPP TS36.211, "Physical Channels and Modulation".

[5] 3GPP TS36.104, "Base Station radio transmission and reception".

[6] 3GPP TS36.101, "User Equipment (UE) transmission and reception".

[7] 3GPP TR 25 967 : "FDD Home Node B (HNB) RF Requirements".

[8] 3GPP TR36.804, "Base Station (BS) radio transmission and reception”.

[9] 3GPP TR36.942, "Radio Frequency (RF) system scenarios".

[10] 3GPP TR25.951, "FDD Base Station (BS) classification".

[11] ITU-R Recommendation SM.329: "Unwanted emissions in the spurious domain".

[12] ITU-R M.1225, Guidelines for evaluation of radio interface technologies for IMT-Advanced

[13] 3GPP TS 36.213: "Physical layer procedures".

3 Definitions, symbols and abbreviations

3.1 Definitions
(Void)

3.2 Symbols
(Void)

3GPP
Release 13 8 3GPP TR 36.922 V13.0.0 (2016-01)

3.3 Abbreviations
For the purposes of the present document, the abbreviations given in TR 21.905 [1] and the following apply. An
abbreviation defined in the present document takes precedence over the definition of the same abbreviation, if any, in
TR 21.905 [1].

HeNB Home Enhanced Node B

HNB Home NodeB
CSG Closed Subscriber Group
GPS Global Positioning System

4 General
This work item has the following objectives:

1. Specify the RF requirements for the E-UTRA TDD Home eNodeB in TS36.104 and the corresponding updates
on the test specification in TS36.141. Some requirements could refer to the outcome of existing/ongoing related
studies.

2. Investigate and find out effective interference control schemes to ensure good performance of both macro layer
and HeNB. Although some of the studies could refer to UTRA HNB related work experience, e.g.
deployment/interference scenarios, amount of studies are needed to find out the effective interference control
schemes due to different physical techniques and system characters between E-UTRA and UTRA. The work
should include but not be limited to the followings,

• The operator has the means to obtain interference control related measurements reports from HeNB and/or
HUE, e.g. the strength of signals and the identity from the macro cell layer and from other HeNBs.

• The operator has the means to set the maximum output power and/or frequency of HeNB. This is expected to
introduce changes to TS36.104.

• The operator has the means to coordinate the HeNB and eNB timing and TDD configuration. This is
expected to introduce changes to TS36.104.

• The operator has guidance on how to control HeNB power and expected performance levels in the relevant
scenarios.

The scope of this work item is limited to the LTE TDD mode.

5 Radio scenarios

5.1 Deployment configurations

In TR 25.967 [7], a number of different deployment configurations have been considered for FDD Home NodeB,
including:

• Open access or CSG (Closed Subscriber Group)

• Dedicated channel or co-channel

• Fixed or adaptive (DL) maximum transmit power

For FDD or TDD Home eNodeB, the following deployment configuration should be considered in addition to the ones
listed above:

• Fixed or adaptive resource partitioning

Specifically, the resource partitioning could be performed in frequency, time or spatial dimensions for interference
coordination.

3GPP
Release 13 9 3GPP TR 36.922 V13.0.0 (2016-01)

Frequency partitioning
Most existing LTE ICIC mechanisms belong to this category, e.g., FFR and SFR. Frequency partitioning can be
combined with power control to achieve better performance. Different from the partial co-channel configuration for
HNB [7], frequency partitioning can be performed at the granularities of RBs within a carrier, as shown in Figure 5.1-1,
which enables more flexible coordination not only between Macro and Home eNodeB, but also between the Home
eNodeBs. For the frequency partitioning method, the Adjacent Channel Power Leakage (ACPL) problem should be
taken into account in performance evaluation, similar to the dedicated channel configuration. If adaptive frequency
partitioning is used, possible information exchanges between Home eNodeBs may need to be supported.

Power

Macro
eNodeB

(a) (b) (c) Frequency

Power

Home
eNodeB 1

(a) (b) (c) Frequency

Power

Home
eNodeB 2

Frequency
(a) (b) (c)

Figure 5.1-1 Frequency Partitioning

Time partitioning
The resources used in Macro and Home eNodeBs can also be partitioned and coordinated in the time dimension.
Different time zone or UL-DL configurations between HeNBs and macro eNBs or among HeNBs under specific
conditions may provide some flexibility for interference coordination. However, it may also bring new interference
risks. Further interference mitigation method based on the time partitioning is FFS.

Spatial partitioning
Due to uplink-downlink channel reciprocity, TDD HeNBs can use beam coordination to improve interference
conditions. For example, the HeNB can avoid beam collision with the Macro or other Home eNBs in a proactive or
reactive way. These mechanisms may require a certain amount of information exchange between the HeNBs.

5.2 Interference scenarios

Table 5.2-1 and Figure 5.2-1 show the possible HeNB related interference scenarios. The listed interference scenarios
are the same for both TDD and FDD. The main difference may exist in how to model the interference, especially for
some control channels that are always present, e.g., BCH, SCH. For both TDD and FDD, we propose to evaluate the
control interference based on the assumption that different base stations are synchronized to ensure the system
performance even under the worst circumstance.

3GPP
Release 13 10 3GPP TR 36.922 V13.0.0 (2016-01)

Table 5.2-1 Interference scenarios

Number Aggressor Victim Priority

1 UE attached to Home eNode Macro eNode B Uplink Yes
B
2 Home eNode B Macro eNode B Downlink Yes
3 UE attached to Macro eNode Home eNode B Uplink Yes
B
4 Macro eNode B Home eNode B Downlink
5 UE attached to Home eNode Home eNode B Uplink Yes
B
6 Home eNode B Home eNode B Downlink Yes
7 UE attached to Home eNode Other System
B and/or Home eNode B

8 Other System UE attached to Home

eNode B and/or Home
eNode B

Figure 5.2-1 Interference Scenarios

6 RF Aspects

6.1 Transmitter characteristics

6.1.1 HeNB output power

6.1.1.1 HeNB maximum output power

6.1.1.1.1 Analysis
From HeNB coverage and capacity point of view, it will be beneficial to define relatively larger output power. However,
as already been demonstrated by numerous contributions, the maximum output power should be limited in order to
control the HeNB->MeNB downlink interference. So, the maximum HeNB output power should be a trade-off between
the HeNB performance and the interference towards close-by MeNB users, which do not have access to the HeNB.
Based on some studies in RAN4, the simulation results show that in some cases the HeNB power can be set up to
20dBm. While in some other cases, e.g. in the case of shared carrier deployment, the HeNB power should be limited to
a relatively low level.

Furthermore, the definition of the total HeNB output power should also consider supporting existing E-UTRAN UEs.
The interfering power level for HeNB ACS requirement is defined -28dBm at 1% blocking probability, which means
the MCL between HeNB and UE is 45dB based on the 23dBm UE maximum power. For UE, the current maximum
tolerable interfering level for ACS is -25dBm. Assuming the total HeNB output power (i.e. the sum over all transmit
antennas) equal to PheNB, the following formula should be true,

PheNB+25<MCL=45dB.

3GPP
Release 13 11 3GPP TR 36.922 V13.0.0 (2016-01)

Seen from the above, HeNB total transmission power of ~20dBm is also applicable from link balance point of view.

So, it is proposed to set the HeNB total maximum output power requirement as 20dBm and the maximum power per
antenna depending on different antenna configurations.

6.1.1.1.2 Minimum requirement

Maximum output power, Pmax, of the base station is the mean power level per carrier measured at the antenna
connector in specified reference condition. The rated output power, PRAT, of the BS shall be as following,

< 20dBm -10*log10 (N)

Where, N is the number of transmitter antenna. N = 1, 2 and 4.

In normal conditions, the base station maximum output power shall remain within +2 dB and -2 dB of the rated output
power declared by the manufacturer.

In extreme conditions, the base station maximum output power shall remain within +2.5 dB and -2.5 dB of the rated
output power declared by the manufacturer.

6.1.1.2 HeNB output power for adjacent UTRA channel protection

The Home BS shall be capable of adjusting the transmitter output power to minimize the interference level on the
adjacent channels licensed to other operators in the same geographical area while optimize the Home BS coverage.
These requirements are only applicable to Home BS. The requirements in this clause are applicable for AWGN radio
propagation conditions.

The output power, Pout, of the Home BS shall be as specified in Table 6.1.1.2-1 under the following input conditions:

- CPICH Êc, measured in dBm, is the code power of the Primary CPICH on one of the adjacent channels present
at the Home BS antenna connector for the CPICH received on the adjacent channels. If Tx diversity is applied on
the Primary CPICH, CPICH Êc shall be the sum in W of the code powers of the Primary CPICH transmitted
from each antenna.

- Ioh, measured in dBm, is the total received power density, including signals and interference but excluding the
own Home BS signal, present at the Home BS antenna connector on the Home BS operating channel.

In case that both adjacent channels are licensed to other operators, the most stringent requirement shall apply for Pout.
In the case when one of the adjacent channels is licensed to a E-UTRA operator while the other adjacent channel is
licensed to a UTRA operator, the more stringent requirement of this subclause and subclause 6.1.1.3 shall apply for
Pout. In case the Home BS's operating channel and both adjacent channels are licensed to the same operator, the
requirements of this clause do not apply.

The input conditions defined for the requirements in this section are specified at the antenna connector of the Home BS.
For Home BS receivers with diversity, the requirements apply to each antenna connector separately, with the other
one(s) terminated or disabled. The requirements are otherwise unchanged. For Home BS(s) without measurement
capability, a reference antenna with a gain of 0 dBi is assumed for converting these power levels into field strength
requirements.

Table 6.1.1.2-1: Home BS output power for adjacent operator UTRA channel protection

Input Conditions Output power, Pout

Ioh > CPICH Êc + 43 dB ≤ 10 dBm

And CPICH Êc ≥ -
105dBm
Ioh ≤ CPICH Êc + 43 dB ≤ max(8 dBm, min(20 dBm,
and CPICH Êc ≥ - CPICH Êc + 100
105dBm dB))

3GPP
Release 13 12 3GPP TR 36.922 V13.0.0 (2016-01)

Note 1: The Home BS transmitter output power specified in Table 6.1.1.2-1 assumes a Home BS reference
antenna gain of 0 dBi, an target outage zone of 47dB around the Home BS for an UE on the adjacent
channel, with an allowance of 2 dB for measurement errors, an ACIR of 33 dB, an adjacent channel UE
CPICH Ec/Io target of -18 dB and the same CPICH Êc value at the adjacent channel UE as for the Home
BS.

Note 2: For CPICH Êc < -105dBm, the requirements in subclauses 6.1.1.1 apply.

Note 3: The output power Pout is the sum transmit power across all the antennas of the Home BS, with each
transmit power measured at the respective antenna connectors.

6.1.1.3 HeNB output power for adjacent E-UTRA channel protection

The Home BS shall be capable of adjusting the transmitter output power to minimize the interference level on the
adjacent channels licensed to other operators in the same geographical area while optimize the Home BS coverage.
These requirements are only applicable to Home BS. The requirements in this clause are applicable for AWGN radio
propagation conditions.

The output power, Pout, of the Home BS shall be as specified in Table 6.1.1.3-1 under the following input conditions:

- CRS Êc, measured in dBm, is the Reference Signal Received Power per resource element on one of the adjacent
channels present at the Home BS antenna connector for the Reference Signal received on the adjacent channels.
For CRS Êc determination, the cell-specific reference signal R0 according TS 36.211 [4] shall be used. If the
Home BS can reliably detect that multiple TX antennas are used for transmission on the adjacent channel, it may
use the average in W of the CRS Êc on all detected antennas.

- Ioh, measured in dBm, is the total received power density, including signals and interference but excluding the
own Home BS signal, present at the Home BS antenna connector on the Home BS operating channel.

In case that both adjacent channels are licensed to other operators, the most stringent requirement shall apply for Pout.
In the case when one of the adjacent channels is licensed to a E-UTRA operator while the other adjacent channel is
licensed to a UTRA operator, the more stringent requirement of this subclause and subclause 6.1.1.3 shall apply for
Pout. In case the Home BS's operating channel and both adjacent channels are licensed to the same operator, the
requirements of this clause do not apply.

The input conditions defined for the requirements in this section are specified at the antenna connector of the Home BS.
For Home BS receivers with diversity, the requirements apply to each antenna connector separately, with the other
one(s) terminated or disabled .The requirements are otherwise unchanged. For Home BS(s) without measurement
capability, a reference antenna with a gain of 0 dBi is assumed for converting these power levels into field strength
requirements.

Table 6.1.1.3-1: Home BS output power for adjacent operator E-UTRA channel protection

Input Conditions Output power, Pout

Ioh > CRS Êc + ≤ 10 dBm

(
10 ⋅ log10 N DL
RB ⋅N RB
sc )
+ 30 dB
and CRS Êc ≥ -127dBm
Ioh ≤ CRS Êc + ≤ max(8 dBm, min(20 dBm,
(
10 ⋅ log10 N RB
DL
⋅ N scRB ) CRS Êc +

+ 30 dB (
10 ⋅ log10 N RB
DL
⋅ N scRB )
and CRS Êc ≥ -127dBm + 85 dB))

Note 1: The Home BS transmitter output power specified in Table 6.1.1.3-1 assumes a Home BS reference
antenna gain of 0 dBi, an target outage zone of 47dB around the Home BS for an UE on the adjacent
channel, with an allowance of 2 dB for measurement errors, an ACIR of 30 dB, an adjacent channel UE
Ês/Iot target of -6 dB and the same CRS Êc value at the adjacent channel UE as for the Home BS.

Note 2: For CRS Êc < -127dBm, the requirements in subclauses 6.1.1.1 apply.

3GPP
Release 13 13 3GPP TR 36.922 V13.0.0 (2016-01)

Note 3: The output power Pout is the sum transmit power across all the antennas of the Home BS, with each
transmit power measured at the respective antenna connectors.
DL
Note 4: N RB is the number of downlink resource blocks in the own Home BS channel.

Note 5: N scRB is the number of subcarriers in a resource block, N scRB = 12 .

6.1.2 Frequency error

Frequency error is the difference between the actual BS transmit frequency and the assigned frequency. The same
source shall be used for RF frequency and data clock generation. Frequencies accuracy is an important RF requirement
for HeNB. A reasonable tradeoff between the cost and system performance should be made to derive the frequency
error for HeNB. Frequency accuracy will affect the system performance in many areas, such as handover performance,
cell throughput and timing etc.

6.1.2.1 Handover performance

Frequency accuracy requirement will affect the measurement precision, which may degrade the handover performance.
For HeNB, there are three kinds of handover scenarios, (1) handover from eNB to HeNB, (2) handover from HeNB to
eNB, (3) handover between HeNBs. Since eNB usually has a better frequency accuracy performance than HeNB, the
handover between HeNBs is the worst scenario. Although the handover scenarios of HeNB have not been defined
clearly in 3GPP, the frequency error should support all possible scenarios.

f1 f0 f0 f2

f3 f1 f0 f0 f2 f4
v

f0: standard carrier frequency;

f1: f0 – maximum frequency error;
f2: f0 + maximum frequency error;
f3: f1 – maximum Doppler shift;
f4: f2 + maximum Doppler shift;

Figure 6.1.2.1-1 Handover between HeNB

For macro cells, eNB can support the handover at the speed of 350km/h. For home environment, the maximum speed of
UE is only 30km/h. Since the frequency error of ±0.05ppm and speed of 350km/h can be supported by EUTRA, the
frequency error of HeNB should be relaxed because of the limited maximum speed. The computation is same to the
analysis for macro eNB. Assuming the operating frequency is 2.6GHz, the frequency accuracy of HeNB should be
within ±0.34ppm.

6.1.2.2 Cell capacity

One of the main motivations to introduce HeNB is providing high data rate services for indoor environment. Therefore,
the high order modulation and coding scheme such as 64QAM 5/6 must be supported by HeNB. However,

3GPP
Release 13 14 3GPP TR 36.922 V13.0.0 (2016-01)

BS frequency error can results in UE demodulation frequency error, thus resulting in performance degradation for
received signal as shown in the following formula.

sin(π N ∆f cTs ) − jπ ( N −1) ∆fcTs N −1 sin(π (m − k − N ∆f cTs ))

Rk =⋅
Sk e + ∑ Sm ⋅ e jπ (1−1/ N )( m − k − N ∆fcTs ) + n
N sin(π∆f cTs ) m =0 N sin(π (m − k − N ∆f cTs ) / N )
m≠k
N −1 N −1

=m 0
= ∑c m−kS m c0 S k + ∑ S m cm
=
m≠k

Where, Rk , S k , N , ∆f c , Ts are received signal, transmit signal, number of sub-carriers, frequency error and sampling
time interval respectively. The first part in the right side is the expected signal on the observed subcarrier and the
second part is the interferences from other sub-carriers other than the observed one. It is seen that frequency error
results in amplitude fading and phase rotation for the expected signal on the observed sub-carriers and ICI between sub-
carriers. Both these two factors will cause performance degradation for UE demodulation. Figure 6.1.2.2-1 shows the
SNR degradation because of the frequency error in OFDM. For example, if the signal to noise ratio (SNR) is 20dB, the
performance degradation can be ignored when the frequency error is less than 100Hz. If the frequency error is 400Hz,
the performance degradation will be larger than 1dB, which implies the signal to inter-subcarrier interference and noise
ratio (SINR) is less than 19dB.

Figure 6.1.2.2-1 SNR degradation VS frequency error

In fact, the performance degradation is related to the Doppler shift and the relative error between HeNB and UE.
Usually, UE has frequency offset estimation and compensation algorithm to follow frequency change due to mobility
and BS transmit frequency error. Therefore, the link performance depends on the performance of the estimation and
compensation algorithms. According to the OFDM performance analysis, as long as the residual frequency error after
compensation is less than one percent of the subcarrier interval, the link performance degradation can be ignored.

In order to evaluate the impact of 0.25ppm BS frequency error on 64QAM, we compare UE performance for the cases
listed in Table 6.1.2.2-1 using a commonly used UE algorithm. Since home eNode B is targeted at use in low delay
spread environment, only relatively low speed environment related EVA and EPA channel type is considered.

3GPP
Release 13 15 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.1.2.2-1 Scenarios for evaluation of HeNB frequency error requirement

Scenario #Propagation Doppler shift Frequency error

NOTE
condition
Scenario 1 EVA 70Hz* 0.05ppm
Scenario 2 EVA 70Hz* 0.25ppm
Scenario 3 EPA 5Hz** 0.05ppm
Scenario 4 EPA 5Hz** 0.25ppm
Note *: corresponding to typical UE speed of ~38km/h in EVA condition.
Note**: corresponding to typical UE speed of ~3km/h in EPA condition.

Usually the short term frequency stability will affect the demodulation performance more. The short term frequency
changing in time domain is in the order of few seconds and 10s may be an acceptable value to capture the frequency
change period. So we use a sine wave to model the frequency change due to frequency stability in the simulations.

∆f c = f max * sin (2π * t / T )

Where T =10s is the periodicity of the sine wave used to modulate the short term frequency stability characteristic in
time domain. f max is the frequency error (0.05ppm or 0.25ppm).

Figure 6.1.2.2-2 gives simulation results for scenario 1 and scenario 2 under EVA propagation channel. Figure 6.1.2.2-3
gives simulation results for scenario 3 and scenario 4.

