Академический Документы
Профессиональный Документы
Культура Документы
17:25
URBAN
RISE
RURAL
ENTERP
HOME
& REMO
TE
VIRTUAL
IZATIO
N
DOCUMENT
077.07.01
SON use cases
February 2014
We are not a standards organization but partner with organizations that inform
and determine standards development. We are a carrier-led organization. This
means our operator members establish requirements that drive the activities
and outputs of our technical groups.
Today our members are driving solutions that include small cell/Wi-Fi
integration, SON evolution, virtualization of the small cell layer, driving mass
adoption via multi-operator neutral host, ensuring a common approach to
service APIs to drive commercialisation and the integration of small cells into
5G standards evolution.
The Small Cell Forum Release Program has now established business cases and
market drivers for all the main use cases. This document is part of
Release 7: HetNet and SON.
All content in this document including links and references are for informational
purposes only and is provided “as is” with no warranties whatsoever including
any warranty of merchantability, fitness for any particular purpose, or any
warranty otherwise arising out of any proposal, specification, or sample.
If you would like more information about Small Cell Forum or would
like to be included on our mailing list, please contact:
Email info@smallcellforum.org
The target audience for this document is systems and network deployment engineers
as well as designers working on SON solution architectures and SON algorithms.
Urban small cells are used both as a capacity layer and for overcoming coverage
holes. They are deployed either outdoors or indoors and, if indoors, in both open and
closed subscriber group modes depending upon the need for the deployment, the
subscription of the end user and other factors. They tend to differ from enterprise
deployments in that they are less planned and offer less opportunity for human
intervention and optimization. They also have a broader charter than strictly
residential small cells and are likely to see a much wider variation in deployment
scenarios (and thus different radio conditions, coverage areas, modes and other
features). This means that the need for a SON function is even more acute.
The background documents [1] and [2] provide recommendations for deployment,
configuration and operation of enterprise small cells and the SON use cases in the
enterprise context. This document covers the SON function as it pertains to urban
small cell networks. The document aims to highlight, in as much detail as possible,
without infringing or architecture or algorithm choices, use cases for the SON function.
The broad goal of SON is to enable the configuration, deployment, optimization,
operation, maintenance and healing of dense networks across multiple scales with
very low or no manual input or feedback; this goal is an underlying theme throughout
the document.
1. Urban SON use cases addressing the two predominant radio access
technologies — UTRAN and EUTRAN
2. SON use cases addressing the full life cycle of the network — from initial
deployment and configuration through self-optimization as well as
maintenance aspects
3. SON functionality specifically addressing the 3GPP-identified areas such as
self-configuration, self-optimization (neighbor discovery, parameters
(PCI/PSC) optimization, resource sharing for interference management and
capacity optimization as final goals, load balancing, robust handoff and
random access channel (RACH) parameter optimization) and self-healing.
The real benefit is, of course, to both the network operators and users in terms of
reduced capex and opex for the same or better level of service.
Figures
Figure 2-1 Example of a centralized (canonical) NM-based SON architecture for
multi-vendor, multi-RAT networks..................................................... 4
Figure 2-2 Example of a distributed SON architecture ......................................... 5
Figure 2-3 Example of a localized SON architecture ............................................ 6
Figure 2-4 Generic hybrid SON architecture ....................................................... 7
Figure 3-1 SON features.................................................................................10
The concepts of small cell networks and enterprise small cell networks are covered in
[1].
The need for and the functionality of SON functions are extensively covered in
literature (see for example LTE Self-Organizing Networks [ 3] and references therein).
Specifically, from an LTE perspective, SON function scope and coverage within the
3GPP context are addressed in Self-Organizing Networks: Concepts and requirements
[ 4] (and references therein).
The SON function is not a specific node or element in the network. Rather, as the
name implies, it is a functional entity, or set of functional entities, that addresses the
specific needs discussed in the references pointed above. In an architecture and
algorithm instantiation sense the function lends itself to a large number of
implementation options.
The various architectures and how they apply to each layer are expanded on in
Chapter 2.
This document draws heavily upon its predecessor, which addresses enterprise SON
use cases [2]. There is a significant amount of similarity and overlap. However, some
of the differences are:
The scope of this document is narrowly to focus upon the SON function addressing the
urban deployments and even more specifically illustrate the SON function through
properly designed use cases, whether they be single RAT (intra and inter-frequency)
or inter-RAT (LTE and UMTS).
Since the document and specifically sections to follow are heavy on acronyms, the
following glossary is added for convenience to the reader.
Term Description
ACS Auto-configuration server
ANR Automatic neighbor relation
CN Core network
DL Downlink (from the base station to the user equipment)
eNB e Node B (3GPP standard name for LTE base stations)
EMS Element management system
EPC Enhanced packet core
IP Internet protocol
IT Information technology
LTE Long term evolution
MME Mobility management entity
MSC Mobile-network switching center
NB Node B (3GPP Standard name for UMTS base stations)
NLM Network listen module (sometimes called a sniffer). this may be located in the
urban small cell and/or in the LTE UE.
