TABLEOFCONTENTSTABLEOFCONTENTS
1 INTRODUCTION ..................................................................................................................................... 5


1.1 Goals of this White Paper ............................................................................................................... 5
1.2 Technology and Market Drivers for SON ........................................................................................ 5
1.3 Reasons for Automation .................................................................................................................. 6
2 3GPP EVOLUTION AND SON ................................................................................................................ 7


2.1 LTE SON High-Level Scope and Timeline ...................................................................................... 7
2.2 SON Development in NGMN .......................................................................................................... 8
2.3 SON Architecture Alternatives ...................................................................................................... 11
3 KEY LTE RELEASE 8, RELEASE 9 AND RELEASE 10 FEATURES SON........................................ 13


3.1 Base Station Self-Configuration .................................................................................................... 13
3.1.1 Benefits ................................................................................................................................. 13
3.1.2 Description............................................................................................................................ 13
3.1.3 Self-Configuration Actions ..................................................................................................... 14
3.1.4 Self-Configuration Status in 3GPP ........................................................................................ 15
3.2 Automatic Neighbor Relation (ANR) ............................................................................................. 15
3.2.1 Benefits ................................................................................................................................. 15
3.2.2 Description............................................................................................................................ 16
3.2.3 Neighbor Relation Discovery ................................................................................................. 16
3.2.4 ANR Actions .......................................................................................................................... 16
3.2.5 Inter-RAT ANR ...................................................................................................................... 17
3.3 Tracking Area Planning ................................................................................................................. 18
3.3.1 Benefits ................................................................................................................................. 19
3.4 PCI Planning ................................................................................................................................ 19
3.4.1 Benefits ................................................................................................................................. 20
3.5 Load Balancing ............................................................................................................................. 20
3.5.1 Benefits ................................................................................................................................. 20
3.5.2 Description............................................................................................................................ 21

3.5.3 Determining A Load Imbalance Condition ............................................................................. 21


3.5.4 Idle Mode Load Balancing ..................................................................................................... 22


3.5.5 Active Mode Load Balancing ................................................................................................. 22


3.5.6 Adapting Handover Configuration ......................................................................................... 23


3.6 Mobility Robustness / Handover Optimization .............................................................................. 24


3.6.1 Benefits ................................................................................................................................. 24


3.6.2 Description............................................................................................................................ 24


3.6.3 Intra-RAT Late HO Triggering ............................................................................................... 25


3.6.4 Intra-RAT Early HO Triggering .............................................................................................. 26


3.6.5 Intra-RAT HO to an Incorrect Cell ......................................................................................... 27


3.6.6 Inter RAT Too Late HO ......................................................................................................... 28


3.6.7 Inter RAT Unnecessary (too early) HO ................................................................................. 29


3.6.8 Algorithm for SON Handover Parameter Optimization.......................................................... 29


3.6.9 Examples of SON Algorithm Interactions .............................................................................. 30


3.7 RACH (Random access channel) Optimization ............................................................................ 31


3.7.1 Benefits ................................................................................................................................. 31


3.7.2 Description ............................................................................................................................. 31


3.8 Inter Cell Interference Coordination .............................................................................................. 32


3.8.1 Benefits ................................................................................................................................. 33


3.8.2 Description of ICIC operation ................................................................................................ 33


3.8.3 Frequency Planning .............................................................................................................. 35


3.8.4 Self-Tuning X2 Independent Algorithms ............................................................................... 35


3.8.5 Performance of ICIC Techniques Relative to Self-Tuning X2 Independent Algorithms ........ 36


3.8.6 Optimization of Coverage/Intererence with H-eNB............................................................... 36


3.8.7 Release-10 ICIC Enhancements ........................................................................................... 38


3.9 Energy Savings ............................................................................................................................. 38


3.9.1 Benefits ................................................................................................................................. 39


3.9.2 Description............................................................................................................................ 39



3.9.3 Deployment Scenarios with LTE in Clusters, Overlaying an Underlying 2G/3G Network .... 40

3.9.4 Scenarios for Further Study .................................................................................................. 41


3.10 Celloutage Detection And Compensation ..................................................................................... 42


3.10.1 Benefits ................................................................................................................................. 42


3.10.2 Cell Outage Detection ........................................................................................................... 43


3.10.3 Cell Outage Compensation ................................................................................................... 43


3.11 Coverage And Capacity Optimization ........................................................................................... 47


3.11.1 Modification of Antenna Tilts ................................................................................................. 47


3.11.2 Minimization of Drive Tests ................................................................................................... 53


3.12 Applications of SON to Address Deployment and Operation of DAS and Small Cells.................56


3.12.1 SON Interactions with Distributed Antenna Systems ............................................................ 56


3.12.2 SON with Picos/Femtos/Relays ............................................................................................ 56


3.12.3 Using SON to Facilitate the Deployment of Repeaters/Relays ............................................. 57


4 4G AMERICAS OPERATOR USECASES ............................................................................................ 59


4.1 4G Operator Usecase: ANR/PCI................................................................................................... 59


4.1.1 ANR/PCI Phase 1 .................................................................................................................. 59


4.1.2 ANR/PCI Phase 2 .................................................................................................................. 59


4.1.3 ANR/PCI Phase 3 .................................................................................................................. 60


4.2 4G Operator Usecase: Cell Outage .............................................................................................. 60


4.2.1 Cell Outage Detection ................................................................................................................ 60


4.2.2 Cell Outage Compensation Through Antenna Tilt Modificaton . PHASE 1 .............................. 60


4.2.3 Cell Outage Compensation Through Antenna Tilt Modification. PHASE 2 .............................. 60
5 RESULTS FROM LIVE OR TRIAL NETWORKS ................................................................................. 62
6 SON CAPABILITIES IN HSPA+ TECHNOLOGY ................................................................................. 65
7 SUMMARY AND CONCLUSION........................................................................................................... 66
LIST OF ACRONYMS ............................................................................................................................... 67
ACKNOWLEDGEMENTS.......................................................................................................................... 68
REFERENCES ........................................................................................................................................... 69



1INTRODUCTION1INTRODUCTION
1.1GOALSOFTHISWHITEPAPER
In today\'s mobile wireless networks, many network elements and associated parameters are manually
configured. Planning, commissioning, configuration, integration and management of these parameters are
essential for efficient and reliable network operation; however, the associated operations costs are
significant. Specialized expertise must be maintained to tune these network parameters, and the existing
manual process is time-consuming and potentially error-prone. In addition, this manual tuning process
inherently results in comparatively long delays in updating values in response to the often rapidly-
changing network topologies and operating conditions, resulting in sub-optimal network performance.