0
PDSCH 1T2R EVA70Hz 5.0MHz- BLER vs SNR
10
Scenario 1
Scenario 2

-1
10
BLER v.s. SNR

-2
10

-3
10

-4
10
16 17 18 19 20 21 22 23 24
SNR in dB

Figure 6.1.2.2-2 Impact of frequency error on UE 64QAM demodulation performance, EVA70Hz

3GPP
Release 13 16 3GPP TR 36.922 V13.0.0 (2016-01)

PDSCH 1T2R EPA5Hz 5.0MHz-BLER vs SNR

0
10
Scenario 3
Scenario 4

-1
10
BLER v.s. SNR

-2
10

-3
10
16 17 18 19 20 21 22 23 24 25 26
SNR in dB

Figure 6.1.2.2-3 Impact of frequency error on UE 64QAM demodulation performance, EPA5Hz

It can be seen from the results, increasing frequency error requirement from current 0.05ppm to 0.25ppm results in
almost no performance degradation for HeNB 64QAM.

6.1.2.3 Timing
As the carrier frequency source is also used to generate the data clock [5], the frequency error is also relative to the
synchronization period when the network listening scheme is applied. As the HeNB cannot capture the GPS timing
signal in most deployment scenarios, network listening scheme is a feasible synchronization solution for HeNB. HeNB
may periodically utilize a synchronization signal such as the primary synchronization sequence (PSS), secondary
synchronization sequence (SSS) and common reference signal (CRS) from eNB to drive its timing. According to the
analysis in RAN4, the maximum synchronization period can be computed and listed in Table 6.1.2.3-1. Note that the
frequency error is the relative frequency error between eNB and HeNB, or HeNB and HeNB. For example, if the
frequency error of HeNB is 0.25ppm, the relative frequency error between eNB and HeNB is range from 0.2ppm to
0.3ppm, and the relative frequency error between HeNBs is range from 0 to 0.5ppm.

Table 6.1.2.3-1 Synchronization maintenance periods with different frequency error values

Frequency error Maximum synchronization period

0.2ppm 7.5s
0.3ppm 5s
0.4ppm 3.75s
0.5ppm 3s

3GPP
Release 13 17 3GPP TR 36.922 V13.0.0 (2016-01)

GPS timing reference

Synchronization Error: 50~100ns

Network Listen
eNB1 eNB2
HeNB1 HeNB2
Synchronization Error < 3us

Figure 6.1.2.3-1 Network listening synchronization scheme

According to the above analysis, if the frequency error of HeNB is stricter than 0.3ppm, the synchronization
maintenance period and related overhead seem to be acceptable.

6.1.2.4 Minimum requirement

The modulated carrier frequency of the HeNB shall be accurate to within ±0.25 ppm observed over a period of one
subframe (1ms).

6.1.3 Adjacent Channel Leakage power Ratio (ACLR)

Adjacent Channel Leakage power Ratio (ACLR) is the ratio of the filtered mean power centred on the assigned channel
frequency to the filtered mean power centred on an adjacent channel frequency.

The requirements shall apply whatever the type of transmitter considered (single carrier or multi-carrier). It applies for
all transmission modes foreseen by the manufacturer's specification. For a multi-carrier BS, the requirement applies for
the adjacent channel frequencies below the lowest carrier frequency transmitted by the BS and above the highest carrier
frequency transmitted by the BS for each supported multi-carrier transmission configuration. The requirement applies
during the transmitter ON period.

Reasonable ACLR requirements should be made to ensure the performance of macrocell operating in adjacent channel.
The relative and absolute ACLR requirements had been studied in RAN4, the main focus is ensuring the downlink
performance of the macrocell and both base station adjacent leakage power and UE blocking characteristics were
considered.

6.1.3.1 Minimum requirement

The ACLR is defined with a square filter of bandwidth equal to the transmission bandwidth configuration of the
transmitted signal (BWConfig) centred on the assigned channel frequency and a filter centred on the adjacent channel
frequency according to the tables below. Either the ACLR limits in the tables below or the absolute limit of -
50dBm/MHz apply, whichever is less stringent.

3GPP
Release 13 18 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.1.3.1-1: Home eNodeB ACLR in unpaired spectrum with synchronized operation

E-UTRA transmitted BS adjacent channel Assumed adjacent Filter on the adjacent ACLR
signal channel centre frequency channel carrier channel frequency and limit
bandwidth BWChannel offset below the first (informative) corresponding filter
[MHz] or above the last bandwidth
carrier centre
frequency transmitted
1.4, 3 BW Channel E-UTRA of same BW Square (BW Config) 45 dB
2 x BW Channel E-UTRA of same BW Square (BW Config) 45 dB
5, 10, 15, 20 BW Channel E-UTRA of same BW Square (BW Config) 45 dB
2 x BW Channel E-UTRA of same BW Square (BW Config) 45 dB
NOTE 1: BW Channel and BW Config are the channel bandwidth and transmission bandwidth configuration of the E-
UTRA transmitted signal on the assigned channel frequency.

6.1.4 Operating band unwanted emissions

The operating band unwanted emission limits are defined from 10 MHz below the lowest frequency of the downlink
operating band up to 10 MHz above the highest frequency of the downlink operating band.

The requirements shall apply whatever the type of transmitter considered (single carrier or multi-carrier) and for all
transmission modes foreseen by the manufacturer's specification.

The unwanted emission limits in the part of the downlink operating band that falls in the spurious domain are consistent
with ITU-R Recommendation SM.329 [11].

In E-UTRA, the SEM is specified in three regions [5].

• Region 1 (First adjacent channel): for 1.4, 3 and 5MHz, the offset between the channel edge frequency and the
centre of the measuring filter -3dB point is defined as the first adjacent channel and for 10, 15 and 20 MHz,
region 1 is defined as 0 to 5MHz. In this region, the emission limit usually has a slope and be relaxed compared
to ACLR1 limit in the first adjacent channel.

• Region 2 (Second adjacent channel): for 1.4, 3 and 5MHz, region 2 is defined as second adjacent channel and
for 10, 15 and 20MHz, region 2 is defined as 5 to 10MHz. In this region, the emission limit has tapered off and is
usually defined with a fixed value which cloud be set to the same order of ACLR2 requirement.

• Region 3 (Spurious emission): in this region, the SEM is usually defined by the spurious emission limit.

The relative ACLR1 and ACLR2 requirements are proposed to 45dBc in RAN4. Assuming 20dBm maximum output
power and 5MHz bandwidth for HeNB, the second adjacent channel leakage power at point B of figure 6.1.4-1 is about
-42dBm/100kHz (20dBm – 45dBc = -25dBm/4.5MHz). We propose to make a 6dB maximum allowed relaxation for
SEM in the first adjacent channel and then get an emission limit of -36dBm/100kHz at point A. Therefore, the emission
6  f _ offset 
mask for first adjacent channel will be defined as −36dBm −  − 0.05  dB , where the f_offset is the
5  MHz 
separation between the channel edge frequency and the centre of the measuring filter. The second adjacent channel
emission mask could be defined as -42dBm/100kHz.

3GPP
Release 13 19 3GPP TR 36.922 V13.0.0 (2016-01)

ACLR1=45dBc

-36dBm/100kHz P=20dBm

-42dBm/100kHz
P=1.5dBm
-50dBm/MHz

A B C
△f(from carrier edge)

0 5 10
(Carrier Edge)

st nd
Figure 6.1.4-1 Spectrum emission mask (1 and 2 adjacent channel)

In region 3, the emission limit is determined by the relative and absolute ACLR requirement whichever is less stringent,
seen in figure 6.1.4-2. The absolute ACLR1&2 requirement is set to -50dBm/MHz. The emission limit in this domain is
proposed to be defined as a function of the maximum output power of HeNB. The upper limit is about -32dBm/MHz
(20dBm – 45dBc – 10log4.5) and the lower limit is -50dBm/MHz. Then the emission limit is specified by equation 2-1.

 P − 52 dB , where 2dBm ≤ P ≤ 20dBm

 (6.2.5-1)
 −50 dBm, where P<2dBm

P=20dBm
Relative ACLR
-32dBm/MHz
-32dBm/MHz

P=2dBm
Absolute ACLR
-50dBm/MHz
-50dBm/MHz

△f(from carrier edge)

10

Figure 6.1.4-2 Spectrum emission mask layout (beyond 10MHz)

6.1.4.1 Minimum requirements

For E-UTRA Home eNodeB emissions shall not exceed the maximum levels specified in the Tables 6.1.4.1-1 to
6.2.5.1-3, where:

3GPP
Release 13 20 3GPP TR 36.922 V13.0.0 (2016-01)

- ∆f is the separation between the channel edge frequency and the nominal -3dB point of the measuring filter
closest to the carrier frequency.

- f_offset is the separation between the channel edge frequency and the centre of the measuring filter.

- f_offsetmax is the offset to the frequency 10 MHz outside the downlink operating band.

- ∆fmax is equal to f_offsetmax minus half of the bandwidth of the measuring filter.

For a multicarrier E-UTRA Home eNodeB the definitions above apply to the lower edge of the carrier transmitted at the
lowest carrier frequency and the higher edge of the carrier transmitted at the highest carrier frequency.

Table 6.1.4.1-1: General operating band unwanted emission limits for 1.4 MHz channel bandwidth

Frequency offset of Frequency offset of Minimum requirement Measurement

measurement measurement filter centre bandwidth
filter -3dB point, ∆f frequency, f_offset (Note 1)
6  f _ offset  100 kHz
0 MHz ≤ ∆f < 1.4 MHz 0.05 MHz ≤ f_offset < 1.45 MHz −30 dBm −  − 0.05  dB
1.4  MHz 
1.4 MHz ≤ ∆f < 2.8 MHz 1.45 MHz ≤ f_offset < 2.85 MHz -36 dBm 100 kHz
2.8 MHz ≤ ∆f ≤ ∆fmax 3.3 MHz ≤ f_offset < f_offsetmax  P − 52dB, 2dBm ≤ P ≤ 20dBm 1MHz

−50dBm, P<2dBm

Table 6.1.4.1-2: General operating band unwanted emission limits for 3 MHz channel bandwidth

Frequency offset of Frequency offset of Minimum requirement Measurement

measurement measurement filter centre bandwidth
filter -3dB point, ∆f frequency, f_offset (Note 1)

−34 dBm − 2 
 f _ offset − 0.05  dB 100 kHz
0 MHz ≤ ∆f < 3 MHz 0.05 MHz ≤ f_offset < 3.05 MHz 
 MHz 
3 MHz ≤ ∆f < 6 MHz 3.05 MHz ≤ f_offset < 6.05 MHz -40 dBm 100 kHz
6 MHz ≤ ∆f ≤ ∆fmax 6.5 MHz ≤ f_offset < f_offsetmax  P − 52dB, 2dBm ≤ P ≤ 20dBm 1MHz

−50dBm, P<2dBm

Table 6.1.4.1-3: General operating band unwanted emission limits for 5, 10, 15 and 20 MHz channel
bandwidth

Frequency offset of Frequency offset of Minimum requirement Measurement

measurement measurement filter centre bandwidth
filter -3dB point, ∆f frequency, f_offset (Note 1)
0 MHz ≤ ∆f < 5 MHz 0.05 MHz ≤ f_offset < 5.05 MHz 6  f _ offset  100 kHz
−36 dBm −  − 0.05  dB
5  MHz 
5 MHz ≤ ∆f < 10 MHz 5.05 MHz ≤ f_offset < 10.05 MHz -42 dBm 100 kHz
10 MHz ≤ ∆f ≤ ∆fmax 10.5 MHz ≤ f_offset < f_offsetmax  P − 52dB, 2dBm ≤ P ≤ 20dBm 1MHz

−50dBm, P<2dBm

NOTE 1 As a general rule, the resolution bandwidth of the measuring equipment should be equal to the
measurement bandwidth. However, to improve measurement accuracy, sensitivity and efficiency, the
resolution bandwidth can be smaller than the measurement bandwidth. When the resolution bandwidth is
smaller than the measurement bandwidth, the result should be integrated over the measurement
bandwidth in order to obtain the equivalent noise bandwidth of the measurement bandwidth.

NOTE 2 The parameter P is defined as the aggregated maximum power of all transmit antenna ports of Home
eNodeB.

3GPP
Release 13 21 3GPP TR 36.922 V13.0.0 (2016-01)

6.1.5 Spurious emissions

6.1.5.1 Mandatory requirements

The requirements of either subclause 6.6.4.1.1 (Category A limits) or subclause 6.6.4.1.2 (Category B limits) of
TS36.104 [5] shall also apply for HeNB.

6.1.5.2 Co-existence with HNB/HeNB operating in other bands

Taking into account the expected deployment scenarios of HeNB, coexistence with other types of base station are not
meaningful. Therefore, only coexistence spurious emission requirements for protection other cross band HNB/HeNB
operating in the same geographic area will be specified.

The assumed scenario for coexistence with other HNB/HeNB is described in Figure 6.1.5.2-1. Two HeNBs are placed
in different rooms and opposite to a wall.

HeNB1 HeNB2

Figure 6.1.5.2-1 Assumed interference scenario for coexistence with other HNB/HeNB

PL(dB) = 127+30log10(R/1000), R in m (6.2.6.2-1)

The path loss model listed above is used to calculate the path loss between two HeNBs, seen equation (6.2.6.2-1). The
minimum separation distance between two HeNBs is assumed to be 1 meter. Assuming 10dB penetration loss and 0dBi
antenna gain of HeNB, we can get a MCL of 47dB for co-location with other HNB/HeNB.

The maximum allowed interference power level is determined based on 0.8dB desensitization criterion. Assuming 13dB
noise figure and 47dB MCL between two HeNBs, we can get a spurious emission limit of -71dBm/100kHz (-174dBm +
50dB + 13dB -7dB + 47dB) to protect other HeNB. However, there are 6dB difference in noise figure between HNB
and HeNB which will bring also 6dB difference in the co-existence requirement. Since operators may deploy HNB and
HeNB in the same operating band in real implementations, it's propose to use -71dBm/100kHz as the common
requirement for protection of other HNB/HeNB operating in other frequency bands to simplify the specification.

6.1.5.2.1 Minimum requirement

The power of any spurious emission shall not exceed the limits of Table 6.1.5.2.1-1 for a HeNB where requirements for
coexistence with other HNB/HeNB in the same geographic area listed in the first column apply.

3GPP
Release 13 22 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.1.5.2.1-1: HeNB Spurious emissions limits for coexistence with other HNB/HeNB in the same
geographic area

Type of coexistence BS Frequency range for co- Maximum Measurement Note

location requirement Level Bandwidth
UTRA FDD Band I or E- 1920 – 1980 MHz -71 dBm 100 kHz
UTRA Band 1
UTRA FDD Band II or E- 1850 – 1910 MHz -71 dBm 100 kHz
UTRA Band 2
UTRA FDD Band III or E- 1710 – 1785 MHz -71 dBm 100 kHz
UTRA Band 3
UTRA FDD Band IV or E- 1710 – 1755 MHz -71 dBm 100 kHz
UTRA Band 4
UTRA FDD Band V or E- 824 – 849 MHz -71 dBm 100 kHz
UTRA Band 5
UTRA FDD Band VI or E- 815 – 850 MHz -71 dBm 100 kHz
UTRA Band 6
UTRA FDD Band VII or 2500 – 2570 MHz -71 dBm 100 KHz
E-UTRA Band 7
UTRA FDD Band VIII or 880 – 915 MHz -71 dBm 100 KHz
E-UTRA Band 8
UTRA FDD Band IX or E- 1749.9 – 1784.9 MHz -71 dBm 100 KHz
UTRA Band 9
UTRA FDD Band X or E- 1710 – 1770 MHz -71 dBm 100 kHz
UTRA Band 10
UTRA FDD Band XI or E- 1427.9 – 1452.9 MHz -71 dBm 100 kHz
UTRA Band 11
UTRA FDD Band XII or 698 – 716 MHz -71 dBm 100 kHz
E-UTRA Band 12
UTRA FDD Band XIII or 777 – 787 MHz -71 dBm 100 kHz
E-UTRA Band 13
UTRA FDD Band XIV or 788 – 798 MHz -71 dBm 100 kHz
E-UTRA Band 14
E-UTRA Band 17 704 – 716 MHz -71 dBm 100 kHz
UTRA TDD in Band a) or 1900 – 1920 MHz -71 dBm 100 kHz
E-UTRA Band 33
UTRA TDD in Band a) or 2010 – 2025 MHz -71 dBm 100 kHz
E-UTRA Band 34
UTRA TDD in Band b) or 1850 – 1910 MHz -71 dBm 100 kHz
E-UTRA Band 35
UTRA TDD in Band b) or 1930 – 1990 MHz -71 dBm 100 kHz
E-UTRA Band 36
UTRA TDD in Band c) or 1910 – 1930 MHz -71 dBm 100 kHz
E-UTRA Band 37
UTRA TDD in Band d) or 2570 – 2620 MHz -71 dBm 100 kHz
E-UTRA Band 38
E-UTRA Band 39 1880 – 1920MHz -71 dBm 100 kHz
E-UTRA Band 40 2300 – 2400MHz -71 dBm 100 kHz

NOTE 1: As defined in the scope for spurious emissions in this clause, the coexistence requirements in
Table 6.1.5.2.1-1 do not apply for the 10 MHz frequency range immediately outside the HeNB transmit
frequency range of a downlink operating band. This is also the case when the transmit frequency range is
adjacent to the Band for the co-location requirement in the table. The current state-of-the-art technology
does not allow a single generic solution for co-location with other system on adjacent frequencies for
30dB BS-BS minimum coupling loss. However, there are certain site-engineering solutions that can be
used. These techniques are addressed in TR 25.942 [9].

6.1.6 Transmitter intermodulation

The transmitter intermodulation requirement is a measure of the capability of the transmitter to inhibit the generation of
signals in its non linear elements caused by presence of the own transmit signal and an interfering signal reaching the
transmitter via the antenna. The requirement applies during the transmitter ON period and the transmitter transient
period.

3GPP
Release 13 23 3GPP TR 36.922 V13.0.0 (2016-01)

6.1.6.1 Minimum requirement

The transmitter intermodulation level is the power of the intermodulation products when an interfering signal is injected
into the antenna connector. The wanted signal channel bandwidth BWChannel shall be the maximum bandwidth supported
by the base station. The offset of the interfering signal from the wanted signal shall be as in Table 6.1.6.1-1.

Table 6.1.6.1-1 Interfering and wanted signals for the Transmitter intermodulation requirement

Parameter Value
Wanted signal E-UTRA signal of maximum channel bandwidth
BW Channel
Interfering signal type E-UTRA signal of channel bandwidth 5 MHz
Interfering signal level Mean power level 30 dB below the mean power
of the wanted signal
Interfering signal centre -BW Channel /2 – 12.5 MHz
frequency offset from -BW Channel /2 – 7.5 MHz
wanted signal carrier -BW Channel /2 – 2.5 MHz
centre frequency BW Channel /2 + 2.5 MHz
BW Channel /2 + 7.5 MHz
BW Channel /2 + 12.5 MHz
NOTE: Interfering signal positions that are partially or completely outside
of the downlink operating band of the base station are excluded
from the requirement.