The recent deployment of LTE to address the growing data capacity crunch, has highlighted the need and
value of self-organizing capabilities within the network that permit reductions in operational expenses
(OPEX), during deployment as well as during continuing operations. Self-optimizing capabilities in the
network will lead to higher end user Quality of Experience (QoE) and reduced churn, thus allowing for
overall improved network performance. SON improves network performance, but in no way replaces the
wireless industry’s important need for more spectrum to meeting the rising mobile data demands from
subscribers.

3GPP initiated the work towards standardizing self-optimizing and self-organizing capabilities for LTE, in
Release 8 and Release 9. The standards provide network intelligence, automation and network
management features in order to automate the configuration and optimization of wireless networks to
adapt to varying radio channel conditions, thereby lowering costs, improving network performance and
flexibility. This effort has continued in Release 10 with additional enhancements in each of the above
areas and new areas allowing for inter-radio access technology operation, enhanced inter-cell
interference coordination, coverage and capacity optimization, energy efficiency and minimization of
operational expenses through minimization of drive tests. This paper discusses the Release 8, Release 9
and Release 10 Self-Organizing Network (SON) techniques and explains how these capabilities will
positively impact network operations in the evolved LTE network. Application of SON techniques by a 4G
Americas member company operator in LTE networks and live results from the application of SON to
HSPA+ networks are described, thus demonstrating the potential benefits of SON. Application of SON
techniques to address deployment and operational challenges with Distributed Antenna Systems (DAS),
and picos, femtos and relays are described to address the impact of SON on a variety of radio systems.
This paper is an update of the paper published in 2009 on “Benefits of SON in LTE” 17 that addressed the
rationale for SON and the description of SON features in 3GPP Releases 8 and 9.

1.2TECHNOLOGYANDMARKETDRIVERSFORSON
Reflecting upon recent wireless industry events, 4G Americas member companies observe several
important trends that are driving additional network complexity and operations effort.

New and emerging classes of mobile devices (smartphones, PC data cards, USB modems, consumer
devices with embedded wireless, machine-to-machine, etc.) are fostering explosive growth of wireless
data usage by public and enterprise users. As a result, wireless service providers have to simultaneously
support a growing number of higher-bandwidth data applications and services on their networks. The list
of user applications and services is quite broad in scope, and includes Internet browsing, Web 2.0, audio,


video, video on demand, gaming, location-based services, social networking, peer-to-peer, advertising,
etc.

On the network side, wireless service provider networks are becoming more complex and heterogeneous.
Projections point to rapidly growing numbers of femto and picocells (in order to drive greater coverage
and/or offload capacity from macrocells), plus increasing prevalence of multi-technology networks (2G,
3G, 4G, plus WiFi). These trends pose potentially significant operational and network complexity
regarding macro/femto and inter-technology handover, as well as management of macro/femto and
macro/pico interference. Taken together, these trends place ever-increasing demands upon service
providers’ networks and their operational staff. Ensuring quality user experience requires more complex
Quality of Service (QoS) and policy implementations while they simultaneously must increase network
throughput in response to the rapid growth in wireless data.

Moreover, wireless data revenue measured on a per-megabit (Mb) basis is decreasing. Fortunately,
spectral efficiency gains are provided by new wireless technologies, and do provide some measure of
relief; however, the data throughput per user is growing (and revenue per Mb is dropping) so rapidly that
spectral efficiency gains alone appear unable to keep up. Consequently, service providers . and
infrastructure vendors . are increasing their focus on operational cost reductions. Reflecting upon these
dramatic trends, it has become clear that traditional network management needs significant improvement
for managing this growing data volume and network complexity in a cost-effective manner.

1.3REASONSFORAUTOMATION
At a high level, there are two underlying operational issues facing service providers. Some processes are
repetitive, while others are too fast or difficult to be performed manually. The rationale for SON
automation can be grouped into two broad categories:

1.
Previously manual processes that are automated primarily to reduce the manual intervention in
network operations in order to obtain operational and/or deployment savings. Automating
repetitive processes clearly saves time and reduces effort. Auto-configuration and self-
configuration fall into this category.
2.
Processes that require automation because they are too fast, too granular (per-user, per-
application, per-flow, as a function of time or loading), and/or too complex for manual intervention.
Automatically collected measurements from multiple sources (e.g., from user devices, individual
network elements, and on an end-to-end basis from advanced monitoring tools) will provide
accurate real time and near real time data upon which these algorithms can operate thus
providing performance, quality, and/or operational benefits.
Consequently, substantial opportunities exist for cross-layer, end-to-end, and per-user/per-application/
per-flow optimizations for extracting additional performance benefits and management flexibility. These
categories need not be distinct, (e.g., a previously manual process that is growing too complex due to the
above trends) may by necessity require automation in order to manage it.

Automation is not a new concept for wireless networks . clearly networks already critically depend on
extensive use of automated processes. For instance, numerous examples abound just in the area of radio
resource management (scheduling, power and/or rate control, etc.) that demonstrate that these
automated features perform well. Thus, the appearance of SON algorithms represents a continuation of
the natural evolution of wireless networks, where automated processes are simply extending their scope
deeper into the network.


23GPPEVOLUTIONANDSON23GPPEVOLUTIONANDSON
2.1LTESONHIGH.LEVELSCOPEANDTIMELINE
Self-Organizing Networks capability is a key component of the LTE network. SON concepts have been
included in the LTE (E-UTRAN) standards starting from the first release of the technology (3GPP Release
8), and expanding in scope with subsequent releases (3GPP Release 9 and 3GPP Release 10). Figure 1
provides the standardization timelines for the different 3GPP LTE releases, throughout which SON
capabilities have been developed and enhanced.


Figure 1 - 3GPP LTE Specifications Timelines. Release 10 specifications were frozen in March
2011 and will be complete in Mid 2011.

A key goal of 3GPP standardization has been the ability to support SON features in multi-vendor network
environments. Therefore, a significant part of the SON standardization has been devoted to defining the
appropriate interfaces to allow exchange of common information which can then be used by each SON
algorithm. The SON specifications have been built over the existing 3GPP network management
architecture, reusing much functionality that existed prior to Release 8. These management interfaces are
being defined in a generic manner to leave room for innovation on different vendor implementations. In
addition to specifying the interfaces, 3GPP has defined a set of LTE SON use cases and associated SON
functions.1 The standardized SON features effectively track the expected LTE network evolution stages
as a function of time, following expected commercial network maturity. As such, the focus of the Release
8 SON functionality was put on procedures associated with initial equipment installation and integration to
support the commercial deployment of the first LTE networks, also known as “eNB self-configuration”.
These procedures included:

. Automatic Inventory

. Automatic Software Download2

.
Automatic Neighbor Relation3

.
Automatic Physical Cell ID (PCI) assignment4

Following this reasoning, the next release of SON (Release 9) provided SON functionality covering
operational aspects of already commercial networks, in particular key aspects related to network
optimization procedures. The Release 9 standardization scope included these additional use cases:


.
Mobility Robustness/Hand Over optimization
. RACH optimization
.
Load Balancing optimization
.
Inter-Cell Interference Coordination


The latest release of SON, being standardized in Release 10, provides a richer suite of SON functions for
macro and metro networks overlaid on and interoperating with existing mobile networks. It includes
enhancements to existing use cases and definition of new use cases as follows:

.
Coverage & Capacity optimization
.
Enhanced Inter-Cell Interference Coordination
.
Cell Outage Detection and Compensation
. Self-healing functions
.
Minimization of Drive Testing
. Energy Savings


The SON standards are a work in progress, and SON-related functionality will continue to expand through
the subsequent releases of the LTE standard, Release 11 and beyond, to cover all key aspects related to
network management, troubleshooting and optimization in multi-layer, multi-RAT heterogeneous
networks.

2.2SONDEVELOPMENTINNGMN
In 2006, a group of operators created the Next Generation Mobile Networks (NGMN) Alliance with the
objective to provide business requirements to the new technologies being developed. In practice, NGMN
recommendations provide guidance to the technical standards being developed for LTE, indicating the
key use cases that are most important for carriers’ day to day operations. These use cases have been
identified by the operators as the typical tasks that will be performed by their engineers in their day-to-day
operations; therefore, a better system of integration and automation would result in a more efficient
utilization of the operator resources, both material (spectrum, equipment, etc.) and human (engineering
time).

NGMN’s first whitepaper included high-level requirements for Self-Optimization network strategies5, and
sometime later a concrete set of use cases was defined, covering multiple aspects of the network
operations including planning, deployment, optimization and maintenance6. On the next page is a
definition of the top 10 use cases indicated by NGMN, many of which have already been introduced in the
3GPP standards.


1 Plug & Play Installation
2 Automatic Neighbor Relation configuration
3 OSS Integration
4 Handover Optimization
5 Minimization of Drive Tests
6 Cell Outage Compensation
7 Load Balancing
8 Energy Savings
9 Interaction home/macro BTS
10 QoS Optimization

NGMN uses cases are defined at a higher level than 3GPP specifications, and are introduced way ahead
of them, providing an initial guidance to the standards development and in some cases complementing
existing standard functionality. A specific example is the interaction between home/macro BTS, for which
NGMN has provided extensive recommendations to avoid excessive interference in the network (see
section 3.8.6 for more details) and which in part has been standardized in 3GPP Release 10.

Another NGMN use case not yet developed by 3GPP is the optimization of QoS related parameters. QoS
functionality is considered very important for wireless operators, especially with the deployment of very
high-speed networks where data consumption is steadily increasing and will be approaching that of fixed
broadband. Typically, the optimization of QoS related parameters governing the different network entities
(scheduler, admission control, mobility, etc.) is very complex and requires expert resources. NGMN
reckons that in the future, the networks should be able to adapt their quality automatically in response to
external factors such as high load in specific areas, or based on special traffic patterns. At this point, the
specific mechanisms to be used are not defined. However, NGMN has proposed a set of information
elements that should be made available to be able to easily define such mechanisms. Below is a set of
performance monitoring (PM) counters recommended for standardization:

.
Number of successful sessions per QoS Class Identifier (QCI)

.
Number of dropped sessions per QCI

.
Cell specific customer satisfaction rate

.
Min/Avg/Max throughput per QCI

.
Min/Avg/Max round trip delay per QCI

.
Packet loss per QCI


.
Mean number of Radio Resource Control (RRC) connected users

.
Mean number of RRC connected UEs with data to send per QCI

.
Percentage of UEs per cell that is not achieving their required GBR and not achieving the
required service data unit (SDU) error ratio per QCI

.
Percentage of UEs for which transfer delay per IP packet was above a particular threshold

.
Percentage of UEs for which average throughput measured at RLC layer for each non-real time
(nRT) QCI was below a particular threshold

.
Percentage of UEs per QCI for which the SDU error ratio is above a certain level

.
Number of RRC connected UEs with measurement gaps configured.

In addition to the definition of use case functionality, the SON group in NGMN has been very active in
defining Operations Support System (OSS) aspects of SON. These include support for more open OAM
interfaces, and procedures to support the integration of non-3GPP elements such as operator databases
and tools.