The transmitter intermodulation level shall not exceed the unwanted emission limits in subclause 6.1.3, 6.1.4 and 6.1.5
in the presence of an interfering signal according to Table 6.1.6.1-1. The measurement may be limited to frequencies on
which third and fifth order intermodulation products appear, considering the width of these products.

6.2 Receiver characteristics

6.2.1 Reference sensitivity level
Reference sensitivity level is the minimum mean power received at the antenna connector at which a throughput
requirement shall be met for a specified reference measurement channel. The main purpose to define the reference
sensitivity requirement is to verify the receiver noise figure. Receiver noise figure will affect the uplink performance of
macrocell and HeNB itself desensitization. These impacts could be studied by system level simulations.

6.2.1.1 Uplink performance degradation of macrocell

6.2.1.1.1 Simulation setup

The simulation parameters and assumptions are discussed and agreed on in RAN4 for HeNB (detailed information
could be got in R4-092042). The hierarchical deployment scenario is illustrated in Figure 6.2.1.1.1-1. 100 HeNBs are
deployed in a sector and each has one active HUE. Since the ACLR of UE is 30dB, the ACIR is assumed to be 30dB for
adjacent interference calculation.

3GPP
Release 13 24 3GPP TR 36.922 V13.0.0 (2016-01)

Figure 6.2.1.1.1-1 Hierarchical deployment scenario of macrocell and HeNB

In the simulations, the uplink power control scheme described in TS36.211 [4] is used. The noise figure will affect the
uplink MCS selection and finally result in different output power setting. High output power of HUE may result in
performance degradation of the macrocell. Therefore, the noise figure should be well planned to ensure the uplink
performance of macrocell.

6.2.1.1.2 Simulation results

The noise floor of macro eNB with 10MHz bandwidth is about -100 dBm (-174 + 60 + 9.5 + 5 = -99.5 dBm). Therefore,
the noise floor of HeNB in the simulation is assumed to be in the range from -99 dBm to -79 dBm. As a function of
HeNB noise floor, the relative uplink throughput loss of macrocell is shown in Figure 6.2.1.1.2-1.

Figure 6.2.1.1.2-1 Relative uplink throughput loss versus noise floor

It is important to control the noise floor in a reasonable range to minimise the impact on uplink performance of
macrocell operating in adjacent channel. Assuming the maximum allowed performance degradation is 3% [10],
compared to macro eNB, it will desensitize the HeNB reference sensitivity by 10 to 13 dB, which corresponds to a
noise floor of -89dBm and -86dBm.

3GPP
Release 13 25 3GPP TR 36.922 V13.0.0 (2016-01)

6.2.1.2 HeNB desensitization

In this section, the impact on sensitivity degradation of HeNB due to interference from MUE is simulated. The
simulation assumptions and deployment scenarios are the same as section 6.3.1.1. In the simulations, one MUE
occupying the whole uplink bandwidth (10MHz) is randomly placed in the building block where HeNBs are deployed
and no RoT control is considered for macrocell. In order to study the impact of different output power of MUE, the
building block with HeNBs and MUE is placed in R/2 and R respectively, where R is the radius of the macrocell.

Figure 6.2.1.2-1 gives the simulation results of uplink noise rise of HeNB versus different block locations. The blue
curve represents noise rise due to uplink interference from other HUEs and the red together with green curve represent
additional noise rise due to uplink interference from MUE. The additional noise rise is calculated based on 40% HeNBs
suffering from the highest interference. Seen from the results, we can find that the additional noise rise is about 9 dB in
the worst case (the MUE is located in cell edge with high output power, D = R) and 5.5 dB in a normal case (the
distance between building block and macro eNB is R/2). Therefore, according to the simulation result shown in Figure
6.2.1.2-1, compared with macro eNB, 7 to 8 dB desensitization seems to be a good tradeoff for TD-LTE HeNB.

Figure 6.2.1.2-1 Noise rise versus different HeNB locations

For another approach in statistics, Figure 6.2.1.2-2 and Figure6.2.1.2-3 demonstrates the CDF of HeNB noise rise with
MUE being in R/2 and R respectively. It is observed that the percentage of HeNBs with the highest interference from
MUE have very big impact on the maximum tolerable noise rise. In Table 6.2.1.2-1 some noise rise values are
summarize for different percentage of HeNB in statistics. If we decide HeNB noise rise based on 40% HeNBs suffering
from the highest interference, the HeNB sensitivity can be degraded by 7~9dB.

3GPP
Release 13 26 3GPP TR 36.922 V13.0.0 (2016-01)

Figure 6.2.1.2-2 HeNB noise rise CDF, MUE distance from MeNB D=R/2

Figure 6.2.1.2-3 HeNB noise rise CDF, MUE distance from MeNB D=R

Table 6.2.1.2-1 Summary of HeNB noise rise due to MUE interference

Probability of HeNB that observe MUE distance from MeNB, MUE distance from MeNB,
highest interference D=R/2 D=R(worst case)
50% 2.5dB 3.59dB
40% 7.43dB 9.4dB
30% 14.9dB 17.14dB

3GPP
Release 13 27 3GPP TR 36.922 V13.0.0 (2016-01)

6.2.1.3 Minimum requirement

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel as specified in
Annex A in TS36.104 with parameters specified in Table 6.2.1.3-1.

Table 6.2.1.3-1: HeNB reference sensitivity levels

Reference sensitivity power level,

E-UTRA
Reference measurement channel PREFSENS
channel bandwidth [MHz]
[dBm]
1.4 FRC A1-1 in Annex A.1 -98.8
3 FRC A1-2 in Annex A.1 -95.0
5 FRC A1-3 in Annex A.1 -93.5
10 FRC A1-3 in Annex A.1* -93.5
15 FRC A1-3 in Annex A.1* -93.5
20 FRC A1-3 in Annex A.1* -93.5
Note*: PREFSENS is the power level of a single instance of the reference measurement channel. This requirement shall
be met for each consecutive application of a single instance of FRC A1-3 mapped to disjoint frequency ranges
with a width of 25 resource blocks each

6.2.2 Dynamic range

The impact of co-channel uplink interference from an uncoordinated UE on the HeNB needs to be studied to derive a
reasonable dynamic range requirement. The co-channel interference from an uncoordinated macro UE (MUE) and
home UE (HUE) are studied based on deterministic analysis and system-level simulations respectively in the following
sections.

6.2.2.1 Deterministic analysis

The assumed scenario for coexistence with uncoordinated MUE is described in Figure 6.2.2.1-1. The HeNB (CSG) is
located on a table in an apartment. A MUE is placed in the same apartment and establishes a call with the macro BS.

Macro BS
HeNB
MUE

Figure 6.2.2.1-1 Assumed interference scenario for coexistence with uncoordinated UE

HeNB will cause a coverage hole (dead zone) to the co-channel deployed macrocell if received interference power at
MUE exceeds the decoding threshold specified in TS36.101 [6]. The size of dead zone is determined by the output
power of HeNB and the received wanted signal power of macro cell. The extension of the dead zone is restricted to be
within several meters. The path loss model listed in equation (6.2.6.2-1) is used to determine the minimum distance that
the MUE is able to go close to the HeNB. Figure 6.2.2.1-2 gives the relationship between the separation distance and
received interference power level at HeNB antenna port. It's proposed to set the maximum received interference power
level to -38dBm assuming 6 meters separation distance (dead zone) and 23dBm (PC 3 MUE) maximum output power.

3GPP
Release 13 28 3GPP TR 36.922 V13.0.0 (2016-01)

Figure 6.2.2.1-2 Minimum separation distance

6.2.2.2 System-level simulations

The assumed scenario and simulation assumptions are the same as section 6.2.1.1.1. The hierarchical deployment
scenario of macrocell and HeNB is illustrated in Figure 6.2.2.2-1. One HeNB building block is randomly placed in a
sector.

Figure 6.2.2.2-1 Hierarchical deployment scenario of macrocell and HeNB

Both co-channel interference from uncoordinated HUE and MUE are considered in our study. The MUE is located in
the cell border and establishes a call with maximum output power. Fifty HUEs are randomly placed into the building
block and each HeNB has an active user. The co-channel interference caused by MUE and HUE are defined as
interference over thermal noise (IoT). In our simulations, uplink power control scheme described in TS36.211 [4] is
used (Alpha = 1.0, P0 = -106dBm/RB). The simulation results are illustrated in Figure 6.2.2.2-2. The red line and blue
line represent IoT level caused by other HUE and both other HUE and uncoordinated MUE respectively. Seen from the
simulation results, we can get the following conclusions.

• Due to the limit coverage and deployment scenario, the HUE is very close to HeNB in most of cases and
maintains the connection at low power level. Therefore, the main interference comes from the uncoordinated
MUE.

3GPP
Release 13 29 3GPP TR 36.922 V13.0.0 (2016-01)

• It's proposed to set the maximum IoT to 55dB to make sure the HeNB could suffer the interference from other
HUE and uncoordinated MUE in most cases (99%).

1
1
0.99
0.8 0.98
0.97
0.6 0.96
CDF

CDF
0.95
0.4 0.94
0.93
0.2 0.92
IoT HUE IoT HUE
0.91
IoT HUE&MUE IoT HUE&MUE
0 0.9
0 20 40 60 80 100 0 20 40 60 80 100
IoT(dB) IoT(dB)

Figure 6.2.2.2-2 IoT level of HeNB

6.2.2.3 Minimum requirement

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel as specified in
Annex A with parameters specified in Table 6.2.2.3-1.

Table 6.2.2.3-1: Dynamic range

E-UTRA Interfering
Reference Wanted signal Type of
channel signal mean
measurement mean power interfering
bandwidth power [dBm] /
channel [dBm] signal
[MHz] BWConfig
FRC A2-1 in
1.4 -31.8 -44.2 AWGN
Annex A.2
FRC A2-2 in
3 -27.9 -40.2 AWGN
Annex A.2
FRC A2-3 in
5 -25.7 -38 AWGN
Annex A.2
FRC A2-3 in
10 -25.7 -35 AWGN
Annex A.2*
FRC A2-3 in
15 -25.7 -33.2 AWGN
Annex A.2*
FRC A2-3 in
20 -25.7 -31.9 AWGN
Annex A.2*

6.2.3 Adjacent channel selectivity (ACS) and narrow-band blocking

Adjacent channel selectivity (ACS) is a measure of the receiver ability to receive a wanted signal at its assigned channel
in the presence of an adjacent channel signal with a specified centre frequency offset of the interfering signal to the
band edge of a victim system. The HeNB receiver must have the ability to against the adjacent channel interference
from the uncoordinated macrocell user. The following sections give the study results.

6.2.3.1 Simulation assumptions

The assumed coexistence scenario of macrocell and HeNB is illustrated in Figure 6.2.3.1-1. The HeNB and macro
eNodeB (MeNB) are working in adjacent channel. The HeNB building block is located in the cell border and 10 MUEs
are randomly placed in each sector. Detailed deployment parameters are listed in Table 6.2.3.1-1.

3GPP
Release 13 30 3GPP TR 36.922 V13.0.0 (2016-01)

10 m

10 m

MBS 10 m

10 m

10 m

Figure 6.2.3.1-1 Assumed interference scenario for coexistence with uncoordinated UE(s)

Table 6.2.3.1-1 Simulation parameters

Macro cell parameters Value

Cellular Layout Hexagonal grid, 3 sectors per site, reuse 1.
Inter-site distance 500 m
UE power class 23 dBm (200 mW)
UE distribution 80% inside the building
HeNB parameters Value
Number of HeNB per row 10
Number of blocks per sector 1
Number of floors per block 6
Number of HeNB 50
activation ratio 100%
Maximum output power 20dBm
Other parameters Value
Propagation model PL(dB) = 127+30log10(R/1000), R in m
Log-normal shadowing standard deviation 10 dB

6.2.3.2 Simulation results

The macro UE will be blocked when the adjacent channel interference power is larger than -20dBm (assuming 5dB
margin based on minimum ACS requirement specified in TS36.101 [6]). Assuming the maximum output power of
HeNB is 20dBm, we can get a minimum separation of 40dB between HeNB and MUE. UEs receiving higher
interference than a blocking threshold of -20dBm will be removed from the UL interference statistics.

Figure 6.2.3.2-1 and 6.2.3.2-2 gives the uplink interference statistics caused by uncoordinated macro UE. Based on the
simulation results, we proposed to define the adjacent channel interference level to -28dBm (24dB higher than EUTRA
macro BS) which results in about 1% blocking probability of HeNB.

3GPP
Release 13 31 3GPP TR 36.922 V13.0.0 (2016-01)

1 1

0.9

DL Blocking Threshold(-20dBm) 0.998

0.8

0.7

CDF of UL interference
CDF of UL interference

0.996
0.6

0.5
0.994
0.4

0.3
0.992 DL Blocking Threshold(-20dBm)
0.2

0.1 0.99

0
-180 -160 -140 -120 -100 -80 -60 -40 -20 -28 -26 -24 -22 -20 -18 -16
dBm dBm

Figure 6.2.3.2-1 CDF of UL interference Figure 6.2.3.2-2 CDF of UL interference (Zoom in

view)

6.2.3.3 Minimum requirements

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel, with a wanted and
an interfering signal coupled to the BS antenna input as specified in Tables 6.2.3.3-1 and 6.2.3.3-2 for narrowband
blocking and in Table 6.2.3.3-3 for ACS. The reference measurement channel for the wanted signal is identified in
Table 6.2.1.3-1 for each channel bandwidth and further specified in Annex A of TS36.104 [5].

Table 6.2.3.3-1: Narrowband blocking requirement

Wanted signal mean Interfering signal

Type of interfering signal
power [dBm] mean power [dBm]
PREFSENS + 14dB* -33 See Table 6.2.3.3-2
Note*: PREFSENS depends on the channel bandwidth as specified in
Table 6.2.1.3-1.

Table 6.2.3.3-2: Interfering signal for Narrowband blocking requirement

Interfering RB
E-UTRA centre frequency
Assigned offset to the Type of interfering signal
BW [MHz] channel edge of the
wanted signal [kHz]
252.5+m*180, 1.4 MHz E-UTRA signal, 1
1.4
m=0, 1, 2, 3, 4, 5 RB*
247.5+m*180,
3 m=0, 1, 2, 3, 4, 7, 10, 3 MHz E-UTRA signal, 1 RB*
13
342.5+m*180,
5 m=0, 1, 2, 3, 4, 9, 14, 5 MHz E-UTRA signal, 1 RB*
19, 24
347.5+m*180,
10 m=0, 1, 2, 3, 4, 9, 14, 5 MHz E-UTRA signal, 1 RB*
19, 24
352.5+m*180,
15 m=0, 1, 2, 3, 4, 9, 14, 5 MHz E-UTRA signal, 1 RB*
19, 24
342.5+m*180,
20 m=0, 1, 2, 3, 4, 9, 14, 5 MHz E-UTRA signal, 1 RB*
19, 24
Note*: Interfering signal consisting of one resource block
adjacent to the wanted signal

3GPP
Release 13 32 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.2.3.3-3: Adjacent channel selectivity

E-UTRA Interfering signal centre

Interfering
channel Wanted signal mean frequency offset from
signal mean Type of interfering signal
bandwidth power [dBm] the channel edge of the
power [dBm]
[MHz] wanted signal [MHz]
1.4 PREFSENS + 27dB* -28 0.7025 1.4MHz E-UTRA signal
3 PREFSENS + 24dB* -28 1.5075 3MHz E-UTRA signal
5 PREFSENS + 22dB* -28 2.5025 5MHz E-UTRA signal
10 PREFSENS + 22dB* -28 2.5075 5MHz E-UTRA signal
15 PREFSENS + 22dB* -28 2.5125 5MHz E-UTRA signal
20 PREFSENS + 22dB* -28 2.5025 5MHz E-UTRA signal
Note*: PREFSENS depends on the channel bandwidth as specified in Table 6.2.1.3-1.

6.2.4 Blocking characteristics

The blocking characteristic is a measure of the receiver ability to receive a wanted signal at its assigned channel in the
presence of an unwanted interferer. The HeNB receiver must have the ability to against the interference from
uncoordinated UE and other co-location HNB/HeNB. The following sections give the study results.

6.2.4.1 General blocking requirement

The general blocking requirement consists of in-band blocking and out-of-band blocking. The unwanted interferer is
presented by E-UTRA signal for in-band blocking and a CW signal for out-of-band blocking.

In E-UTRA [8], the mean power of the E-UTRA interfering signal is equal to -43dBm which is a compromise between
the 30dBm and 24dBm maximum output power assumption in TR36.942 [9] under worst case MCL condition.

The assumed coexistence scenario of macrocell and HeNB is illustrated in Figure 6.2.4.1-1. The HeNB and macro
eNodeB (MeNB) are working in adjacent channel. The HeNB building block is located in the cell border and 10 MUEs
are randomly placed in each sector. Detailed deployment parameters are listed in Table 6.2.4.1-1.

10 m

10 m

MBS 10 m

10 m

10 m

Figure 6.2.4.1-1 Assumed interference scenario for coexistence with uncoordinated UE(s)

3GPP
Release 13 33 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.2.4.1-1 Simulation parameters

Macro cell parameters Value

Cellular Layout Hexagonal grid, 3 sectors per site, reuse 1.
Inter-site distance 500 m
UE power class 23 dBm (200 mW)
UE distribution 80% inside the building
HeNB parameters Value
Number of HeNB per row 10
Number of blocks per sector 1
Number of floors per block 6
Number of HeNB 50
activation ratio 100%
Maximum output power 20dBm
Other parameters Value
Propagation model PL(dB) = 127+30log10(R/1000), R in m
Log-normal shadowing standard deviation 10 dB

The macro UE will be blocked when the interference power is larger than -25dBm (assuming 5dB margin based on
minimum blocking requirement specified in TS36.101 [6]). Assuming the maximum output power of HeNB is 20dBm,
we can get a minimum separation of 45dB between HeNB and MUE. UEs receiving higher interference than a blocking
threshold of -25dBm will be removed from the UL interference statistics.

Figure 6.2.4.1-2 and 6.2.4.1-3 gives the uplink interference statistics caused by uncoordinated macro UE. Based on the
simulation results, we proposed to define the channel interference level to -27dBm for blocking requirement which
results in about 0.8% blocking probability of HeNB. In the meantime, we observe that -15dBm out-of-band blocking
requirement seems to be also sufficient for HeNB.

1 1

0.9

0.998
0.8

0.7
CDF of UL interference
CDF of UL interference

DL Blocking Threshold(-25dBm)
0.996
0.6

0.5
0.994
0.4

0.3 0.992

0.2
DL Blocking Threshold(-25dBm)

0.1 0.99

0 -30 -29 -28 -27 -26 -25 -24 -23 -22

-180 -160 -140 -120 -100 -80 -60 -40 -20
dBm dBm

Figure 6.2.4.1-2 CDF of UL interference Figure 6.2.4.1-3 CDF of UL interference (Zoom in

view)

6.2.4.1.1 Minimum requirement

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel, with a wanted and
an interfering signal coupled to BS antenna input using the parameters in Table 6.2.4.1.1-1 and 6.2.4.1.1-2. The
reference measurement channel for the wanted signal is identified in Table 6.2.1.3-1 for each channel bandwidth and
further specified in Annex A of TS36.104.