The objective of the OAM effort is to ensure a true multi-vendor ecosystem, where entities from different
manufacturers could operate together in an automatic fashion.16 For this purpose, a more open definition
of the Northbound Interface (Itf-N) in 3GPP has been requested to reduce the integration efforts between
the Network Management System (NMS) and the Element Management System (EMS), together with
real-time reporting requirements to allow a third party vendor to take timely action in the network. In
addition to generic requirements, the OPE recommendations define use-case specific OAM actions, as is
the case of the Handover Optimization and Cell Outage Compensation. Other aspects covered by this
project are the standardization of performance management information (counters and key performance
indicators or KPIs) and their exchange formats.

The integration of OSS tools is also a very active area in NGMN. In every network deployment, there is a
multitude of non-3GPP related elements that the operator needs to integrate with their existing network
elements, such as RF planning databases, work flow databases and optimization tools (cell planning
tools, third party SON tools, etc.). In order to integrate such entities, NGMN provides a set of generic
guidelines including the following:

.
Capability to run SON functionality in open (operator controlled) and closed (fully automatic) loop
modes

.
Deactivation of SON features in the Network Elements to pass control to the operator or external
entities

.
Support of centralized, distributed and hybrid architectures

.
Real time synchronization with NMS

.
Supply of relevant statistics and historical view to NMS

.
Customization of SON policies.


In addition to generic guidelines, NGMN provides specific use-cases requirements for each of the use
cases defined in 3GPP and NGMN. For example, in the case of the Automatic Neighbor Relation (ANR)
usecase, NGMN requirements for full OSS integration are listed below:

.
Support of CM Northbound Interface 3GPP BulkCM IRP (Bulk Configuration Management
Integration Reference Point). ANR based changes in the eNB shall be \"online\" synchronized with
EMS.

.
Support of real time relationship configuration to ensure that HO is possible a few seconds after
neighbor detection.

.
OSS should be able to indicate which entity (such as the eNB or some other entity) takes control
of the configuration for a given feature, say automatic neighbor relations (ANR).

.
OSS should provide monitoring support of the main ANR steps: neighbor cell detection, X2 setup,
neighbor cell configuration adaptation and ANR optimization

2.3SONARCHITECTUREALTERNATIVES
The specification covering the SON overview15 identifies three different architectures for SON
functionality: (a) Distributed, (b) Centralized and (c) Hybrid as shown in Figure 2.


Figure 2 . Different SON architecture approaches:
Centralized (left), Distributed (center) and Hybrid (right)


In a centralized architecture, SON algorithms for one or more usecases reside on the Element
Management System or a separate SON server that manages the eNBs. The output of the SON
algorithms namely, the values of specific parameters, are then passed to the eNBs either on a periodic
basis or when needed. A centralized approach allows for more manageable implementation of the SON
algorithms. It allows for usecase interactions between SON algorithms to be considered before modifying
SON parameters. However, active updates to the usecase parameters are delayed since KPIs and UE
measurement information must be forwarded to a centralized location for processing. Filtered and
condensed information are passed from the eNB to the centralized SON server to preserve the scalability
of the solution in terms of the volume of information transported. Less information is available at the SON


server compared to that which would be available at the eNB. Higher latency due to the time taken to
collect UE information restricts the applicability of a purely centralized SON architecture to those
algorithms that require slower response time. Furthermore, since the centralized SON server presents a
single point of failure, an outage in the centralized server or backhaul could result in stale and outdated
parameters being used at the eNB due to likely less frequent updates of SON parameters at the eNB
compared to that is possible in a distributed solution.

There are three key time intervals associated with Centralized SON.

.
The Collection Interval is the period during which statistics are collected and uploaded. This is
also the smallest available granularity for data analysis. This interval is most likely determined by
the vendors OAM statistics throughput limitations. Most Network Management solutions would
typically support a five minutes interval.

.
The Analysis Interval is the time period considered in the decision process for parameter
adjustment. It is beneficial to consider more than a single collection interval in the analysis. While
the latest collection interval should have the greatest impact on the analysis, the output should be
damped to take into account results from previous intervals.

.
The Change Interval is the period between changes applied to the network by SON. System
performance constraints may limit the number of cells for which changes are applied at any given
time. This could result in Change Intervals that do not align directly with the Collection Intervals.
These limiting factors don’t always apply, but centralized solutions either need to have vastly over
provisioned processing and networking capability, or intelligent change management.

In a distributed approach, SON algorithms reside within the eNB’s, thus allowing autonomous decision
making at the eNBs based on UE measurements received on the eNBs and additional information from
other eNBs being received via the X2 interface. A distributed architecture allows for ease of deployment
in multi-vendor networks and optimization on faster time scales. Optimization could be done for different
times of the day. However, due to the inability to ensure standard and identical implementation of
algorithms in a multi-vendor network, careful monitoring of KPIs is needed to minimize potential network
instabilities and ensure overall optimal operation.

In practical deployments, these architecture alternatives are not mutually exclusive and could coexist for
different purposes, as is realized in a hybrid SON approach. In a hybrid approach, part of a given SON
optimization algorithm are executed in the NMS while another part of the same SON algorithm could be
executed in the eNB. For example, the values of the initial parameters could be done in a centralized
server and updates and refinement to those parameters in response to the actual UE measurements
could be done on the eNBs. Each implementation has its own advantages and disadvantages. The
choice of centralized, distributed or hybrid architecture needs to be decided on a use-case by use case
basis depending on the information availability, processing and speed of response requirements of that
use case. In the case of a hybrid or centralized solution, a practical deployment would require specific
partnership between the infrastructure vendor, the operator and possibly a third party tool company.
Operators can choose the most suitable approach depending upon the current infrastructure deployment.


3KEYLTERELEASE8,RELEASE9ANDRELEASE10FEATURESSON3KEYLTERELEASE8,RELEASE9ANDRELEASE10FEATURESSON
3.1BASESTATIONSELF.CONFIGURATION
The deployment of a new network technology is a major investment for any service provider. In addition
to the spectrum and equipment costs, the operator faces multiple challenges related to the network
planning, commissioning and integration that often result in higher costs than the infrastructure equipment
itself. Today, there are a number of computer-aid design tools that an operator uses to simplify these
tasks, such as propagation tools, automatic cell planning (ACP) or automatic frequency planning (AFP)
tools. However, much of the process related to network element integration and configuration is still
performed manually. When a new base station (eNB) is installed, it requires that most aspects of its
configuration are provided by the engineer(s) on site, including the setup of the transport links, adding the
node to the corresponding concentration node (BTS or RNC), and establishing the connectivity with the
core network. This is in addition to the configuration of all the radio-related parameters such as the cable
and feeder loss adjustments, antenna type and orientation, transmit power, neighbor relations, etc. All
these processes are cumbersome, time-consuming, error-prone, and, in general, will require the
presence of more than one expert engineer, all the above resulting in an inefficient and costly process.