3GPP
Release 13 34 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.2.4.1.1-1: Blocking performance requirement for HeNB

Operating Centre Frequency of Interfering Interfering Wanted Signal Interfering signal Type of
Band Signal [MHz] Signal mean power centre frequency Interfering
mean power [dBm] minimum Signal
[dBm] frequency offset
from the
channel edge of
the wanted
signal [MHz]
1-7, 9-11, (FUL_low -20) to (FUL_high +20) -27 PREFSENS +14dB* See table See table
13-14, 33- 6.2.4.1.1.1-2 6.2.4.1.1-2
40 1 to (FUL_low -20) -15 PREFSENS +14dB*  CW carrier
(FUL_high +20) to 12750
8 (FUL_low -20) to (FUL_high +10) -27 PREFSENS +14dB* See table See table
6.2.4.1.1-2 6.2.4.1.1-2
1 to (FUL_low -20) -15 PREFSENS +14dB*  CW carrier
(FUL_high +10) to 12750
12 (FUL_low -20) to (FUL_high +12) -27 PREFSENS +14dB* See table See table
6.2.4.1.1-2 6.2.4.1.1-2
1 to (FUL_low -20) -15 PREFSENS +14dB*  CW carrier
(FUL_high +12) to 12750
17 (FUL_low -20) to (FUL_high +18) -27 PREFSENS +14dB* See table See table
6.2.4.1.1-2 6.2.4.1.1-2
1 to (FUL_low -20) -15 PREFSENS +14dB*  CW carrier
(FUL_high +18) to 12750
Note*: PREFSENS depends on the channel bandwidth as specified in Table 6.2.1.3-1.

Table 6.2.4.1.1-2: Interfering signals for blocking performance requirement for HeNB

Interfering signal
E-UTRA centre frequency
channel minimum offset to
Type of interfering signal
BW the channel edge of
[MHz] the wanted signal
[MHz]
1.4 2.1 1.4MHz E-UTRA signal
3 4.5 3MHz E-UTRA signal
5 7.5 5MHz E-UTRA signal
10 7.5 5MHz E-UTRA signal
15 7.5 5MHz E-UTRA signal
20 7.5 5MHz E-UTRA signal

6.2.4.2 Co-location with other HNB/HeNB

The assumed scenario for co-location with other HNB/HeNB is described in Figure 6.2.4.2-1. Two HeNBs are placed in
different rooms and opposite to a wall.

34
Release 13 35 3GPP TR 36.922 V13.0.0 (2016-01)

HeNB1 HeNB2

Figure 6.2.4.2-1 Assumed interference scenario for co-location with other HNB/HeNB

The minimum coupling loss between two co-located HeNBs is assumed to be 47dB.Assuming the maximum output
power of HeNB is also 20dBm as HNB, the co-location blocking requirement to against other nearby cross-band
HNB/HeNB will be -27dBm.

Based on above analysis, it's proposed to set the co-location blocking requirement to -27dBm to against other nearby
cross-band HNB/HeNB. In addition, the power difference between wanted signal and interference signal is proposed to
be the same as EUTRA macro BS co-location blocking test. These requirements will be specified in section 7.6.2.1 of
TS36.104 [5].

6.2.5 Receiver Intermodulation

6.2.5.1 Analysis
Receiver inter-modulation can occur when two interfering signals with a particular relationship are applied to a BS
receiver. Two large interfering signals at the same time occur less frequently than a single interfering signal. Due to
lower probability of two large interfering signals, the power level of the interfering signals for the inter-modulation
requirement should be lower compared to Blocking requirement. For the Macro eNB, the level of IM interfering signals
is -52dBm which is 9 dB lower compared to Blocking requirement of -43dBm. It is proposed to use the same relative
values also for the home eNode B.

In the TR, the blocking interference level for Home eNode B is proposed to be -27dBm. Adopting the same relative
values of 9dB also for Home eNode B inter-modulation requirement, the following interfering signals level for inter-
modulation is proposed:

- Interfering signals: -36 dBm for both modulated and CW interferer.

As for the wanted signal level, it is proposed to keep the same relative value with interfering signal as that for MeNB in
order not to put more stringent requirement for HeNB. So the following wanted signal level is proposed.

- Wanted signal: PREFSENS+14dB

6.2.5.2 Minimum requirement

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel, with a wanted signal
at the assigned channel frequency and two interfering signals coupled to the BS antenna input, with the conditions
specified in Tables 6.2.5.2-1 and 6.2.5.2-2 for intermodulation performance and in Table 6.2.5.2-3 for narrowband
intermodulation performance. The reference measurement channel for the wanted signal is identified in section 6.3.1 for
each channel bandwidth and further specified in Annex A of [5].

35
Release 13 36 3GPP TR 36.922 V13.0.0 (2016-01)

Table 6.2.5.2-1: Intermodulation performance requirement

Interfering signal
Wanted signal mean power [dBm] mean power Type of interfering signal
[dBm]
PREFSENS + 14dB* -36 See Table 6.2.5.2-2
Note*: PREFSENS depends on the channel bandwidth as specified in section 6.3.1.

Table 6.2.5.2-2: Interfering signal for Intermodulation performance requirement

Interfering signal
E-UTRA centre frequency
channel offset from the
Type of interfering signal
bandwidth channel edge of the
[MHz] wanted signal
[MHz]
2.1 CW
1.4
4.9 1.4MHz E-UTRA signal
4.5 CW
3
10.5 3MHz E-UTRA signal
7.5 CW
5
17.5 5MHz E-UTRA signal
7.5 CW
10
17.7 5MHz E-UTRA signal
7.5 CW
15
18 5MHz E-UTRA signal
7.5 CW
20
18.2 5MHz E-UTRA signal

Table 6.2.5.2-3: Narrowband intermodulation performance requirement

Interfering Interfering RB
signal mean centre
E-UTRA
power[dBm] frequency offset
channel Wanted signal Type of interfering
from the
bandwidth mean power [dBm] signal
channel edge of
[MHz]
the wanted
signal [kHz]
-36 270 CW
1.4 PREFSENS + 14dB* 1.4 MHz E-UTRA
-36 790
signal, 1 RB**
-36 275 CW
3 PREFSENS + 14dB* 3.0 MHz E-UTRA
-36 790
signal, 1 RB**
-36 360 CW
5 PREFSENS + 14dB* 5 MHz E-UTRA
-36 1060
signal, 1 RB**
-36 415 CW
PREFSENS + 14dB*
10 5 MHz E-UTRA
(***) -36 1420
signal, 1 RB**
-36 380 CW
PREFSENS + 14dB*
15 5MHz E-UTRA signal,
(***) -36 1600
1 RB**
-36 345 CW
PREFSENS + 14dB*
20 5MHz E-UTRA signal,
(***) -36 1780
1 RB**
Note*: PREFSENS is related to the channel bandwidth as specified in section 6.3.1.
Note**: Interfering signal consisting of one resource block positioned at the stated offset.
Note***: This requirement shall apply only for a FRC A1-3 in [5] mapped to the frequency range
at the channel edge adjacent to the interfering signals

36
Release 13 37 3GPP TR 36.922 V13.0.0 (2016-01)

6.2.6 In-channel selectivity

Receiver in-channel selectivity requirement is a measure of the receiver ability to receive a wanted signal at its assigned
resource block locations in the presence of an interfering signal received at a larger power spectral density.

6.2.6.1 Analysis
For Home eNode B, the same method as that for MeNB can be used for defining this requirement. The UL signals are
just defined for 2 users, one being the "wanted" signal and the other one being the "interfering" signal at elevated power.
Regarding the interferer level, a 16QAM "interfering" signal is proposed 25dB above its noise floor to mask the impact
of receiver's own noise floor. The "wanted" signal was defined as a QPSK modulated FRC, for which ≥95% T-put
should be achieved in the presence of the interfering signal. The only difference between MeNB and HeNB is the power
level setting for wanted signal and interfering signal. Since the noise figure has been agreed as 13dB (8dB degradation
compared to MeNB), the wanted signal and interfering signal levels for home eNode B is shown in Table 6.2.6.1-1.

Table 6.2.6.1-1: RB allocations and power settings for wanted signal and interferer for HeNB

E-UTRA channel RBs RBs Wanted signal Interfering signal

BW [MHz] Wanted signal Interfering signal level [dBm] level [dBm]
1.4 3 3 -98.9 -79
3 9 6 -94.1 -76
5 15 10 -92.0 -73
10 25 25 -90.5 -69
15 25 25 -90.5 -69
20 25 25 -90.5 -69

6.2.6.2 Minimum requirement

The throughput shall be ≥ 95% of the maximum throughput of the reference measurement channel as specified in
Annex A of [5] with parameters specified in Table 6.2.6.2-1 for HeNB.

Table 6.2.6.2-1 E-UTRA Home BS in-channel selectivity

E-UTRA
Reference Wanted signal Interfering
channel Type of interfering
measurement mean power signal mean
bandwidth signal
channel** [dBm] power [dBm]
(MHz)
A1-4 in Annex 1.4 MHz E-UTRA
1.4 -98.9 -79
A.1 signal, 3 RBs
A1-5 in Annex 3 MHz E-UTRA
3 -94.1 -76
A.1 signal, 6 RBs
A1-2 in Annex 5 MHz E-UTRA
5 -92.0 -73
A.1 signal, 10 RBs
A1-3 in Annex 10 MHz E-UTRA
10 -90.5 -69
A.1 signal, 25 RBs
A1-3 in Annex 15 MHz E-UTRA
15 -90.5 -69
A.1* signal, 25 RBs*
A1-3 in Annex 20 MHz E-UTRA
20 -90.5 -69
A.1* signal, 25 RBs*
Note*: Wanted and interfering signal are placed adjacently around DC
Note**: the reference channel A1-x is defined in [5]

6.3 Performance requirement

Compared with macro eNodeB, HeNB is usually deployed for indoor scenarios. The propagation conditions will
contain more multi-paths with smaller multi-path delay due to the rich scattering characteristic of indoor environment.
In LTE systems, which is based on multiple carrier frequencies with CP to combat the multi-path fading, the indoor
demodulation performance can be expected at least as good as that under outdoor channel model. Moreover, from a
general point of view, HeNB can also be deployed for small-scale enterprise solutions. An outdoor hot spot model
without very large multi-path delay, e.g. EVA, should also be considered for demodulation performance.

37
Release 13 38 3GPP TR 36.922 V13.0.0 (2016-01)

Furthermore, UE attached to HeNB is usually considered to move at a speed no faster than 30km/h, which corresponds
to a maximum Doppler frequency of about 70Hz. Thus, it is feasible to define the HeNB performance requirements by
utilizing some specific macro eNodeB test cases with low speeds.

For multi-path fading propagation conditions shown in B.2 of TS36.104 [5], EPA and EVA model with a maximum
Doppler frequency no larger than 70Hz is considered for TD-LTE HeNB demodulation performance and the
performance requirements remain the same as that in TS36.104 [5] accordingly.

6.4 Synchronization requirement

6.4.1 Synchronization Accuracy

6.4.1.1 Synchronization error analysis

For LTE TDD, the inter-cell interferences of eNB (HeNB) to eNB (HeNB) and UE to UE are related to the cell
synchronization. In order to overcome the above interferences, strict synchronization is required. For HeNB, the
interference case of UE to UE at the uplink-to-downlink switch point is the crucial factor to the synchronization
requirement because of the limited coverage. At the switch point, two kinds of interferences should be taken into
account. One is MUE downlink disturbed by HUE uplink, the other is HUE downlink disturbed by MUE uplink.

If the HeNB coverage is up to tens meters, the synchronization error should be smaller than 1us. For the network
listening scheme, such strict requirement will increase the implementation difficulty and synchronization overhead.
However, as the minimum cyclic prefix of LTE is far larger than the sum of delay spread and propagation delay for
indoor scenarios, a little interference caused by inaccurate synchronization will not result in performance degradation.

For LTE, there are two kinds of cyclic prefix (CP), i.e. the normal CP and extend CP, with their periods listed in
Table.6.4.1.1-1.

Table 6.4.1.1-1 the CP period of LTE

Normal Extend

CP period 4.7/5.2us 16.7us

Since the most important application of HeNB focuses on the indoor scenarios such as home and office, ITU indoor
channel models [12] are taken into account for reference, in which the maximum delay spreads of LOS (light-of-sight)
and NLOS (non light-of-sight) scenarios are listed in Table.6.4.1.1-2.

Table 6.4.1.1-2 the delay spread of indoor channel

ITU InH LOS ITU InH NLOS

Maximum delay spread 0.13us 0.225us

In addition, the propagation delay is limited up to 1us because of the restricted transmission power and complex indoor
scenarios. Thus, there is

[ synchronization error ]HeNB =

[CP ]min − [ propagation delay ]max − [ delay spread ]max
= 4.7us − 1us − 0.225us
= 3.475us
Thus in many scenarios, a 3 us synchronization requirement can be adopted. The 3us requirement is also compatible
with the macro cell. But it is also important that practical synchronization schemes (including GPS, IEEE 1588v2, open
and closed loop network listening) are not excluded. Thus, if the HeNB derives its synchronization from a larger cell,
then the propagation distance is larger and a different synchronization requirement is required compared with that in
small cell.Therefore, it is important to have a synchronization requirement that is strict and practical.

Network listening is one essential practical scheme, as it works when GPS doesn't work (e.g. indoors) and IEEE 1588v2
is not available (e.g. with consumer-grade backhaul). Network listening can be performed in open loop or close loop
fashion. Based on some studies in RAN4, the advantages of open loop vs. closed loop are explained and it is concluded

38
Release 13 39 3GPP TR 36.922 V13.0.0 (2016-01)

that open loop network listening is essential for TD-LTE HeNBs. When synchronization is acquired using open loop
network listening, the synchronizing HeNB is automatically offset by the propagation delay compared to the donor eNB
or HeNB. Some requirements that take this fact into account are necessary. A 3us requirement for small cells is based
on a propagation delay of 1.67 us and an implementation margin of 1.33 us. The same margin can be used in the large
cell scenario as well. This will result in a synchronization requirement of 1.33 us + the propagation delay between the
HeNB and the donor cell.

It should be noted that the guard period in the DwPTS subframe should be chosen so as to accommodate the
propagation delay. The analysis in RAN4 shows that if the guard period is equal to twice the maximum propagation
delay, an additional timing advance can be used to prevent UE-UE interference. The following figure demonstrates this.
Note that un-accessed UEs can not know the additional timing advance, and therefore the UpPTS channel of un-
accessed UEs must be disturbed by the additional timing advance, but with a small additional timing advance, this issue
will be mitigated in some extent. The issue is specific to the open-loop scheme, FSS. For close-loop scheme, the
additional timing advance is unnecessary.

GP of MeNB

MeNB TX TX

MUE TX TX D TX
UL

DL
GP of HeNB TAoffset or UE rampup

BS ramp down
HeNB TX TX UE ramp down

Prop delay

HUE TX TX TX TX e_TA for UL-DL guard

GP of HeNB

HeNB with TX TX
UL e_TA

HUE with
UL e_TA
TX TX C TX TX

Figure 6.5-1 TDD HeNB Timing using Network

Figure Listening
6.4.1.1-1 (Void)and Extra Timing Advance

6.4.1.2 Synchronization requirement

The synchronization requirement for a HeNB is defined as the difference in radio frame start timing, measured at the
transmit antenna connectors, between the HeNB and any other HeNB or eNB which has overlapping coverage. The
synchronization requirement shall be set to 3 us in all cases, except when the HeNB gets its synchronization when
performing network listening off cells with propagation distance greater than 500m. This requirement shall apply
independent of the synchronization technique used (GPS, IEEE 1588 v2, Network Listening). In scenarios where
synchronization is obtained via network listening off cells with propagation distance greater than 500m, the
synchronization requirement shall be 1.33 μs plus the propagation delay between the HeNB and the cell selected as the
network listening synchronization source (e.g. when the propagation distance is 2.6km, the synchronization requirement
is 10 us). In terms of the network listening synchronization source selection, the best accurate synchronization source to
GNSS should be selected.

6.4.2 Techniques for Synchronization

Three synchronization techniques have been identified for HeNB synchronization.

39
Release 13 40 3GPP TR 36.922 V13.0.0 (2016-01)

GPS. If a HeNB contains a GPS receiver and can acquire the GPS synchronization signals, then GPS provides the most
accurate synchronization accuracy (on the order of 100ns). However, GPS receivers do not always work in some
important scenarios (e.g. indoors.)

IEEE 1588 v2. Under good backhaul conditions (e.g. operator controlled fiber / Ethernet), IEEE 1588 v2 can provide
sub-microsecond level accuracy. However, such good backhaul conditions may not always be possible. In particular
backhauls over cable and DSL modems have significant jitter and delay variations. Note that the upstream packet delay
δ1 is often not equal to the downstream delay δ2 creating an error of (δ1 – δ2)/2. This resulting error may be up to many
milliseconds, rendering IEEE 1588v2 restricted for the application of TD-LTE synchronization.

Network Listening. Network listening can be used in scenarios where GPS and IEEE 1588 v2 do not work. For this
reason, network listening is an essential synchronization scheme for TD-LTE HeNBs in those scenarios.

6.4.2.1 Synchronization using Network Listening

The technique in which a HeNB derives its timing from a synchronized eNB or HeNB (which in turn may be GNSS-
synchronized) is referred to here as "synchronization using network listening." A HeNB that uses network listening (say
HeNB1) may utilize a synchronization or reference signal from another eNB (say sync eNB) to derive its timing as in
Fig. 6.4.2.1-1(a). Such single hop synchronization for HeNB is the most common case under good macro coverage
based on analysis in RAN4. But when a HeNB can not acquire synchronization from a primary synchronization source
(an eNB or HeNB with GNSS synchronization) then multiple hops could be supported. This concept is illustrated in Fig.
6.4.2.1-1(b) where HeNB2 acquires synchronization from HeNB1 which in turn acquires synchronization from eNB.

In the case of multihop synchronization, the concept of synchronization stratum can be introduced. The synchronization
stratum of a particular HeNB is defined as the smallest number of hops between the HeNB and the GPS source. It
should be noted that the synchronization stratum of a particular HeNB is one greater than its donor (H)eNB, i.e., the
(H)eNB that it is tracking. In the figure below, sync eNB has stratum 0, HeNB1 has stratum 1 and HeNB2 has stratum 2.

Sync eNB HeNB2 Sync eNB HeNB2

HeNB1 HeNB1

(a) (b)

Fig. 6.4.2.1-1 Synchronization using Network listening

The HeNB may periodically track one or more signals from the donor cell (e.g. Primary and Secondary Synchronization
Signals, Common Reference Signal, Positioning Reference Signal) to maintain its synchronization. Of course, tracking
the PSS and SSS could come at the cost of some backward compatibility since a HeNB would need to shut down its
PSS/SSS transmission to monitor the PSS/SSS of the donor (H)eNB. Two fully backward compatible schemes for
tracking the Common Reference Signal (CRS) have been proposed, one that uses MBSFN subframes and one that uses
the guard period between DL and UL transmission. A description of these schemes is given in the section 6.5.3.1.2 and
6.5.3.1.3.

6.4.2.1.1 Interference Problems with Network Listening and Solutions

When a HeNB obtains synchronization through network listening, it has to stop transmitting and monitor the signals of
its donor (H)eNB, this process is susceptible to interference. In particular, cells that are in the vicinity of others using
network listening may not be able to receive the synchronization signals from a farther off cell due to strong
interference from these cells. This is shown in Fig. 6.4.2.1.1-1.

40
Release 13 41 3GPP TR 36.922 V13.0.0 (2016-01)

sync eNB Tx Tx Tx Tx … Tx Tx

HeNB1 Rx Tx Tx Tx … Rx Tx
Interference Interference
HeNB2 Tx Tx Rx Tx … Tx Tx

Fig. 6.4.2.1.1-1 Interference problem in network listening

One solution is to use appropriate DL Power Control to mitigate the interference from neighbour nodes when
synchronization tracking. The interference from neighbour cell will be controlled in an acceptable level, which could
ensure the network listening.

An alternate solution is to coordinate the tracking time between cells. Fig. 6.4.2.1.1-2 shows an example in which the
tracking times are coordinated among different nodes. Here HeNB1 tracks sync eNB without interference from HeNB2.
Additionally, HeNB2 tracks the synchronization signals from HeNB1. The results discussed in RAN4 show that
virtually all HeNBs can obtain synchronization via coordinated silence.

sync eNB Tx Tx Tx Tx … Tx Tx

HeNB1 Rx Tx Tx Tx … Rx Tx

HeNB2 Tx Rx Tx … Tx

Fig. 6.4.2.1.1-2 Interference problem mitigation using Coordinated Silence

Note that synchronization maintenance can be done at very low periodicity as the clock drifts are 250ppb or less. In
order to achieve satisfactory performance, the nodes must co-ordinate their silence periods, and utilize these
opportunities to achieve and maintain synchronization. Coordinated information should be conveyed to the cells for
synchronization tracking meanwhile these cells should have an initial common reference time, e.g. aligned SFN (system
frame number), to ensure the execution. This initial reference time including SFN could be obtained at HeNB bootup
by observing the time of the nearest cell, which may or may not be the same cell that the HeNB chooses to track later on.

6.4.2.1.2 MBSFN Subframe based Network Listening

The proposed scheme uses MBSFN subframes for tracking synchronization. An HeNB stops transmitting for a
subframe to track synchronization. To minimize the impact on UEs, the HeNBs declare this subframe to be an MBSFN
subframe. This method allows for multiple hops in the synchronization path. Also, all the nodes can track in a
coordinated fashion (all declaring MBSFN subframes at the same time), thus minimizing interference.

Fig. 6.4.2.1.2-1 Tracking using MBSFN Subframes

Furthermore, it ensures that the entire network uses the same synchronization source (e.g. GNSS) and that loops are not
created. This is because each HeNB declares its stratum as one greater than that of its donor (H)eNB. It should be noted

41
Release 13 42 3GPP TR 36.922 V13.0.0 (2016-01)

that the stratum number of a HeNB is self-configured, and that the HeNB tries to track the lowest available stratum
node. This in turn allows the HeNB to be as close to GNSS time as possible. Furthermore, the stratum number is a
dynamic quantity that could vary with changing RF conditions (if HeNB1 in the above example is turned off, then
HeNB2 could obtain synchronization via a different route, say eNB0  HeNB3  HeNB4  HeNB2, in which case it
would have a stratum number of 3.) A HeNB should preferably synchronize to the lowest possible stratum. A flow chart
to demonstrate deriving the stratum and using MBSFN subframes for tracking is given in the subsequent figure.

Fig. 6.4.2.1.2-2 HeNB Procedure for Synchronization using MBSFN Subframes

The overhead incurred by this scheme depends on the number of hops and would be equal to the number of hops times
one subframe in every 320 subframes. (320ms corresponds to the highest configurable periodicity of MBSFN
subframes). For a stratum-1 HeNB, the overhead is a little under 0.3%. It should also be noted that the MBSFN
subframe based method can be used for FDD as well for deriving frequency synchronization (and potentially time
synchronization if required in future releases).

6.4.2.1.3 TDD Special Subframe based Network Listening

To avoid asynchronous interference, another simple method is introduced for TDD system to achieve HeNB
synchronizing to macro layer eNB, which there is no need HeNB install satellite receiver. In this solution, Home eNB
and macro layer eNBs utilize different special subframe configuration, macro layer configure with more OFDM
symbols in DwPTS, and Home eNB with less OFDM symbols in DwPTS, so Home eNB can utilize the GP to track
macro layer eNB common reference signal (CRS) in DwPTS without additionally impact on its normal transmission,
and CRS tracing can be done every radio frame to generate a statistic tuning value, which ensure more robust
synchronization. Meanwhile considered HeNB is mainly used for providing high data rate service, the DL resource is
relative not stringent to provide such overhead for robust synchronization.

With this solution HeNB can only read CRS for synchronization, however this should be enough. When HeNB is power
on, HeNB may follow the UE cell search process and get the accurate synchronization from the macro eNB while is
assumed as accurate synchronization resource with satellite receiver. And then symbol timing, radio frame timing and
eNB cell ID can be get by HeNB, which enable the HeNB conduct the aforementioned CRS based synchronization
tracking procedure, which is mainly a process to track the synchronization on a finer time scale.

When HeNB is operating, its location is stable, so there is no need to always repeat the cell search process to get the
timing. Only tracking the CRS periodically to maintain the synchronization with macro layer is enough for HeNB.

42
Release 13 43 3GPP TR 36.922 V13.0.0 (2016-01)

There are two CP lengths defined in TS 36.211, so the analysis is provided separately for the two cases.

1. Normal CP case

CRS on antenna port 0 and antenna port 1 are located in 1st and 5th OFDM symbol of each slot. Macro layer eNB can be
configured unifiedly with more DwPTS symbols (i.e. config1, 2, 3, 4, 6, 7, 8, detailed configurations are shown in
Table 6.4.2.1.3-1). HeNB use other different configuration to pair with macro layer configuration, such as config 0 or 5,
in these configurations, HeNB only transmit 3 OFDM symbols. When HeNB track the timing, after its DwPTS
transmission finish, HeNB transit to receiver state, normally the HeNB eNB DL->UL switching time is less than 15us,
one symbol is enough for the switching, HeNB will receive the CRS from the 5th OFDM symbol. Also by configuring
both macro eNB (config4) and HeNB (config2) with more symbols in DwPTS, the user data transmission is improved,
such as HeNB can start to receive CRS from the 12th symbol of the special subframe.

Example of Macro eNB and HeNB configuration refer to Figure 6.4.2.1.3-1.Considering HeNB DL-> UL switching
time, HeNB can receive macro eNB CRS in 5th OFDM symbol successfully.

Macro eNB config 1 R0/R1 R0/R1 R0/R1

HeNB config 0 R0/R1 R0/R1 R0/R1

DwPTS symbol GP symbol UpPTS symbol Symbol for switching

R0/R1: CRS in port 0 or 1

Figure 6.4.2.1.3-1: Normal CP case, Macro eNB and HeNB DwPTS/GP/UpPTS configuration

Table 6.4.2.1.3-1: DwPT/GP/UpPTS configuration

Config Normal CP Extended CP

DwPTS GP UpPTS DwPTS GP UpPTS
0 3 10 1 3 8 1
1 9 4 1 8 3 1
2 10 3 1 9 2 1
3 11 2 1 10 1 1
4 12 1 1 3 7 2
5 3 9 2 8 2 2
6 9 3 2 9 1 2
7 10 2 2
8 11 1 2

2. Extended CP case

CRS on antenna port 0 and antenna port1 are located in 1st and 4th OFDM symbol of each slot. Macro layer eNB can
configure with more DwPTS symbols (i.e. config1, 2, 3, 5, 6). HeNB use different configuration, such as config 0 or 4,
in these configurations, HeNB only transmit 3 OFDM symbols. When HeNB track the timing, after it's DwPTS
transmission finish, HeNB transit to receiver state, normally the HeNB eNB DL->UL switching time is less than 15us,
one symbol is for the switching, HeNB will receive the CRS from the 7th OFDM symbol. Also by configuring both
macro eNB (config3) and HeNB (config1) with more symbols in DwPTS, the user data transmission is approved, such
as HeNB can start to receive CRS 10th symbol of the special subframe.

Example of Macro eNB and HeNB configuration refer to Figure 6.4.2.1.3-2. Considering HeNB DL-> UL switching
time, HeNB can receive macro eNB CRS in 7th OFDM symbol successfully.

43
Release 13 44 3GPP TR 36.922 V13.0.0 (2016-01)

Macro eNB config 1 R0/R1 R0/R1 R0/R1

HeNB config 0
R0/R1 R0/R1

DwPTS symbol GP symbol UpPTS symbol Symbol for switching

R0/R1: CRS in port 0 or 1

Figure 6.4.2.1.3-2: Extend CP case, Macro eNB and HeNB DwPTS/GP/UpPTS configuration

This solution based on macro cell layer and HeNB layer deployed different DwPTS/GP/UpPTS configurations. As for
HeNB configuration, that can be fixed by operator before distribution. There is no impact on the air interface
specifications. For two layers cell, only special subframe configuration is different, there is no interference issue
between macro layer eNB and HeNB or related connected UE-UE. Also there is no backward compatibility issue for
Rel8 eNB and UE, only requirement is to home eNB which need odell the cell search and track macro layer CRS
function. Tracking period can be set by the HeNB, for the oscillator frequency stability is affected by the ambient
temperature.

The feasibility of this scheme is relative to the macro cell configuration. If macro layer eNBs are configured with max
GP, there is no way that a HeNB could use macro common reference signals for synchronization tracking.

For the scenario that HeNB are not able to synchronize directly to an eNB that is GNSS-synchronized, utilizing the
special subframe configuration pairs, this solution also can fulfill 2 hops synchronization in some configurations. Take
extended CP case for example, macro cell can be set with configuration 3, the first hop HeNB is set with configuration
1 or 5 and the second hop HeNB is set with configuration 0 or 4.

6.4.2.1.4 Indication of Stratum Level and Synchronization Status

The HeNB should be aware of its neighbours synchronization hierarchy (stratum info), and then correspondingly decide
its own stratum number. Also, the HeNB needs to let others know its own synchronization status and stratum info. Two
solutions are proposed to fulfil this function as below.

Note that while the solutions are described in the context of the MBSFN-subframe based scheme, the use of these
solutions for other schemes (including schemes not listed in the TR) is not precluded.

RAN4 endorses both backhaul signalling and blind detection schemes for indication of stratum level and
synchronization status, and their adoption depends on the operator deployment choice.

It is up to the operator to choose either backhaul signalling or blind detection depending on the deployment.

6.4.2.1.4.1 Stratum Indication Using Backhaul Signalling

The optional backhaul signalling of time synchronization status and stratum level are TDD HeNB specific. Optional
only means that it is up to the operator to decide whether to use the signalling or not depending on the deployment.

An HeNB can get information of neighboring eNB's time synchronization status and stratum level over the backhaul by
using the S1-AP eNB configuration transfer procedure and the S1-AP MME configuration transfer procedure.

The HeNB_1 initiates S1-AP eNB configuration transfer procedure by sending the MME the eNB CONFIGURATION
TRANSFER message containing the target eNB ID and Time Sync Info request. The MME forwards the request to the
target eNB with the MME CONFIGURATION TRANSFER message.

44
Release 13 45 3GPP TR 36.922 V13.0.0 (2016-01)

When the (H)eNB_2 receives the MME CONFIGURATION TRANSFER message with Time Sync Info request, it
replies to the MME with eNB CONFIGURATION TRANSFER message containing its stratum level and sync status.
The MME forwards the received information to the HeNB1.

HeNB_1 MME (H)eNB_2

1. ENB CONFIGURATION
TRANSFER
(Time Sync Info request) 2. MME CONFIGURATION
TRANSFER
(Time Sync Info request)

3. ENB CONFIGURATION
TRANSFER
4. MME CONFIGURATION (stratum level, sync status)
TRANSFER
(stratum level, sync status)

Figure 6.4.2.1.4.1-1 Stratum Indication by Backhaul Signalling

6.4.2.1.4.2 Stratum Indication by Blind Detection

Blind detection, as an alternative mechanism, is also proposed in RAN4, which can fulfil the convey requirement for
stratum info and synchronization status without signalling when the OAM configures or all HeNBs embedded pre-
configure the same muting places (e.g. MBSFN subframes) for a given value of stratum and status, while configures
different muting places for other values of stratum and status. Optional OAM signalling of MBSFN subframes as a
function of stratum is available. If with all HeNBs embedded pre-configurion, then it is no need to send the OAM
signalling. For blind detection, all HeNBs should well know the mapping relationship of each stratum and its muting
place, here one instance of the muting place can be subframes declared as MBSFN for this stratum. One mapping
example is illustrated on Fig 6.4.2.1.4.2-1, that HeNB stratum 1 will trace CRS in SF#2 of RF#1, HeNB stratum 2 will
trace CRS of stratum 1 in SF#2 of RF#2, also mute to avoiding interference to HeNB of stratum 1 (this muting could be
omitted if power control is appropriately utilized then interference will be mitigated); similar ruling is taken to the
following strata.

HeNB do blind detection for the existence of CRS on muting place for all possible stratum (normally on booting stage)
and contrast the mapping table to recognize the strata of its surrounding base stations, and basing some strategy to
decide its synchronization source, thus also decide its own stratum and muting place.

On normal working, HeNB execute network listening on its specific muting place according stratum, while that is also
indicating its stratum and synchronization info for new booted neighbour who is doing blind detection.

Periodically, this HeNB may reserve all muting places for one or several rounds, and detect whether any change occurs
which may impact its stratum, e.g. synchronization source node shutting down or new node booting up providing lower
stratum than current source, and adapt its own stratum accordingly. Non-GNSS synchronized stratum and GNSS
synchronized stratum can be differentiated if different muting places are used for the two which can ensure smooth
stratum change between both types.

45
Release 13 46 3GPP TR 36.922 V13.0.0 (2016-01)

Fig. 6.4.2.1.4.2-1 Explanation for blind detection on stratum info and synchronization status

46
Release 13 47 3GPP TR 36.922 V13.0.0 (2016-01)

Booting Up or
Periodic timer Expire?
no
yes

Do not transmit on
muting place for all
possible strata

Receiving Monitor set of selected

CRS cells’ CRS in these
muting places

Analysis surrounding cells’

stratum and status info, Do not transmit
Define its (new) own on muting place
stratum of its own
stratum

Fig. 6.4.2.1.4.2-2 HeNB procedure for blind detection on stratum info and synchronization status

6.4.2.1.5 Scheme Comparison

A brief comparison of the proposed schemes is shown in Table 6.4.2.1.5-1.

Table 6.4.2.1.5-1: Comparisons for different network listening schemes

Network Listening
Scheme 1 Scheme 2
schemes

 Use MBSFN subframes for  Use DwPTS for tracking

Principle of the
tracking CRS of donor CRS of donor (H)eNB
scheme
(H)eNB
Meets requirements Meets requirements
Performance
Provides flexible overhead- CRS tracing can be done every
(e.g. synchronization
tracking periodicity tradeoff Radio Frame, which ensure
accuracy, speed, etc)
robust synchronization
0.3% for stratum-1 nodes when Maximum 12.86% with 2 switch
HeNB Overhead
using the lowest periodicity of point per RF
(e.g. OFDM symbols only 1 MBSFN for tracking per Minimum 1.43% with 1 switch
per [320ms]) 320ms point per RF
1, up to 2 with some cases of
Number of multi- 4 special SF configuration (e.g.
hops supported Normal CP SSF Conf.4 Conf.
2 Conf. 5)
Compatibility and Fully backward compatible Fully backward compatible
impacts on current
network

47
Release 13 48 3GPP TR 36.922 V13.0.0 (2016-01)

Optional backhaul signalling will No extra signalling is needed

be specified, and either backhaul when supporting only single
Impacts on
signalling or blind detection hop
specifications
scheme used depending on
operator deployment.
Could be used by a HeNB Could be used by a HeNB
Others capable of either FDD or TDD capable of TDD mode
mode

RAN4 endorses both the GP based solution and the MBSFN subframe based solution.

7 Interference control

7.1 HeNB measurements

Several types of measurements that HeNB can perform are listed in the following subsections. The objectives of the
HeNB measurements are

- to provide sufficient information to the HeNB for the purpose of interference mitigation

- to provide sufficient information to the HeNB such that the HeNB coverage can be maintained.

According to the measurement type, some of these measurements can be collected through Connected Mode UEs
attached to the HeNB or via a DL Receiver function within the HeNB itself. Such DL receiver function is also called
Network Listen Mode (NLM), Radio Environment Measurement (REM) or "HeNB Sniffer".

These measurements can also be used during the HeNB self-configuration process.

7.1.1 Measurements from all cells

This section identifies the potential measurements performed by HeNB during self-configuration and normal operation.
Based on the measurements in Table 7.1.1-1, the HeNB can obtain useful information from its surrounding cells for
purposes such as interference management.

Table 7.1.1-1: HeNB measurements from surrounding cells

Measurement Type Purpose Measurement Source(s)

Calculation of UL interference towards HeNB

Received Interference Power HeNB UL Receiver
(from MUE)

HeNB could use the Received Interference Power measurement to monitor the uplink interference. For example, a
Received Interference Power measurement value larger than a pre-defined threshold would mean that at least an MUE
which is interfered by a HeNB is close to the HeNB and that the MUE's Tx power would cause significant interference
towards the HeNB. This measurement value may be used in calculating path loss between the HeNB and the MUE
assuming that a single MUE dominates the interference. It is also preferable for the HeNB to distinguish between UL
interference from the MUE and wanted signals from HUEs to improve the accuracy of interference measurement.

7.1.2 Measurements to identify surrounding cell layers

This section identifies the potential measurements performed by HeNB during self-configuration and normal operation.
Based on the measurements in Table 7.1.2-1, the HeNB can obtain useful information to identify the layer of its
surrounding cells and indirectly identifies other HeNBs nearby for purposes such as mobility handling.

48
Release 13 49 3GPP TR 36.922 V13.0.0 (2016-01)

Table 7.1.2-1: HeNB measurements from surrounding cells

Measurement Type Purpose Measurement Source(s)

Cell reselection priority Distinction between cell types based on
HeNB DL Receiver
information frequency layer priority

Distinction between cell layers based on

CSG status and ID HeNB DL Receiver
CSG, and self-construction of neighbour list,

7.1.3 Measurements from macro cell layer

This section identifies the potential measurements performed by HeNB during self-configuration and normal operation.
Based on the measurements in Table 7.1.3-1, the HeNB can obtain useful information from its surrounding macro cells
for purposes such as interference management.

Table 7.1.3-1: HeNB measurements from surrounding macro cells

Measurement Type Purpose Measurement Source(s)

Calculation of co-channel DL interference towards
macro UEs (from HeNB)
Calculation of co-channel UL interference towards
macro layer (from HUEs) HeNB DL Receiver
RSRP Calculation of co-channel UL interference towards HUE
HeNB (from MUEs) based on estimated MUE Tx MUE (in case of hybrid cell)
power
Determine coverage of macro cell (for optimization
of hybrid cell configuration)
HeNB DL Receiver
Determine quality of macro cell (for optimization of
Co-channel RSRQ HUE
hybrid cell configuration)
MUE (in case of hybrid cell)
Reference Signal Transmission
Estimation of path loss from HUE to MeNB HeNB DL Receiver
Power
Allow HeNB to Instruct UEs to measure specific HeNB DL Receiver
Physical + Global Cell ID cells. HUE
Allow UE to report discovered cells to HeNB.