The objective of the Self-Configuration SON functionality is to reduce the amount of human intervention in
the overall installation process by providing “plug and play” functionality in the eNBs. As will be seen in
later sections, the scope of self-configuration functionality is expected to expand and evolve with
upcoming versions of the LTE standard.

3.1.1
BENEFITS
Self-Configuration of eNBs will reduce the amount of manual processes involved in the planning,
integration and configuration of new eNBs. This will result in a faster network deployment and reduced
costs for the operator in addition to a more integral inventory management system that is less prone to
human error.

3.1.2
DESCRIPTION
Self-Configuration is a broad concept which involves several distinct functions that are covered through
specific SON features, such as Automatic Software Management, Self Test and Automatic Neighbor
Relation configuration.

The Self-Configuration algorithm should take care of all soft-configuration aspects of the eNB once it is
commissioned and powered up for the first time. It should detect the transport link and establish a
connection with the core network elements, download and upgrade the corresponding software version,
setup the initial configuration parameters including neighbor relations, perform a self-test and finally set
itself to operational mode.

In order to achieve these goals, the eNB should be able to communicate with several different entities, as
depicted in the figure below.


DHCP/DNS Server
Configuration,
performance &
New eNB
Existing eNB
SGW
MME
OSS
DHCP/DNS Server
Configuration,
performance &
New eNB
Existing eNB
SGW
MME
OSS
Figure 3 - Self-Configuration of eNB in LTE

To be able to successfully achieve all functions the following prerequisites should at least be met prior to
the installation of the new node:

1.
A network planning exercise for the cell should have been completed resulting in a set of RF
parameters, including location, cell identities, antenna configuration (height, azimuth & type),
transmit power, maximum configured capacity and initial neighbor configuration. This information
should be made available in the configuration server.
2.
The transport parameters for the eNB should be planned in advance, including bandwidth, VLAN
partition, IP addresses, etc. The IP address range and Serving Gateway address corresponding
to the node should be made available in the configuration server.
3.
An updated software package download should be made available from the OSS.
The specific set of actions involved in the process will be covered in the next section.

3.1.3
SELF.CONFIGURATION
ACTIONS
The Self-Configuration actions will take place after the eNB is physically installed, plugged to the power
line and to the transport link. When it is powered on, the eNB will boot and perform a Self Test, followed
by a set of self discovery functions, which include the detection of the transport type, Tower-Mounted
Amplifier (TMA), antenna, antenna cable length and auto-adjustment of the receiver-path.

After the self-detection function, the eNB will configure the physical transport link autonomously and
establish a connection with the DHCP/DNS servers, which will then provide the IP addresses for the new
node and those of the relevant network nodes, including Serving Gateway, MME and configuration
server. After this, the eNB will be able to establish secure tunnels for OAM, S1 and X2 links and will be
ready to communicate with the configuration server in order to acquire new configuration parameters.


One of the OAM tunnels created will communicate the eNB with a dedicated management entity, which
contains the software package that is required to be installed. The eNB will then download and install the
corresponding version of the eNB software, together with the eNB configuration file. Such configuration
file contains the pre-configured radio parameters that were previously planned.

Note that at the time of the installation most of the radio parameters will have the default vendor values. A
finer parameter optimization will take place after the eNB is in operational state (self-optimization
functions). The configuration of neighbor relations can optionally be performed through an automated
SON functionality that is covered in a separate section of this paper, otherwise the initial setup will be
done according to the output of the network planning exercise.

After the node is properly configured, it will perform a self-test that will include hardware and software
functions, and will deliver a status report to the network management node. Also, the unit will be
automatically updated in the inventory database that will incorporate the unique hardware identifier, as
well as the current configuration and status of the node.

3.1.4
SELF.CONFIGURATION
STATUS
IN
3GPP
Current LTE standards incorporate functionality related to the self-configuration of eNB, including
Automatic Software Management,7 Self Test,8 Automatic Neighbor Relation,9 and Automatic Inventory
Management.10 It is expected that the first versions of the eNB self-configuration functionality in the eNB
will have vendor-dependent aspects, as 3GPP has not fully specified a standardized self-configuration
functionality. Examples of open areas in the standards include:

.
A defined interface between operator planning tools, equipment inventory and network
management entities

.
Configuration of transport parameters

.
Specific message formats for implementing the overall process

3.2AUTOMATICNEIGHBORRELATION(ANR)
One of the more labor-intense areas in existing radio technologies is the handling of neighbor relations for
handover. It is a continuous activity that may be more intense during network expansion but is still a time-
consuming task in mature networks. The task is multiplied with several layers of cells when having
several networks to manage. With LTE, one more layer of cells is added; thus optimization of neighbor
relations may be more complex. Even with the best methods at hand, due to the sheer size of large radio
networks . with several hundred thousands of neighbor relations for a single operator . it is a huge
undertaking to maintain the neighbor relations manually. Neighbor cell relations are therefore an obvious
area for automation, and Automatic Neighbor Relation (ANR) is one of the most important features for
SON. To explore its full potential, ANR must be supported between network equipment from different
vendors. ANR is, therefore, one of the first SON functions to be standardized in 3GPP.11

3.2.1
BENEFITS
ANR will remove, or at least minimize, the manual handling of neighbor relations when establishing new
eNBs and when optimizing neighbor lists. This will increase the number of successful handovers and lead
to less dropped connections due to missing neighbor relations.