If a HeNB has receiver capability, then it is able to measure the received CRS Êc, measured in dBm, which is the
Reference Signal Received Power per resource element present at the Home BS antenna connector for the Reference
Signal received on the co-channel. For CRS Êc determination, the cell-specific reference signal R0 according TS 36.211
shall be used. If the HeNB can reliably detect that multiple TX antennas are used for transmission on the co-channel, it
may use the average in W of the CRS Êc on all detected antennas. On start-up, the HeNB can measure the CRS Êc
power from the most dominant co-channel deployed macro cell.

Table 7.1.3-2: HeNB measurements from surrounding macro cells

Measurement Type Purpose Measurement Source(s)

Measurement is used to determine whether HeNB is

Co-channel received CRS Êc
close to dominant Macro cell, or whether it is close HeNB DL Receiver
(measured in dBm)
to macro-cell-edge border.

7.1.4 Measurements of other HeNB cells

This section identifies the potential measurements performed by HeNB during self-configuration and normal operation.
Based on the measurements in Table 7.1.4-1, the HeNB can obtain useful information from its adjacent HeNBs for
purposes such as interference management.

49
Release 13 50 3GPP TR 36.922 V13.0.0 (2016-01)

Table 7.1.4-1: HeNB measurements from adjacent HeNBs

Measurement Type Purpose Measurement Source(s)

Calculation of co-channel DL interference towards
neighbour HUEs (from HeNB) HeNB DL Receiver
Co-channel RSRP
Calculation of co-channel UL interference towards HUE
neighbour HeNBs (from HUEs)
Reference Signal Transmission Estimation of path loss from HUE to HeNB HeNB DL Receiver
Power

Physical + Global Cell ID Allow HeNB to Instruct UEs to measure specific HeNB DL Receiver
cells HUE
Allow UE to report discovered cells to HeNB.

7.2 HeNB self-configuration

7.2.1 Information Exchange between eNBs and HeNBs
The provision of information exchange between eNBs ↔ HeNBs and HeNBs ↔ HeNBs has potential benefits in
allowing HeNBs to take account of uplink and downlink conditions at nearby eNBs and HeNBs when configuring
power and/or resource blocks to use in uplink and downlink.

We consider several relevant metrics to compare these approaches:

(1) Latency: It was recognized in several contributions that a reliable low latency scheme is desirable for
interference management. In RAN4, it was discussed that the adaptation of HeNB parameters could be relatively
slow, such that changes in interference/loading at eNB are not tracked on a sub-frame by sub-frame, or frame by
frame, but rather more slowly as the traffic load varies on the eNB. Simulation results in RAN4 showed that with
50ms latency such relatively slow adaptation still offers significant performance benefits. Similarly simulation
results also showed significant performance benefits at comparable latencies. Further benefits can be obtained by
faster interference coordination, especially in the case of bursty traffic.

(2) Scalability and Complexity: It is desirable to have the network complexity scale in a manageable manner with
increasing number of HeNBs, UEs etc. Furthermore, different approaches are expected to have different
implementation impacts at different network entities (eNB, MME etc.).

(3) Overhead: The signalling overhead for exchanging interference management messages (for both the backhaul
and Over-the-Air methods discussed subsequently) should be small.

Possible approaches for performing the information exchange are illustrated in Figure 7.2.1-1 and their potential
benefits and drawbacks are described in the following. Flexible operations should be allowed to choose one or
combination of information exchange approaches in HeNB deployment.

50
Release 13 51 3GPP TR 36.922 V13.0.0 (2016-01)

HeNBs in Information over the NW

condominium
HeNB GW

MeNB Information over the air

HUE HeNB
HeNB HeNB
HeNB optimization over the NW HeNB optimization over the air

Figure 7.2.1-1: Illustration of information exchange for Over-the-Air and Network based approaches

Option1. Over-the-air information, direct eNB to HeNB

Potential benefits of this approach include:

• No impact to network load.

• Low latency for information signalled from eNB to HeNBs.

• Predictable timeline (independent of backhaul conditions), can be used for coordinated scheduling/transmission.

While this approach may offer low latency, there are several drawbacks:

• The eNB may not always be visible from the HeNB, even though there are victims requiring protection.

• For some advanced approaches for managing interference, it may be desirable to send different information to
different groups of eNBs or HeNBs. An over the air broadcast would preclude such operation.

• The downlink would need to be interrupted whenever information is read over the air.

• Requires changes to eNB implementation

Option2. Over-the-air information, (H)eNB to HeNB via UE

For the DL, a victim UE forwards interference coordination related information from its serving (H)eNB to the
aggressor HeNB.

For the UL: An aggressor UE forwards interference coordination related information from the victim (H)eNB to its
serving (H)eNB.

Potential benefits of this approach:

• The downlink would not need to be interrupted to receive information over the air.

• Lower latency compared to backhaul solutions (i.e. Option3 and 4) (higher latency relative to Option 1).

• Predictable timeline (independent of backhaul conditions), can be used for coordinated scheduling/transmission.

• Different information can be sent to different HeNBs

Potential drawbacks to this approach:

• Rel8 UEs can't be used to relay the messages.

• Requires changes to (H)eNB implementation

51
Release 13 52 3GPP TR 36.922 V13.0.0 (2016-01)

• Can increase the number of UEs that need to be handled by the HeNB.

Option3. X2 based interface between eNB and HeNB, and between HeNBs

The potential benefits of this approach:

• Higher accuracy of information received at destination than the Over-the-Air approach

• Different information can be sent to different groups of eNBs or HeNBs

The potential drawbacks of this approach:

• The eNB may have large numbers of HeNBs within its coverage area which potentially means the macro would
need to deal with many messages to/from HeNBs. Ways in which this could be mitigated could be considered by
the relevant working groups for further study e.g. X2 could be between macro eNBs and HeNBs via HeNB
gateways only, with the HeNB gateways performing a distribution/aggregation function towards the HeNBs. To
reduce the complexity further the set of supported X2AP procedures could be limited, e.g. no handover over X2,
and only sending Load Indication (OI, HII, RNTP) in the direction macro eNB to HeNB.

• Potentially large latency.

Option4. S1 based interface between eNB and HeNB, and between HeNBs

In some cases it is likely that direct physical links would not exist between (H)eNBs and HeNBs, and as such X2 would
be a logical interface sharing a similar physical path to S1. With this in mind it could be argued that the information
exchange could be made over S1 instead of X2. If compared to the X2 based approach there are some potential benefits
to this approach:

• Higher accuracy of information received at destination than the Over-the-Air approach

• Different information can be sent to different groups of eNBs or HeNBs

• S1 signalling interface already exist in the current specifications

Potential drawbacks of this approach:

• Increased functionality and processing load at the MME.

• Increased latency

• Lack of alignment between eNB↔eNB, eNB↔HeNB and HeNB↔HeNB SON/interference management.

• Potential lack of alignment with likely future evolutions of interference management in Release 10 and beyond
(assuming that these are less likely to be based on S1)

7.3 Uplink interference control

7.3.1 Control Channel Protection

7.3.1.1 HeNB Uplink Control Channel Protection

In the uplink, physical uplink control channel (PUCCH) interference from

• HUE (aggressor) to macro-eNB (victim),

• MUE (aggressor) to HeNB (victim), and

52
Release 13 53 3GPP TR 36.922 V13.0.0 (2016-01)

• HUE (aggressor) to HeNB (victim)

can be mitigated by enabling orthogonal transmissions. Uplink control signalling (PUCCH, CQI) reliability can be
maintained for both HeNBs and macro-eNBs by making use of PUCCH offsets for enabling orthogonal PUCCH
assignments between the HeNB and macro-eNB users. For PUCCH transmissions, over-provisioning can be made use
of to ensure orthogonality of control channels between a HeNB UE and a macro-eNB UE as shown in Figure 7.3.1.1-1.
It is possible to employ this method for Release-8 UEs without changing the physical layer design or RAN2 signalling.

macro-eNB control macro-eNB control

HeNB control HeNB control

HeNB control HeNB control

macro-eNB control macro-eNB control

Figure 7.3.1.1-1 UL control interference mitigation by PUCCH orthogonalization

7.3.1.2 Signalling offset over the backhaul

It would be desirable for the macro-eNB to signal an offset to all HeNBs within its coverage area in order that
transmissions from UEs connected to HeNBs do not cause interference at the macro-eNB receiver (e.g., a HeNB
deployed in close range of a macro-eNB). Conversely, a macro-eNB UE that is at the cell edge and therefore
transmitting close to its maximum transmit power can interfere severely with a HeNB UE and the signalling offset can
be made use of to mitigate interference. Alternately, a HeNB gateway can signal over S1, the offsets that each HeNB
should use, thus providing the capability of configuring orthogonal PUCCH transmissions in neighboring HeNBs
thereby avoiding HeNB (aggressor) to HeNB (victim) interference on the uplink.

One option for the HeNBs is to not allocate PUCCH resources on edge RBs as shown in Fig. 7.3.1.1-1 using over-
provisioning. A typical macro-eNB deployment is likely to have PUCCH transmission on the band-edges to maximize
the number of contiguous RBs that can be allocated to PUSCH. However, unlike macro-eNBs, utilizing the full uplink
bandwidth may not be critical for HeNBs as they serve only a few users at a time. Therefore, the PUCCH resources in
HeNBs can be "pulled" inward. The edge RBs not used by the HeNBs can be used by the macro-eNB for PUCCH for
its UEs. Also, the macro-eNB, being aware of the RBs used by HeNBs in its coverage area, can schedule some users
(e.g. UEs close to the macro and not near any HeNB) on RBs that overlap with HeNB UE PUCCH region. This results
in reduced interference from macro-eNB UEs to HeNB UE PUCCH.

7.3.2 Smart Power Control based on Path Loss to Worst Victim Macro
eNodeB
Interference from the Home UE (HUE) to the Macro eNodeB (MeNB) is particularly significant if the HUE is located
close to the MeNB. On the other hand, an indoor HUE near its serving HeNB and far from the MeNB may be harmless.
As discussed in RAN4, the HUE transmission power should be controlled based on path loss (PL) from the HUE to its
worst victim MeNB (i.e. nearest neighbour MeNB).

The PL from HUE to MeNB can be estimated from HUE measurements of Reference Signal Received Power (RSRP)
and MeNB Reference Signal (RS) Transmission (Tx) power. HeNB might know MeNB RS Tx power by means of

53
Release 13 54 3GPP TR 36.922 V13.0.0 (2016-01)

decoding the variable "referenceSignalPower" in System Information Block Type2 (SIB2) message broadcasted from
MeNB.

During this work item, such smart power control methods are proposed and their performance gain is investigated. The
methods are as follows.

7.3.2.1 Power Cap Method

In this method, the maximum transmission power density (i.e. power cap) of HUE is restricted based on the interference
rise at MeNB. The power cap is calculated as the function of PL from the HUE to its worst victim MeNB. The HUE is
power-controlled based on PL to its serving HeNB, up to the level of the power cap.

Simulation results have been generated for an urban deployment model with varying HeNB density and for either full
buffer or bursty traffic based on FDD. Similar performance trends also apply to TDD. The results are assume a fixed
power cap of either 0.2 dB (labelled "tight") or 7 dB (labelled "loose", and it should be noted that this is a very loose
cap, for which in practice the home UE power will likely be set considering coverage requirements of the HeNB alone
rather than also considering interference to the macro layer).

7.3.2.1.1 Simulation Assumptions

The simulation parameters largely follow the assumption in section 6.2.1.1.1 with the following specific parameters.

Table 7.3.2.1.1-1: Simulation Parameters

Parameter Assumption
Deployment Urban signalling
Macro layer has 7 sites (21 sectors) with wrap-around,
500m ISD.
0% (urban) of home UEs are outdoors and 20% of macro
UEs are outdoors.
Number of macro UEs per sector 20
Exterior wall loss 20dB
Shadowing correlation (one BS to multiple UEs) Correlated shadowing
Macrocell power uplink control Max power based on limiting noise rise to macro
neighbours
Femtocell uplink power control Max power based on limiting noise rise to macro
neighbours (a similar approach to that described in [7]
section 7.3.1 for WCDMA).
Link to system mapping Per sub-carrier capacity approach
Scheduler Frequency selective/Proportional fair
Traffic model Full buffer or Bursty
In the case of bursty traffic being modelled, 70% of UEs
use the bursty traffic model, the remaining UEs are full-
buffer.
Apartment block model Dual stripe, 6 floors (=240 apartments), one "dual stripe"
randomly dropped per macro sector. A variable probability
of having active femto in each apartment.
Pathloss model Full (rather than simplified) model

7.3.2.1.2 Simulation Results

Figure 7.3.2.1.2-1 below shows the average macrocell sector throughput as a function of the probability that there is an
active HeNB in an apartment. Results are shown for two values of the target maximum "noise rise" that the home UE
should generate at the macro eNB ("tight" and "loose"). It can be seen that with a low density of active HeNBs the
"loose" approach provides adequate protection whereas at higher densities the "tight" cap is appropriate. This goes for
both the full buffer and the bursty traffic models.

54
Release 13 55 3GPP TR 36.922 V13.0.0 (2016-01)

Macro Layer Uplink, Av. Sector Throughput

12000
full buffer loose
full buffer tight
11000 bursty loose
bursty tight

10000

Throughput (kb/s) 9000

8000

7000

6000

5000
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-1: Macrocell uplink average sector throughput

Figure 7.3.2.1.2-2 below shows the cell edge (5 percentile) macro user throughput as a function of the probability that
there is an active HeNB in an apartment. Again it can be seen that with a low density of active HeNBs the "loose"
approach provides adequate protection whereas at higher densities the "tight" cap is appropriate. This goes for both the
full buffer and the bursty traffic models.

Macro Layer Uplink, 5 percentile User Throughput

300
full buffer loose
full buffer tight
bursty loose
250
bursty tight

200
Throughput (kb/s)

150

100

50

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-2: Macrocell uplink 5 percentile user throughput

Figure 7.3.2.1.2-3 below shows the mean Interference over Thermal (IoT) at the macro eNB. It can be seen that the IoT
is controlled more with the "tight" cap particularly at high HeNB densities.

55
Release 13 56 3GPP TR 36.922 V13.0.0 (2016-01)

Macro Layer Uplink, Av. IoT (dB)

13
full buffer loose
full buffer tight
bursty loose
12
bursty tight

11

Av. IoT (dB)

10

7
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-3: Macrocell Interference over Thermal

Figure 7.3.2.1.2-4 below shows the average HeNB sector throughput as a function of the probability that there is an
active HeNB in an apartment. It can be seen that the "loose" cap results in a higher throughput. This goes for both the
full buffer and the bursty traffic models.

4
x 10 Femto Layer Uplink, Av. Cell Throughput
2
full buffer loose
1.8 full buffer tight
bursty loose
1.6 bursty tight

1.4
Throughput (kb/s)

1.2

0.8

0.6

0.4

0.2

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-4: Femtocell uplink average sector throughput

Figure 7.3.2.1.2-5 below shows the cell edge (5 percentile) home user throughput as a function of the probability that
there is an active HeNB in an apartment. Again it can be seen that the "loose" cap results in a higher throughput.

56
Release 13 57 3GPP TR 36.922 V13.0.0 (2016-01)

Femto Layer Uplink, 5 percentile User Throughput

16000
full buffer loose
full buffer tight
14000
bursty loose
bursty tight
12000

10000
Throughput (kb/s)

8000

6000

4000

2000

0
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-5: Femtocell uplink 5 percentile user throughput

Figure 7.3.2.1.2-6 shows the mean Interference over Thermal (IoT) at the HeNB.

Femto Layer Uplink, Av. IoT (dB)

4

full buffer loose

full buffer tight
2
bursty loose
bursty tight

-2
Av. IoT (dB)

-4

-6

-8

-10
0 0.05 0.1 0.15 0.2 0.25 0.3 0.35 0.4 0.45 0.5
Prob. active femto in apt.

Figure 7.3.2.1.2-6: Femtocell Interference over Thermal

7.3.2.1.3 Discussion of Results

For low densities of HeNB a "loose" power cap is sufficient which allows higher HeNB throughputs than the "tight"
power cap which is required for higher HeNB densities.

7.3.2.2 Power Control based on PL from HUE to its serving HeNB and PL from HUE
to its worst victim MeNB
The UE specific term of the transmission power density PO_PUSCH should be defined as the function of PL from HUE
to its serving HeNB ( PLHUE− HeNB ) and PL from HUE to its worst victim MeNB ( PLHUE− MeNB ) because the uplink
transmission power is explicitly defined as the form using PL from UE to its serving eNodeB in the current

57
Release 13 58 3GPP TR 36.922 V13.0.0 (2016-01)

specification [13]. For example, the power control where the UE specific term of PO_PUSCH is set to
PLHUE − MeNB − a × PLHUE − HeNB + interference_rise_at_MeNB (in dB) corresponds to the power cap method (The
term − α × PLHUE− HeNB is cancelled by path loss compensation term and the parameter α is path loss compensation
coefficient [13]). In general, the UE specific term of PO_PUSCH might be non-decreasing function of PLHUE− MeNB and
the dependency of PLHUE− HeNB is implementation issue.

One realization of such power control is proposed during this work item; PL difference based power control. In this
method, the UE specific term of PO_PUSCH is defined as the non-decreasing function of PL difference
∆PL = PLHUE − MeNB − PLHUE − HeNB (in dB). The explicit form of the UE specific term of PO_PUSCH is also given in
RAN4.

7.3.2.2.1 Simulation Assumptions

The simulation parameters largely follow the suburban model with the following specific parameters. The following
simulation is performed based on FDD. Similar performance trends also apply to TDD.

Table 7.3.2.2.1-1: Simulation Parameters

Parameter Assumption
Deployment Suburban model
Macro layer has 7 sites (21 sectors) with wrap-around,
500m ISD.
10% of home UEs are outdoors and all macro UEs are
indoors.
Number of macro UEs per sector 10
Exterior wall Loss 20dB
Shadowing correlation (one BS to multiple UEs) Correlated shadowing
Macrocell power uplink control Closed loop ICIC based on overload indicator, targeting the
IoT value to 10 dB
Femtocell uplink power control PL difference based TPC and
FPC (for comparison)
Link to system mapping EESM, same β value for all MCS
Scheduler Frequency selective / Proportional fair
Traffic model Full buffer
Pathloss model Full model

7.3.2.2.2 Simulation Results

Figure 7.3.2.2.2-1 and Figure 7.3.2.2.2-2 show the MUE and HUE throughputs for various HeNB densities, which is the
number of HeNB per macro sector. The power control based on PL difference (PL-diff.) and conventional fractional
power control (FPC) are compared.

Figure 7.3.2.2.2-1 indicates that the PL difference based power control mitigates the degradation of MUE throughput
than FPC. Figure 7.3.2.2.2-2 (right) shows the PL difference based power control can keep the HUE average throughput
at the same level of FPC. Its cost is the degradation of HUE 5 percentile throughput as shown in Fig. 7.3.2.2.2-2 (left).
In the suburban model with 10 % outdoor HUE, the HUE that is correspond to HUE 5 percentile throughput is mainly
located outdoors.

58
Release 13 59 3GPP TR 36.922 V13.0.0 (2016-01)

1000
1000
FPC 900 FPC
900 PL-diff.

MUE throughput (5 %ile) [kbps]

PL-diff.
800
MUE throughput (Ave.) [kbps]

800
700
700
600
600
500
500
400
400

300 300

200 200

100 100

0 0
0 5 15 25 0 5 15 25
HeNB density HeNB density

Figure 7.3.2.2.2-1: MUE throughput (Left: Average, Right: 5 percentile)

12000 12000
FPC FPC
11000 PL-diff. 11000 PL-diff.
10000 10000
HUE throughput (5 %ile) [kbps]
HUE throughput (Ave.) [kbps]

9000 9000
8000 8000
7000 7000
6000 6000
5000 5000
4000 4000
3000 3000
2000 2000
1000 1000
0 0
5 15 25 5 15 25
HeNB density HeNB density

Figure 7.3.2.2.2-2: HUE throughput (Left: Average, Right: 5 percentile)

7.3.2.2.3 Discussion of Results

The power control based on PL difference can mitigate the degradation of MUE throughput. Its cost is the degradation
of HUE 5 percentile throughput which is mainly correspond to outdoor HUE.

7.3.2.3 For Future Releases

The above smart power controls require no interference coordination between eNodeBs. As the result, it might be
difficult to manage the interference from HUE as HeNB density increases. For future releases (LTE Release 10 or LTE-
Advanced), the adaptive power control by means of X2 or S1 signalling between MeNB and HeNB or between HeNBs
should be investigated (e.g. to take account of the density of active femtocells within a macrocell coverage area). Notice
that during this work item, the adaptive power controls are proposed and their performance gain is investigated.

59
Release 13 60 3GPP TR 36.922 V13.0.0 (2016-01)

7.4 Downlink interference control

7.4.1 Control Channel Protection
Several techniques have been considered for data interference management. However control channel interference
management is equally important since improved data SINR is not useful if the UEs cannot receive control channels.
Thus, it is vital to have techniques that address control channel interference.

Downlink control channel (PDCCH) interference can occur in two directions in co-channel HeNB deployments.
• HeNB (aggressor) to macro-UE (victim), and

• macro-eNB (aggressor) to HUE (victim) if the UE is connected to a weaker HeNB cell (e.g. to access local
information at the HeNB).

This can lead to problems both in connected mode and in idle mode such as:

1. UE being unable to reliably decode paging channel resulting in missed pages and therefore a user's inability to
receive UE-terminated calls,

2. UE being unable to read common control channels, and

3. throughput degradation or degraded PDSCH performance.

The following are some of the techniques that could be used for control channel protection. It should be noted that some
of these aspects may require UE implementation changes and should be considered for Rel 10 and beyond. It is possible
that these methods offer gains for Rel 8/9 UEs; however, this needs to be studied further.

7.4.1.1 Control of HeNB downlink interference towards macro eNB control channels
by frequency partitioning with per-subband interference estimation
Frequency partitioning, or carrier offsetting, where HeNBs are confined to use only a part of the bandwidth can be used
to mitigate interference problems. This scenario is shown in Fig. 7.4.1.1-1. By using scheduling techniques that would
avoid data transmissions on those parts of the bandwidth, the levels of interference as seen at the receiver can be
reduced. This could resolve the interference problem for the data transmissions, however, control channels such as
PDCCH that span the entire bandwidth would still be affected.

Fig. 7.4.1.1-1 Partial Bandwidth Coexistence

The effects of the high interference seen in one of the subbands can be mitigated if the interference estimation is done
on a per-subband basis. This would confine the influence of the interference only to that subband and not allow it to
affect the entire bandwidth. This in turn would mean that only some of the coded bits are affected. When wideband
interference estimation is used, all the bits are affected and the probability of successfully decoding the message
decreases. Assuming sufficient number of CCEs are used (i.e., enough code protection), the PDCCH BLER
performance would be slightly degraded. But the transmission would likely be reliable enough not to significantly affect
normal operation.

60
Release 13 61 3GPP TR 36.922 V13.0.0 (2016-01)

To illustrate the performance of this scheme, some simulation results are given. A simulation was performed to evaluate
the impact on control channel performance of high interference on one of the subbands. Results for the cases of per-
subband interference estimation and wideband interference estimation are presented.
The simulation considers a HeNB that uses one fourth of the bandwidth of the macro as shown in Fig. 7.4.1.1-1. A UE
connected to a macro-eNB and receiving PDCCH transmission from it, sees high interference on one of the subbands.
The level of interference is varied as a parameter relative to the noise level. The PDCCH error rate is compared for the
cases when wideband interference estimation and per-subband interference estimation are used. The simulation
parameters are given in Table 7.4.1.1-1. Only the results for 4 CCE PDCCH are given here but similar results were
observed for other PDCCH sizes.

Table 7.4.1.1-1: Simulation parameters used

Parameter Assumption
Information payload size 40 bits
Coding 1/3 rate TBCC with rate matching
Macro Bandwidth 5 MHz
HeNB bandwidth 1/4 of macro Bandwidth
Channel model TU, 3km/h
Channel estimation 2-D MMSE channel estimation
Interference estimation Ideal

Fig. 7.4.1.1-2 CCE PDCCH BLER with wideband interference estimation

Fig.7.4.1.1-3 4CCE PDCCH BLER with subband interference estimation

61
Release 13 62 3GPP TR 36.922 V13.0.0 (2016-01)

7.4.1.2 Control of HeNB downlink interference among neighboring HeNBs control

channels by frequency partitioning
Unlike data (PDSCH, PUSCH), there is no HARQ for control channel transmissions which must typically target fairly
low BLER of 1% or less. HeNBs that are in close proximity of each other will not have reliable downlink control
channels (e.g. PDCCH, PHICH, PCFICH, PBCH, P/S-SCH). One way to solve this is to segment the LTE carrier and
allow the interfering HeNBs to transmit their control signalling in separate frequency domain resources. For example, if
the LTE carrier is 20MHz then it would be segmented into two 10MHz carriers on the downlink with each of the two
interfering HeNBs transmitting its control signalling (PDCCH, PHICH, PCFICH, P-SCH, S-SCH, PBCH) on one of the
10 MHz carriers.

Both Release-8 UEs and Release-9 UEs would access the HeNB as a 10 MHz carrier and receive control and broadcast
signalling from HeNB within 10 MHz. However, Release-10 UEs can additionally be assigned PDSCH resource on the
remaining 10 MHz frequency resources using carrier aggregation. Therefore, while Release-8/9 UEs are limited to
allocations of 50 RBs, Release-10 UEs could be assigned any portion of the 100 RBs.

7.4.2 Data Channel Protection

7.4.2.1 Control of HeNB Downlink Interference towards macro eNB data channels by
frequency partition
Frequency partition between Macro eNB and HeNB can be utilized to mitigate the interference from HeNB to Macro
eNB. HeNB can get frequency partition information of its neighbour Macro eNB through air link measurement if
additional receiver is enabled on HeNB. Alternatively, a semi-static scheme can be adopted if a pre-configuration of the
frequency partition can be determined by Macro eNB management server. For example, Macro eNB will schedule
resource blocks to Macro UE based on its location. When HeNB gets its own location information, it will know which
resource blocks will be assigned to a nearby macro UE.

With the knowledge of the frequency partition information, for example, HeNB knows which set of resource blocks will
be used for Macro eNB cell center users (CCU), and which set of resource blocks will be used for Macrocell cell edge
users (CEU), HeNB can coordinate its transmission to avoid its interference to nearby Macro UE by giving high
scheduling priority to resource blocks not used by the nearby Macro UE. For example, if HeNB is located at the edge of
the Macro eNB, HeNB will give higher priority to resource blocks used by macro center UEs for downlink transmission.
If HeNB is located at the center of the Macro eNB, HeNB will give higher priority to resource blocks used by macro
edge UEs for downlink transmission as shown in Figure 7.4.2.1-1.
HeNBl a CCU

a CEU HeNB

Macro eNB
Macro eNB

(a) (b)

Figure 7.4.2.1-1 Examples of HeNB and macro UE location

7.4.2.2 Control of HeNB Downlink Interference among neighboring HeNBs

HeNBs listen to neighboring HeNBs' control channel and reference signal transmissions, determines the cell ID of each
neighboring HeNB and measure the path loss from each of them. In addition, the HeNBs could also use reports from
UEs.

Based on this information, HeNBs could use fractional frequency reuse (FFR) to orthogonalize the resources used and
increase the overall performance of the network.

7.4.2.2.1 Centralized coordination

The centralized coordinator can form an adjacency graph of all HeNBs based on the reports from each HeNB.

62
Release 13 63 3GPP TR 36.922 V13.0.0 (2016-01)

Each HeNB estimates the fraction of time it needs to transmit according to the traffic load and channel conditions of its
UEs, and reports this ratio to the centralized controller via S1 signalling.

• For mixed traffic with both delay sensitive traffic and delay tolerant traffic, two ratios which correspond to both
traffic types will be reported.

• Each HeNB needs to update its report when at least one of the following event happens:

▪ New traffic session initiation

▪ UE channel condition variation over a pre-defined threshold

Given the adjacency graph and the reported ratios from each HeNB, the centralized coordinator determines

Option 1: the subframes that each HeNB is allowed to transmit, and notifies each HeNB of its transmission pattern via
S1 signalling. A HeNB needs to properly configure DRX parameters of its UEs according to the transmission pattern
notified by the centralized coordinator.

Option 2: the subbands or carrier frequency that each HeNB is allowed to transmit, and notifies each HeNB of its
transmission pattern via S1 signalling.

Note that the S1 signalling load between HeNBs and the centralized coordinator could be large if the number of HeNB
connections per coordinator is significant. To reduce S1 signalling, it is preferable to limit the number of HeNB reports
to the centralized coordinator. For example, the centralized coordinator can assign a HeNB a lot more resources than it
actually needs, and the HeNB will not send a report to the centralized controller unless it uses up all the assigned
resources. As HeNBs are generally lightly loaded, a HeNB may rarely send a report if it is assigned a large fraction of
resources (i.e. subframes, subbands or carrier frequency).

7.4.2.2.2 Distributed Dynamic Frequency Partitioning

Based on the information collected, the HeNBs can construct a "Jamming Graph" where each node denotes an active
HeNB and an edge denotes jamming condition between two HeNBs. A jamming condition is declared when the channel
gain difference between the interfering and serving links exceeds a certain threshold. The distributed fractional
frequency reuse planning problem is now converted into a graph coloring problem, which could be solved in a
distributed manner at low complexity.

Examples of such algorithms and brief performance analysis are given in several RAN4 contributions for both the FDD
and TDD case.

This algorithm could be an adaptive algorithm in which resources are negotiated and adaptively allocated for different
nodes, based on a utility function that enables nodes to quantify the benefit or loss due to each resource coordination
action. These utility values can then be used at each node to select the right resource coordination requests to be sent to
their neighbors, or to select the best requested coordination action from among multiple received requests, and hence to
grant/reject the requests based on their quantified benefit to the network.

To support this adaptive algorithm, network nodes need to exchange information such as subbands reuse updates and
utility information. The performance depends significantly on the latency of the messaging, especially in the case of
non-full buffer traffic. This information should be taken into account while analyzing the different options for such
information exchange (e.g. X2, S1, over-the-air, over-the-air via UE).

7.4.2.3 Adaptive Frequency Selection

To minimize the mutual interference between HeNB and surrounding cells, RSRP measurements and cell re-selection
priority information in SIBs or RRC message from neighbour cells could be utilized for HeNB to adaptively select its
operating carrier frequency.

The original purpose of cell-reselection priority information is for UE to properly select its camping cell based on the
frequency layer priority of each carrier. For interference avoidance, this information can be utilized by HeNB to
appropriately select its operating frequency by decoding SIBs or RRC message from its neighbour cells during normal
operation or in self-configuration mode. At the same time, HeNB should also measure co-channel or adjacent channel
RSRP of the neighbour cells to make a proper selection. Furthermore, if the priority information of the neighbour cells
is changed at a certain time after the HeNB has already started up, the HeNB should periodically update the priority
information, measure the RSRP and adjust its operating frequency (if necessary).

63
Release 13 64 3GPP TR 36.922 V13.0.0 (2016-01)

Figure 7.4.2.3-1 below depicts a scenario where the HeNB is overlaid by both higher and lower priority frequency
layers as shown for HeNB cell 1, and another scenario where the HeNB is overlaid by only one frequency layer as
shown for HeNB cell 2.

For HeNB cell 1, when the measured RSRP of both higher and lower priority frequency layers exceed a predefined set
of thresholds which implies that it is overlaid by more than one frequency layers, the HeNB should select its operating
frequency that is being used by the cell that has the lower priority for interference mitigation. As illustrated in Figure
7.4.2.3-1, the HeNB cell 1 should select RF1 which is the lower priority carrier frequency. On the other hand, if the
HeNB is overlaid by only one priority carrier frequency as shown for HeNB cell 2 in Figure 7.4.2.3-1, the HeNB should
select a carrier frequency that corresponds to the smaller measured RSRP. For example, the HeNB cell 2 in Fig.7.4.2.3-
1 should select RF2 which has smaller RSRP.

HeNB cell2 HeNB cell 1

RF1: lower priority RF2: higher priority

RF2 selected RF1 selected

Figure 7.4.2.3-1 Examples of adaptive frequency selection scheme

The adaptive frequency selection scheme is applied most effectively to HeNBs in non-allowed CSG mode for avoiding
the interference. In the case when the HeNB is operating in an open or hybrid access mode, it may also apply the
scheme for interference avoidance.

7.4.2.4 Downlink interference management based on mapping between PCIs and

transmission patterns
Interference Management based on mapping between PCIs and transmission patterns requires a centralized coordinator.
The centralized coordinator generates a number of transmission patterns and a function which maps transmission
patterns to PCIs, and then sends to all cells the information of transmission patterns and the mapping function. The
essentials of this scheme are as follows:

- When powered on, a HeNB will listen to neighboring cell transmissions, determine the Cell ID of these
neighboring HeNBs, and report them to the centralized coordinator.

- The centralized coordinator can form an adjacency graph of all HeNBs based on the reports from each HeNB as
well as its UEs.

- Given the adjacency graph, the centralized coordinator determines the followings semi-statically:

▪ a set of transmission patterns; A pattern shows a profile of maximum Tx power for each PRB or group of
PRBs

▪ a function which maps transmission patterns to PCIs

- The centralized coordinator notifies to HeNBs and MeNBs of the followings via S1 signalling.

▪ the set of transmission patterns

▪ the function which maps transmission patterns to PCIs

- This notification makes it possible that all cells share the information of the transmission patterns and the
mapping between the patterns and PCIs.

- HeNB should use PRBs according to the transmission pattern corresponding to its PCI.

Downlink interference management can be done as follows:

64
Release 13 65 3GPP TR 36.922 V13.0.0 (2016-01)

If a UE connected to a serving (H)eNB is close to a neighbouring HeNB, the downlink of the UE can be severely
degraded due to high interference from the neighbouring HeNB. The serving (H)eNB can be macro eNB or HeNB. The
UE detects the PCI and measures the RSRP of the neighbouring HeNB, and reports these to the serving (H)eNB.
Because of the sharing of the information of the transmission patterns and the mapping between these patterns and PCI,
the serving (H)eNB knows the transmission pattern of the neighbouring HeNB that the UE is close to. This makes it
possible for the serving (H)eNB to allocate downlink resource such that the UE is not severely interfered by the
neighbouring HeNB on its downlink.

7.4.2.5 Control of HeNB Downlink Interference by dynamically changing HeNB CSG

ID
To mitigate HeNB interference as well as to prevent free HeNB usage by neighbors, one option is to change the CSG
ID of a closed access HeNB dynamically between

• Its default CSG ID which is assigned when it is deployed and

• A dedicated CSG ID which is configured by the operator.

This dedicated CSG ID is included in the Operator CSG list for every UE in the operator network. When the closed
access HeNB changes its CSG ID to the dedicated CSG ID, it actually becomes accessible to each passing-by UE, thus
alleviating its interference towards the macro cell. Note that this scheme would only be beneficial when a large fraction
of UEs support CSG.

By dynamically changing the CSG IDs of closed access HeNBs in a coordinated manner, it might be possible to ensure
that the number of closed access HeNBs using default CSG IDs under each macro cell does not go beyond a pre-defined
upper bound at any time instant.

• A centralized controller (e.g. HeNB Gateway or HeNB management server) can set the upper bound, and
configures, for each HeNB, the time intervals it supports the default CSG ID and the dedicated CSG ID,
respectively. Then each HeNB dynamically varies its CSG ID based on the configuration parameters provided
by the centralized controller.

▪ Different macro cells may have different or common upper bounds.

▪ Whenever a HeNB is deployed, the centralized controller can determine its adjacent macro cell based on its
location and configure the time intervals it will employ the default CSG ID and the dedicated CSG ID,
respectively.

▪ `The mechanisms of when and how to select a set of HeNBs is FFS. It may be possible to carefully arrange
time intervals to employ default CSG IDs for different HeNBs to ensure that the number of closed access
HeNBs using default CSG IDs under every macro cell can be upper limited at any time instant.

- E.g. some HeNBs may use default CSG IDs the first 30 minutes of every hour, but others may not;
alternatively, some HeNBs may use default CSG IDs 9AM-11PM everyday, but others may do.

▪ The upper bound can be determined by

- Simulation/numerical analysis

- HeNB subscription profile

> E.g. make HeNBs "less open" if the subscription rate is high and the subscribers desire to do so.

- Macro cell load

> E.g. make HeNBs "more open" when macro cell is heavy loaded, and "less open" when macro cell is
lightly loaded.

Further study of this proposals and comparison to alternatives (e.g. hybrid cells) may be needed. In particular, the UE
and core network impact of dynamically changing the CSG ID is for further study by the relevant working groups.

65
Release 13 66 3GPP TR 36.922 V13.0.0 (2016-01)

7.4.3 Power Control

7.4.3.1 HeNB power control based on HeNB-MUE path loss

HeNB should adjust the downlink transmit power by taking into account the path loss between the HeNB and an
outdoor neighbour MUE including penetration loss in order to provide better interference mitigation for the MUE while
maintaining good HeNB coverage for HUEs.

HeNB should set the transmit power of reference signal P_tx as follows:

P_tx (dBm) = MEDIAN( P_m + P_offset, P_tx_upp, P_tx_low )

Where:

- P_m (dBm) is RSRP from the nearest MeNB measured by the HeNB. P_m is dependent on path loss which
includes the penetration loss between the nearest MeNB and the HeNB.

- P_offset (dB) is the power offset described below in detail.

- P_tx_upp/P_tx_low (dBm) is the upper/lower limit value for the transmit power of the reference signal. The
maximum and the minimum total transmit power of HeNB should follow HNB in [7].

The HeNB can also set the maximum downlink transmit power in proportion to the transmit power of the reference
signal. As the RSRP decreases, which means the HeNB is located close to the edge of the macro cell, the transmit
power should be small in order to mitigate the downlink interference to the MUE.

P_offset above should be defined based on path loss between the HeNB and the MUE. The path loss may consist of
indoor path loss between the HeNB and cell edge of HeNB cell and the penetration loss. Therefore, P_offset should be
formulated as follows:

P_offset = MEDIAN( P_offset_o + K*LE, P_offset_max, P_offset_min )

Where:

- P_offset_o (dB) is a predetermined power offset value corresponding to the indoor path loss. Typical value range
between 50 and 100dB, and can be determined by the averaged measurement value.