3.2.2
DESCRIPTION
Target eNB
Serving eNB
S1
X2
SAE
UE
Figure 4 - Automatic Neighbor Relation (ANR) in LTE

The ANR in LTE allows automatic discovery and setup of neighbor relations when a user (UE) moves
from a serving eNB to another (target) eNB. ANR also automatically sets up of the LTE unique X2
interface between eNBs, primarily used for handover.

There are two LTE distinctive functions that make ANR possible:

1.
The UEs in LTE do not require a neighboring list and the reporting of unknown cells is fast
enough to be used during handover preparation. It enables ANR to receive handover
measurements on unknown cells that are not yet known by the serving eNB.
2.
The possibility for the eNB to request the UE to make a full identification of a cell. It allows eNB to
determine an unambiguous identity of a neighboring cell.
3.2.3
NEIGHBOR
RELATION
DISCOVERY
The UE is ordered to report measurements to the serving eNB directly after the RRC connection is set up

(i.e. is attached to the cell) and continues to do so while staying in RRC connected mode. The UE reports
all detected PCIs (Physical Cell Identities) . the short identity of the LTE cell . that fulfill the measurement
criteria set by the eNB at RRC connection. The UE may also measure on legacy radio technologies if it
supports multi-mode operation.
If there is an unknown cell included in the measurement report then ANR may begin actions to make the
cell known and potentially enable handover to the cell.

3.2.4
ANR
ACTIONS
If a PCI is reported by a UE that does not correspond to any of the serving eNBs’ defined neighbor cells

(i.e. it is not a neighbor cell), the ANR function in the serving eNB may request the UE to retrieve the
Global Cell Identity (GCI) of the cell with the unknown PCI in order to identify the cell. This cell is from
now called target cell (see figure 4 above). The UE reads the GCI, which is broadcast by the target cell
16


and reports it to the serving eNB. When the serving eNB receives the GCI, it can . with help from MME,
one part of SAE . retrieve the target eNB’s IP address, which makes it possible for the serving eNB to
contact the target eNB.

The serving and target eNBs are now in contact with each other and X2 can be setup. The serving eNB
requests X2 setup to the target eNB and includes all necessary cell data to create a neighbor relation (i.e.
PCI, GCI, TAC, PLMN-id and frequency) from the target cell to the serving cell. The target cell adds the
serving cell to its neighbor list and the target eNB sends the corresponding data for the target cell (PCI,
GCI, TAC, PLMN-id and frequency) to the serving cell which in turn adds the target cell to its neighbor list.

With the X2 interface in place, it is possible to use X2 for all future handovers between the cells. For
handover from LTE to legacy systems (i.e. GSM and WCDMA), ANR works in the same way with the
exception that it only needs to setup a neighbor relation to the target cell and not the X2 since the
handover to non-LTE systems is always performed over SAE.

ANR can automatically remove unused neighbor relations based on the relation usage, handover
performance or a combination thereof.

When adding and removing neighbors, ANR is under control of policies set by the operator. The black
listing allows the operator to decide neighbor relations that ANR may never add as neighbors. The white
listing allows the operator to decide permanent neighbor relations that ANR may never remove. These
policies are controlled from an Element Management System (EMS) such as OSS.

3.2.5
INTER.RAT
ANR
The inter-RAT ANR is applicable when different mobile network standards such as GSM, UMTS/HSPA,
CDMA/EV-DO and LTE are deployed to cover the same geographical area.

In this section, the ANR Inter-RAT function is explained for the case when LTE is overlaid on one of the
other networks, hereafter referred to as underlying networks. LTE is normally deployed on the same sites
as the underlying network’s radio base stations. It will therefore partially or fully cover the same area as
the underlying network. Inter-RAT handover between LTE and an underlying network could be used for
circuit switch (CS) fallback to the existing networks for voice calls and SMS, and load balancing of UEs
when traffic on a network exceeds a threshold. This means that the UEs could handover between LTE
and the underlying network even when the LTE cell coverage is good. Based on the above assumptions,
the ANR function in LTE must be able to find the underlying network’s cells with the same coverage as
the LTE cells. To find these neighbors, an additional ANR method in LTE needs to be introduced whereby
the UEs are ordered to report best serving cell periodically.

The ANR function in LTE orders the ANR capable multi-standard UEs, connected to LTE, to report best
serving cell measurements from the underlying network periodically. There are different possible solutions
for ordering the multi-standard UEs depending on which technology is used in the underlying network.
The new SON unique report configuration, “ReportStrongestCellsForSon” could be used for UMTS/HSPA
and CDMA/EV-DO. The existing “ReportStrongestCells” could be used for GSM. They are specified in
3GPP 36.331 (Release-8) and are used for specifying criteria for triggering of inter-RAT measurement
reporting events B1 and B2. Event B1 refers to the event when inter-RAT neighbor has a higher value of
RSRP/RSRQ than a threshold. Event B2 refers to the event when the serving cell has a lower of
RSRP/RSRQ than a threshold, and inter RAT neighbor has a higher value than a threshold.


To minimize the load on the UEs, the ANR function in LTE should be able to adapt the number of UE
measurements to the actual need. In the beginning, when many new cell relations are found, a more
intense measurement activity is used. This enables a quick creation of necessary neighbor relations. After
some time when less neighbor candidates are found, the ANR function should involve fewer UEs in the
measurements. Even when new neighbors are not detected at all or very rarely, it is important that ANR
maintains some minimal level of periodical measurements. This enables the ANR function to rapidly
detect any network changes.

The LTE ANR procedure following the discovery of new neighbors is similar to the one for intra-LTE,
described earlier. After the UE reports a strongest cell (identified with its PCI) that is not listed as a
neighbor in the neighboring cell list, the ANR function on the eNB will add it to the neighboring cell list for
the cell it was discovered in. Before doing so, the eNB needs to know the unique identity of the cell i.e.
the Cell Global Identity, CGI. It can add the missing neighbor cell relation immediately if the cell CGI is
already known by eNB. Otherwise the ANR function in the eNB first needs to identify the cell by finding
out its CGI. In this case, ANR orders a CGI measurement from the UE, and once it is reported back, the
new neighbor is added.

After the new IRAT neighbor is added to the eNB cell, it can be used for mobility handovers and for the
other cases where neighbor cells are used. It should be noted that the neighboring cell relationship is only
one way i.e. from the LTE cell to the cell in the underlying network. The underlying network’s ANR
function also needs to discover the LTE cell. A manual procedure could be used where the operator,
when notified about the new neighboring cell relationship, adds the other direction manually or has a
centralized solution that automates the procedure.

3.3TRACKINGAREAPLANNING
Wireless networks partition a typical market into non-overlapping Traffic Areas (TAs). Each TA is uniquely
identified by the TA Identifier (TAI). Each and every User Equipment (UE) in the Power-ON state is
mapped to one (or more) TAs. TAs were constructed to facilitate the Paging procedure. Note that
whenever the switch (MME) receives a call for mobile M, it looks up the TA of mobile M . as TA(M)) .
sends a page to all the eNBs in TA(M). Each eNB faithfully broadcasts the message on the Paging
channel, which is received by UEs in Power-ON mode. When mobile M receives the page, it realizes that
there is a terminating call (data transfer) for it, and it sends a paging response to its serving eNB .
eNB(M). The eNB(M) responds with an affirmative to the switch, which goes on to direct the call towards
eNB(M). Subsequently, call setup procedure is followed between mobile M, eNB(M), and MME/S-GW.

In order to ensure that the MME has the most recent information for each mobile in terms of its current
TAI, all UEs are required to provide TAU as soon as they realize that their current serving eNB has a
different TAI. Such an update is sent on the Random Access Channel (RACH). (Border eNB’s are
basically the eNBs which are on the border of a TA). Such a structure leads to a tradeoff between the
RACH and paging channel. Observe that if each TA is kept small, then a moving mobile would cross
through many TAs, and would need to make a random access attempt in one of the Border eNB of each
TA. However, if the number of eNBs in a TA is large, then the RACH load on the Border eNBs would be
less, but each terminating call/data transfer to a mobile M would lead to a broadcast of paging message
from the MME to each eNB in that TA. This would certainly put additional pressure on the backhaul link.
Additionally, on each of the eNBs, a page will also need to be sent using up the paging channel.
Therefore, determining TA size and demographic is a tradeoff between the RACH load on the Border
eNBs and the paging load on the backhaul and RF of the eNBs. Note that the RACH load affects only one
cell, but the paging load translates into a broadcast message on all the eNBs belonging to that TA.


3.3.1
BENEFITS
Present day wireless operators have been forced to take an offline approach due to lack of any
mechanism for effective and efficient adjustment of tracking areas. Due to the cumbersome nature of
such a process, most carriers hardly change the tracking areas of their cells. In other words, TAIs for
each cell are decided at the time of deployment based on rules-of-thumb, anticipated traffic patterns, etc.,
and are only altered in the event of extreme performance degradations. SON TA feature has the ability to
change that, both at the time of deployment using Tracking Area Planning (TAP) and during the
subsequent network optimization using Tracking Area Optimization (TAO).

At the time of deployment, TAP algorithm prepares the initial deployment plan for the cell sites of a
market in an (semi) autonomous fashion. The output of the TAP drives the choice of tracking area that an
eNB belongs to. The corresponding TAI is delivered to each eNB during the initialization phase. The
inputs to such a deployment plan could be market geographical data, TAI range and values, eNB
locations, market size, etc.

Once initial deployment is complete, the TAO algorithm actively monitors the Tracking Area Updates
(TAU) and the load on the radio access channel (RACH) to continuously identify the eNBs that are most
suited for a change in their TAI. The intention is to capture some of the mobility patterns for each eNB.
For example, if a highway passes through a cluster of eNBs, it might make sense to ensure that tracking
area boundary cleanly intersects with the highway, and avoids a UE in the car to ping-pong between
multiple TAs. TAO algorithm has the ability to identify such eNBs and allocate them to appropriate TA.

3.4PCIPLANNING
In order for the UEs to uniquely identify the source of a receiving signal, each eNB is given a signature
sequence referred to as Physical Cell ID (PCI). Based on the LTE specification of the physical layer
detailed in 3GPP TS 36.211-840, there are a total of 504 unique physical layer cell identities. These
physical layer cell identities are grouped into 168 unique physical layer cell identity groups, where each
group contains three unique identities. The overall signature PCI is constructed from primary and
secondary synchronization IDs as follows:

(1) (2)
PCI
.
3N
.
N

ID ID

(1) (2)
Where, N is in the range of 0~167, representing the physical layer cell identity group, and N is in

ID ID
the range of 0~2, representing the physical layer identity within the physical layer cell identity group. The
hence constructed PCI is allocated to each eNB at the time of installation. Based on the allocated IDs, the
eNB transmits the PCI on the downlink preamble. The UEs in its service area receive the preamble, and
are able to identify the eNB, and the corresponding signal quality. It is possible, however, that a UE finds
that there are two eNBs that have the same PCI. This is possible since the PCIs are reused by multiple
eNBs. Note that there are only 504 PCIs, and a typical market might have 200 to 300 cell sites, assuming
three eNBs per cell site leads to as many as a thousand eNBs in a market. Therefore, the service
provider must carefully determine the PCI of each eNB to make sure that such conflicts do not happen, or
are minimized.

Typical operators use an offline planning tool or depend on manual determination to develop a PCI
deployment plan for a market. The plan uses basic information such as eNB location, potential neighbors,
etc., to determine the PCI for each eNB. Such an allocation is carefully reviewed to ensure that the
market does not have any PCI conflicts; hence the determined PCI values are communicated to each


eNB during the installation using the configuration files or manually inputted by the staff. Needless to say,
such a process does not lend itself to subsequent changes and is prone to human error.

3.4.1
BENEFITS
SON mechanisms enable the operator to automate this tedious process described above in section 3.4.
In the SON framework, as soon as the eNB is powered up during the auto-configuration phase, it is
allocated to a PCI (that is a primary and a secondary synchronization ID). Such a PCI is determined using
a PCI Planning Tool (PPT) that not only uses the estimated coverage area information for each eNB, but
also enforces significant margin and separation between two eNBs that are allocated to the same PCI.
Additional considerations could also be included when determining such a plan. Nonetheless, SON
ensures that each eNB has a PCI value at the time of installation without requiring explicit human
intervention.