- K is an adjustable positive factor can be determined by the priority of HeNB operation. This value should be high
to increase the total transmit power (MeNB is more acceptable to higher interference) and low to reduce the
interference to MeNB operation.

- LE (dB) is estimated penetration loss as below.

- P_offset_max/P_offset_min (dB) is the maximum/minimum value of the P_offset by which the estimated and
calculated P_offset can be prevented from being too large or too small. This value is dependent of the actual wall
penetration loss plus P_offset_o. And the typical wall penetration loss ranges between 10 and 30dB.

If the path loss between the HeNB and the MUE can be estimated, then the transmit power of the HeNB should be set
accordingly.

The path loss between the HeNB and the MUE should be estimated based on the difference between the estimated UL
transmit power and the UL reception power (as the Received Interference Power) of the MUE. The estimated UL
transmit power is based on the assumption that UL power control is applied for both MUE and HeNB as a UE. Then the
UL transmit power can be calculated by the DL propagation loss from the surrounding MeNB to the HeNB utilizing the
RSRP measurement.

7.4.3.2 Smart power control based on interference measurement from macro BS

In the absence of alternative techniques such as resource partitioning, a fixed maximum power setting configured by the
network is not sufficient to ensure minimum HeNB coverage range while protecting macro in all cases. Hence, a
sensing of the HeNB environment can be used for setting the transmit power.

The HeNB adjust its maximum DL transmit power as a function of air interface measurements to avoid interfering with
macro cell UEs. Examples of such measurements are total received interference, received CRS Êc from the most

66
Release 13 67 3GPP TR 36.922 V13.0.0 (2016-01)

dominant macro cell eNB, etc. The scheme is open loop, and does not involve the UEs and signalling between network
nodes.

The HeNB shall adjust its maximum transmit power according to the following formula,

Ptx = max(min(a × (CRS Eˆ C + 10 log( N RB

DL
× N SC
RB
)) + β , Pmax ), Pmin ) [dBm] (1)

where parameters Pmax and Pmin is the maximum and minimum HeNB transmit power settings, CRS Êc is measured in
dBm, which is the Reference Signal Received Power per resource element present at the Home BS antenna connector
DL
received from the strongest co-channel macro cell. N RB is the number of downlink resource blocks in the HeNB
channel. N scRB is the number of subcarriers in a resource block ( N scRB = 12 ). Parameter a is a linear scalar that allows
altering the slope of power control mapping curve, β is a parameter expressed in dB that can be used for altering the
exact range of CRS Êc covered by dynamic range of power control.

Parameters Pmin , a, and β are considered to be HeNB configuration parameters, and Pmax corresponds to the HeNBs
maximum transmit power capability.

For the special case where the HeNB is unable to detect any co-channel deployed macro cells. The HeNB is free to use
its maximum transmit power, or apply pre-set parameterized value based on the requirement of operator.

7.5 Hybrid Cells

Hybrid cells are being included in the 3GPP release 9 specifications. Hybrid HeNBs may provide different service
levels to UEs that are members of the HeNB and non-member UEs. In RAN4, extensive deployment scenarios of
hybrid cells have been discussed. The interference scenarios apply to most of the deployments under consideration.

For the scenario where HeNBs are on a shared carrier with eNBs, the interference management considerations are
different between closed and hybrid access modes. For the closed access mode the used HeNB resources (e.g. power,
RBs) are selected as a trade-off between performance at the HeNB/HUEs and interference caused to the macro
eNB/MUEs. For the hybrid access mode the trade-off is between overall system performance (including both macro
eNB and HeNB layers), and resources consumed at the HeNB by "visiting" (i.e. non-CSG member) UEs. These aspects
are considered in more detail in the following sub-sections.

7.5.1 Hybrid Access Level of Service

Hybrid HeNB may provide different service levels to UEs that are members of the HeNB and non-member UEs. The
lowest level of services is paging service, where a hybrid cell allows a non-member UE to access the cell to receive
pages. A paging only hybrid cell is an interesting alternative to pure CSG cells. Since CSG cells have separate PCID
space, switching between CSG and hybrid mode would have impact on both idle state and connected state home UEs.
On the contrary, a hybrid cell could with paging-only service provides similar functionality as a CSG cell without
incurring CSG-hybrid switching penalty.

If a hybrid cell only provides paging services to non-member UEs, data channel interference is similar to CSG HeNBs.
The difference is that the hybrid cell has more information about the victim UE than a CSG cell. When the hybrid cell
decides to handover (HO) the UE to a macro cell, interference coordination could be negotiated with the macro as part
of the HO procedure. Some examples, are:

1. DL interference: The hybrid cell could reduce transmit power such that the UE handed over to the target cell has
sufficient DL C/I to receive DL control channels from the target cell. The hybrid cell could also engage in
fractional frequency reuse (FFR) with the target cell to enhance DL data rate of the victim UE.

2. UL interference: A hybrid cell and HO target cell could choose the power setting of this UE such that UL
interference could be coordinated. UL control channels of the HO sUE could also be orthogonalized with the
PUCCH of the source hybrid cell. UL data channel coordination through UL FFR could be configured on a semi-
static basis.

Note that hybrid cells also have additional information on the channel quality of active UEs. Hence the adjustment
made at the hybrid cell could fully take into account on the impact to ongoing traffic. In some extreme scenarios, hybrid
cell could handover both the member UEs and non-member UEs to the macro cell.

67
Release 13 68 3GPP TR 36.922 V13.0.0 (2016-01)

If a hybrid cell provides data services to non-member UEs, the hybrid HeNB is similar to a pico cell with lower Tx
power and different service level for members and non-members. In addition, a hybrid HeNB is different from a
operator deployed pico cell in the following areas

1. A release 9 hybrid HeNB does not support X2 interface.

2. A hybrid HeNB could be customer deployed without proper RF planning

3. The density of hybrid HeNB could have much high density than operator deployed pico cells.

Given the challenges mentioned above, non-Rel-8 interference coordination schemes should be investigated for hybrid
cells.

7.5.2 DL Performance Evaluation

Full buffer performance analysis is performed for CSG and hybrid HeNB deployments based on FDD. The dense-urban
model corresponds to densely-populated areas where there are multi-floor apartment buildings with smaller size
apartment units. Similar performance trends can be observed for TDD.

The set of simulation parameters are shown below:

• System bandwidth 5 MHz

• Macro Power = 43dBm

• HeNB power between [-10 dBm, 10 dBm].

• Case 1, the HeNB power is fixed to 8 dBm

• Case 2, the adaptive HeNB power setting is used to reduce the MUE outage

• ISD of 1km

• Noise power = -99dBm

• 57 cell wrap around model with 3 center cells simulated for traffic

• 10 macro UEs per cell

• HeNB penetration rate of 5%, and activity factor of 13%, this leads to 12 active home UEs per cell

The association algorithm in all cases is based on the best RSRQ among allowed cells. In the case of CSG cells, UEs
are only allowed to associate with the macro cell or the HeNB in the same CSG group. In the case of hybrid cells, all
cells are open.

The C/I and throughput distribution for CSG and hybrid cells are shown in Figures 7.5.2-1 to 7.5.2-4. As discussed in
RAN4, adaptive power control could reduce outage for CSG cells. In the case of hybrid cells, there is no outage even
without adaptive power control due to open association. Note that if different service levels are enforced for group
member and non-group members, the fairness could be different from those shown in the figures.

Key mobile statistics (outage, 20% throughput and median throughput) are shown in Table 7.5.2-1. It is noted that
hybrid cells improve the outage and edge user performance, while making little difference in high throughput region.
This is consistent with the expectation that with hybrid cells, the network would be able to offload macro UEs in poor
channel conditions to close by hybrid cells. It is also interesting to note that the improvements due to hybrid cell is
much higher when adaptive HeNB power control is not available (800% gain versus 60% gain).

Note that backhaul limitation of CSG and hybrid cells are not modelled in the simulations. For practical deployments,
users close by a hybrid cell is likely to be backhaul limited rather than air-interface limited.

68
Release 13 69 3GPP TR 36.922 V13.0.0 (2016-01)

1
1
no HeNBs 0.9
0.9 with HeNBs
0.8
0.8

0.7
0.7

0.6 0.6

CDF
CDF

0.5 0.5

0.4 0.4

0.3 0.3 no HeNBs

with HeNBs
0.2 0.2

0.1 0.1

0 0
-60 -50 -40 -30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30
C/I C/I

(a) CSG cells (b) Hybrid cells

Figure 7.5.2-1 C/I for CSG cells and hybrid cells deployments with 8 dBm HeNB Tx power

1
1
0.9
0.9
0.8
0.8

0.7
0.7

0.6 0.6
CDF
CDF

0.5 0.5

0.4 0.4

0.3 0.3

0.2 0.2 no HeNBs

no HeNBs with HeNBs
0.1 with HeNBs 0.1

0 0
5 6 7 8 5 6 7 8
10 10 10 10 10 10 10 10
UE Avg Throughput [bps] UE Avg Throughput [bps]

(a) CSG cells (b) Hybrid cells

Figure 7.5.2-2 Throughput for CSG cells and hybrid cells deployments with 8 dBm HeNB Tx power

1
1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6
CDF

0.5
CDF

0.5

0.4
0.4

0.3 0.3
no HeNBs
with HeNBs
0.2 0.2
no HeNBs
0.1 0.1 with HeNBs

0 0
-40 -30 -20 -10 0 10 20 30 -30 -20 -10 0 10 20 30
C/I C/I

(a) CSG cells (b) Hybrid cells

Figure 7.5.2-3 C/I for CSG cells and hybrid cells deployments with adaptive HeNB Tx power

69
Release 13 70 3GPP TR 36.922 V13.0.0 (2016-01)

1
1
0.9
0.9
0.8
0.8
0.7
0.7
0.6
0.6

CDF
0.5
CDF

0.5

0.4 0.4

0.3 0.3

0.2 no HeNBs 0.2

with HeNBs no HeNBs
0.1 0.1 with HeNBs

0 0
5 6 7 8 5 6 7 8
10 10 10 10 10 10 10 10
UE Avg Throughput [bps] UE Avg Throughput [bps]

(a) CSG cells (b) Hybrid cells

Figure 7.5.2-4 Throughput for CSG cells and hybrid cells deployments with adaptive HeNB Tx power

Table 7.5.2-1 Summary of results

Outage Probability Worst 20% mobile Median throughput

(SNR < -6 dB) throughput (kbps) (kbps)
No HeNB 12.7% 35 150
CSG HeNB with fixed Tx
18.9% 100 5600
power of 8 dBm
CSG HeNB with adaptive Tx
9.8 % 250 3300
power
Hybrid HeNB with fixed Tx
2% 900 5100
power of 8 dBm
Hybrid HeNB with adaptive
3% 400 3400
Tx power

7.5.3 Hybrid Cell RB Resource Management

For hybrid cells, non-CSG members consume RB resources at the HeNB, the amount of which will depend on the
number of non-CSG UEs and the service level provided to the non-CSG UEs. One possible method of managing the RB
resources used at a HeNB for non-CSG UEs is to reserve some RB resources for use by non-CSG UEs.

In hybrid access mode, if a HeNB accepts non-CSG members as temporary users, it would degrade CSG members'
capacity similar to the open access HeNB. Moreover, when HeNB is under heavily loading, non-CSG UEs may be
blocked first and diverted to macro eNBs. These diverted non-CSG UEs that are still within the coverage of the hybrid
access HeNB may experience strong interferences from the HeNB.

In order to manage RB resource and mitigate the DL interference of the hybrid cells, a method called "Resource Priority
Region (RPR)" may be used which guarantees a small percentage of HeNB resources for non-CSG members.

The RPR for the hybrid access HeNB divides radio resources of a HeNB into two regions for non-CSG members and
CSG members respectively. The detailed definitions for each resource region are:

1. Non-CSG member priority region – non-CSG members have higher priority than CSG members.

2. CSG member priority region – CSG members have higher priority than non-CSG members.

70
Release 13 71 3GPP TR 36.922 V13.0.0 (2016-01)

Fig.7.5.3-1 Resource Priority Region

A threshold – Priority Region Threshold (PRT) is set to separate resources between two priority regions. The PRT
could be a time or physical resource block (PRB) in radio frames, and PRT could be statically or dynamically adjusted
by exchanging ICIC messages between HeNB and macro eNB. The HeNB could autonomously define the threshold for
release 9, and further enhancements allowing the threshold to be adapted e.g. based on S1 signalling could be
considered for release 10.

The hybrid access HeNB with RPR efficiently decreases the blocking probability to non-CSG members when HeNB is
exhausting its resource. Also, this method guarantees the CSG members throughput that HeNB are not affected by
sharing the resource with non-CSG members.

7.5.4 Hybrid Cell Power Management

The optimum power setting for hybrid cells is likely to be different than for closed cells, in that for closed cells the
power is set as a compromise between HeNB coverage and interference caused to neighbour "victim" cells, whereas for
hybrid HeNB it is set as a compromise between overall system performance versus resources used at the HeNB by non-
CSG UEs.

Measurements made by the HeNB of neighbour cells ("sniffing") can be used to set an appropriate downlink power.
However the propagation conditions between a neighbouring (H)eNB and its associated UEs may differ significantly
from the propagation conditions between a neighbouring (H)eNB and the HeNB as measured during "sniffing".
Furthermore the propagation conditions between the HeNB and nearby non-served UEs will not be known. These
differences will result in uncertainty when estimating the coverage of HeNB and neighbouring (H)eNBs to non-served
UEs.

One potential way to allow a hybrid access mode HeNB to get a more accurate picture of its local environment is for the
HeNB to request a UE to measure RSRP and/or RSRQ of both source and target cells immediately after a UE hands-in
(active state) or registers with (idle state) the HeNB. This would apply particularly to non-CSG UEs but could also
apply to CSG UEs. In this way the HeNB could, for example, determine if the hand-in or re-selection is a) due to poor
signal level from the source (e.g. macro) cell or b) due to high interference from the HeNB. This would then allow the
HeNB to determine its output power appropriately. For example if the signal is particularly poor on the source (e.g.
macro) cell, the HeNB could use a relatively high power and/or provide relatively high access priorities for non CSG
UEs, compared to the case where the signal level on the source cell is not so poor.

This basic approach would require no standards changes. Future standards releases could consider more sophisticated
approaches aimed at improved performance e.g. based on UE storing measurement reports and/or events prior to a
hand-in or reselection of a HeNB, with subsequent reporting to the HeNB.

71
Release 13 72 3GPP TR 36.922 V13.0.0 (2016-01)

Annex A (informative):
Change History

Change history
Date TSG # TSG Doc. CR Rev Subject/Comment Old New
2009-03 RAN4#50bis R4-091231 TR skeleton 0.0.1
2009-06 RAN4#51bis R4-092253 Agreed TPs in RAN4#51: 0.0.1 0.1.0
R4-091787, "Text proposal for TDD HeNB related interference
scenarios and deployment configurations", CMCC
2009-08 RAN4#52 R4-092863 Agreed TPs in RAN4#51bis: 0.1.0 0.2.0
R4-092256, "Text proposal on ACLR requirements of TD-LTE
HeNB", CMCC
R4-092258, "Text proposal on frequency error of TD-LTE
HeNB", CMCC
R4-092259, "Text proposal on performance requirements of TD-
LTE HeNB", CMCC
2009-08 RAN4#52 R4-093360 Agreed TPs in RAN4#52: 0.2.0 0.3.0
R4-093333, "Text proposal on section 6.2.3 frequency error",
CATT
R4-093325, "Text proposal on TD-LTE HeNB receiver
sensitivity", CMCC, CATT
R4-093326, "Text proposal on TD-LTE HeNB spurious
emission", CMCC
R4-093327, "Text proposal on TD-LTE HeNB ACS requirement",
CMCC
R4-093328, "Text proposal on TD-LTE HeNB blocking
requirements", CMCC
R4-093329, "Text proposal on TD-LTE HeNB dynamic range",
CMCC
2009-11 RAN4#53 R4-094562 Agreed TPs in RAN4#52bis: 0.3.0 0.4.0
R4-093532, "Home eNode B receiver in channel selectivity
requirement", CATT
R4-093973, "Text proposal on HeNB receiver requirements",
CMCC
R4-093974, "Text proposal on HeNB spurious emission
requirement", CMCC
R4-094003, "Text proposal on TD-LTE HeNB operating band
unwanted emissions", CMCC
R4-094004, "Text proposal on HeNB transmitter
intermodulation", CMCC
R4-094005, "Text proposal on HeNB ACS and narrow band
blocking requirements", CMCC
R4-094007, "Home eNode B receiver intermodulation
requirement", CATT
R4-094074, "Home eNode B Maximum output power", CATT
R4-094077, "Text proposal on TD-LTE HeNB synchronization
requirements", CMCC, Qualcomm
2009-12 RAN#46 RP-091149 Submitted to RAN plenary for information 0.4.0 1.0.0
2010-01 RAN4 R4-100176 Agreed TPs in RAN4 #53: 1.0.0 1.1.0
AdHoc#1 R4-094563, "Text proposal on TD-LTE HeNB performance
requirement in TR36.922", CMCC, CATT
R4-094985, "TDD HeNB synchronization requirement for large
propagation distance case", Qualcomm Europe, CMCC, Nokia
Siemens Networks, Nokia
2010-02 RAN4#54 R4-101074 Agreed TPs in RAN4 AdHoc#1: 1.1.0 1.2.0
R4-100049, "Text Proposal for TR 36.922: TDD HeNB
Synchronization using Network Listening", Nokia Siemens
Networks, Nokia, Qualcomm Incorporated, CMCC

72
Release 13 73 3GPP TR 36.922 V13.0.0 (2016-01)

R4-100177, "Text proposal on LTE TDD HeNB synchronization

requirement", CMCC, Nokia Siemens Network, Nokia,
Qualcomm Incorporated
R4-100178, "Text proposal on LTE TDD HeNB interference
control", CMCC
R4-100234, "Text Proposal for TR 36.922: Interference control
for LTE Rel-9 HeNB cells", Nokia Siemens Networks, Nokia,
Panasonic
Agreed TPs in RAN4 #54:
R4-100328, "Text Proposal for TR 36.922: TDD HeNB
Synchronization using Network Listening", Nokia Siemens
Networks, Nokia, Qualcomm Incorporated, CMCC
R4-100704, "Text proposal on LTE TDD HeNB interference
control", CMCC
R4-100705, "Text proposal on Home BS adjacent channel
protection", CMCC
2010-03 RAN4#54 R4-101078 Indices update and editorial modifications 1.2.0 1.2.1
2010-03 RAN#47 RP-100192 Submitted to RAN plenary for approval 1.2.1 1.3.0
2010-03 RAN#47 RP-100192 Approved by RAN 1.3.0 9.0.0
2010-06 RP-48 RP-100627 001 Editorial modifications on TDD HeNB TR36.922 9.0.0 9.1.0
SP-51 Upgraded unchanged from Rel-9 9.1.0 10.0.0
2012-09 SP-57 - - - Update to Rel-11 version (MCC) 10.0.0 11.0.0
2014-09 SP-65 - - - Update to Rel-12 version (MCC) 11.0.0 12.0.0
2016-01 SP-70 - - - Update to Rel-13 version (MCC) 12.0.0 13.0.0

73