Subsequently, during the operational phase, each eNB collects the information pertaining to any PCI
conflicts. Observe that PCI conflicts might happen due to errors during the initial PCI Planning phase,
deployment of new eNBs, changes in the demographics of a market, power of eNBs, etc. Whenever an
LTE UE receives power from two eNBs with the same PCI, it informs the serving eNB about the conflict.
Such an alarm is relayed to the OSS/SON mechanism, which collects and logs the details of such
conflicts. The operator can then decide on a suitable time interval for activating the PCI Optimization Tool
(POT), (e.g., it might make sense to schedule such an activity during a lightly-loaded night-time period).
The POT algorithm uses the collected logs, alarms and the updated coverage maps in order to identify
the eNBs for which the PCI needs to be changed and the associated new PCI value. Furthermore, the
SON algorithm ensures that the information is relayed to the correct eNBs. Upon reception, eNBs could
wait for a hold period before they begin to deploy the newly allocated PCI values.

3.5LOADBALANCING
Load Balancing refers to the process whereby similar network elements that are intended to share traffic,
share the load. The similar network elements can be anything from packet gateways to MMEs to base
stations and sectors. In LTE, MME pools are expected to share user traffic load across different MMEs as
load increases, while eNBs may have RRM functions that share/offload traffic to neighboring cells in order
to increase system capacity. As a result, different real-time algorithms at different nodes can
simultaneously provide Load Balancing of user traffic per network element as required. Additionally, long-
term traffic behavior of each node can be monitored so that traffic may be “directed” a-priori by a
centralized entity in the network. For instance, this could be a desirable feature for markets where
periodic or scheduled concentrations of users regularly occur (e.g. sporting events, conventions, daily
commutes, etc.).

The decision to re-balance a cell or move a particular user must take in to consideration the target for the
user(s). It is not desirable to send a user to an alternate location (i.e. neighbor or co-located frequency) if
that user will then have a reduced QoS or lower performance than remaining in the source, or if the
resulting re-balance will result in reduced system capacity/utilization.

3.5.1
BENEFITS
The objective of Mobility Load Balancing is to intelligently spread user traffic across the system’s radio
resources as necessary in order to provide quality end-user experience and performance, while
simultaneously optimizing system capacity. Additionally, MLB may be desirable to shape the system load


according to operator policy, or to “offload” users from one cell or carrier in order to achieve energy
savings. The automating of this minimizes human intervention in the network management and
optimization tasks.

3.5.2
DESCRIPTION
The term Mobility Load Balancing (MLB) is used in this section to refer specifically to the network cell
(eNB) level only, not core entities such as the MME, gateways, etc. The goal of MLB is to spread user
traffic across system radio resources in order to provide quality end-user experience and higher system
capacity. This can be accomplished by one or a combination of algorithms that perform Idle or Active
balancing of users. These SON algorithms for offloading traffic from one element to another can include
intra-carrier, inter-carrier, or inter-technology resources, as long as there is software intelligence to ensure
radio admission and continuity of service on the target element. The actual transfer of users is
accomplished by modification of handover threshold parameters. This can require coordination with
competing SON algorithms and standardized messaging with multi-vendor equipment to ensure
robustness and stability.

LTE is better suited to a distributed algorithm utilizing the X2 interface, while technologies with a
BSC/RAN architecture and/or macro-diversity may favor a more centralized approach. The text in this
section is suited to single-link technologies such as LTE.

.
Distributed LB: Algorithms run locally in the base stations. Load information is exchanged
between base stations so that Idle/Active HO (handover) parameters may be adjusted and/or
adjustments to RRM functionality can be made.

.
Centralized LB: Algorithms run in a core network element. Base stations report load information
to a central entity which then responds with appropriate modifications to idle/active HO
parameters.

In either case (distributed or centralized), it is assumed there will be centralized Operations,
Administration and Management (OA&M) control for an operator to enable/disable and configure relevant
algorithm settings.

3.5.3
DETERMINING
A
LOAD
IMBALANCE
CONDITION
Load balancing mechanisms must work together with the scheduler and admission control. For non-
Guaranteed Bit Rate (GBR) users, there is no constraint on the minimum performance those users
receive except within the scope of the maximum number of users per cell (admission control) and
perhaps a vendor-imposed minimum throughput (scheduler). For GBR users, the scheduler must ensure
that all radio bearers are granted resources in a manner that satisfies their specific service. Therefore, a
system may be considered “in balance” as long as there are no users being denied resources and all
active services are being supported within the scope of their QoS needs.

Simple thresholds can be implemented where low, medium and high load conditions equate to a given
number of active users in the cell for the non-GBR case. These can serve as triggers to modify idle mode
parameters and/or to handover active users to neighbors (i.e. cell-edge intra-carrier, collocated inter-
carrier or collocated inter-technology handover). However, more intelligent metering is needed for GBR
users since it is possible for a small number of such users to “load” a cell depending upon their
requirements.


3.5.4
IDLE
MODE
LOAD
BALANCING
The LTE system does not have a real-time, per-cell view of idle mode users. The only time the system
becomes aware of the exact cell a user is in, while in idle mode, is when the Tracking Area of the user
changes and a TAU message is sent by the UE. Therefore, while parameters that control how and when
a UE performs cell reselection (idle handover) are modifiable, there is no direct measurement mechanism
for the system to determine when there are “too many” idle users. Note that this “too-many idle user
condition” has no direct bearing on either system capacity or user experience besides increased signaling
on core network nodes.

The way around this immeasurable condition is for the system to adjust cell reselection parameters for the
idle users based on the current active user condition. As real-time traffic and/or QoS demands increase in
a cell, it would be possible for the cell to adjust the cell reselection parameters in order to force users
nearest the cell edge to select their strongest neighbor to camp on, or to force a handover to a co-located
carrier that has more resources available.

Care must be