NMS Network management system
NRT Neighbour relations table
PCI Physical cell identifier (PCI in LTE)
PLMNID Public land-mobile network identifier
PSC Primary scrambling code (in UMTS)
RSI Root sequence index
S1 An LTE standards defined interface between eNB and EPC. It has two ‘planes’ —
a user plane and an MME plane.
SC Small cell
SGSN Serving GPRS support node
TNL Transport network layer
TX Transmit or transmitter
UL Uplink (from the user equipment to the base station)
UMTS Universal mobile transmission system
X2 An LTE standards-defined interface between eNBs.
Table 1–1 Glossary of terms
In general, the following four different architectures can be identified for SON
functionality: centralized, distributed, localized and hybrid.
In order to minimize the impact of slow updates to SON parameters, depending on the
use cases, the interactions between SON algorithms can be allowed through an
intelligent management entity prior to active modification of SON parameters.
Additionally, operators often greatly over-provision the centralized SON servers for
enhanced processing and greater network capabilities. An accurate analysis of
cost/performance issues — as well as a high tolerance to failures — needs to be
adequately addressed to avoid overloading of the centralized SON controller.
Although a distributed architecture offers less control for operators over SON
parameters, this approach provides better scalability, less latency and reduced load on
NMS. It also allows for ease of deployment in mixed-vendor scenarios. Most EMS
vendors expose their APIs that allow the retrieval and the setting of their information
elements that can propagate the parameters down to the network elements. The
distributed architecture can also make the benefits of self-management functions for
enhanced robustness avoiding single points of failure.
In localized SON architecture, the algorithms are distributed locally within individual
network elements at a relatively low level in the architecture. In this approach, the
SON functionality is entirely autonomous — that is, the decisions are made
autonomously and operators have (in most cases) no control over SON functionality.
Autonomous SON configuration for local self-optimization loops such as enhanced
plug-and-play functions, hand-over thresholds and energy saving modes can be
executed with low overheads and latency on the devices.
This architecture is the most scalable option among the possible SON architectures.
For frequent local SON functions, tight coupling topologies are required in order to
reduce the delays in communications between levels of hierarchies. Localized SON
provides the fastest response time to short-term (<1ms) functions such as (inner-
loop) power control algorithms. In localized SON architecture, well-defined interfaces
between SON function entities are not required.
The most efficient SON realization for an urban HetNet is hybrid architecture.
Scalability and real-time response of SON functions must be supported through a lean
and reliable management framework. This architecture can best exploit the benefits of
SON functions and smoothly integrate with operators’ existing practices. The hybrid
SON can be distributed at small cells, element management systems or core networks.
This architecture is, in effect, a trade-off between the centralized and distributed
topologies.
In a hybrid approach, part of a given SON algorithm can be executed in operation and
maintenance (OAM) servers (or NMS) while another part can be executed in small
cells. The existence and cooperation of SON functions both on the RAN side and on the
backend management side can allow a high degree of automation. For example, the
values of the initial parameters, boundary conditions and operators’ policies could be
configured in a centralized server, and updates and refinement to those parameters in
response to the actual UE measurements could be done in the (e)NBs. Figure 2-4
shows a generic (canonical) hybrid SON architecture for a complete multi-RAT, multi-
vendor and multi-layer HetNet.
Operators can migrate from a centralized to a distributed SON option as they gain
more confidence about multi-vendor IOT. In general, distributed or hybrid SON
approaches are favored in coordinated networks since they enable lower delay, less
signaling, and lower cost, even though they may risk losing some performance gains
compared to their centralized counterparts. Certain SON functions such as mobility
load balancing can have an improved performance in a coordinated and
hybrid/distributed SON architecture. In general, hybrid SON architectures require
more complicated mechanisms for implementing various information elements. Often
in multi-operator/multi-vendor deployment scenarios, bilateral or multilateral
agreements need to be facilitated by operators/vendors.
SON
# Description Pros Cons
architectures
One-to-many More manageable Less scalable
topology implementation Large amount of
Typically, Often preferred in signalling information
resides in multi-layer/multi- needs to be transported
OAMs, vendor networks due to to centralized SON
controllers or coordination difficulties servers
servers between macro-cells Slower
OSS-centric and small cells as well reaction/response time
approach as lack of IOT between to network dynamics
1 Centralized mixed vendors Potentially less accurate
SON servers often need
to be over-provisioned
Need to address
cost/performance and its
tolerance to failure
issues to avoid
overloading of SON
controller
Many-to-many Minimizes risk of In inter-RAT access
topology network instability points, new interfaces
Resides in Increased scalability need to be defined
multiple Less latency Operators have less
2 Distributed locations Reduced load on NMS control over SON
Algorithms are allowing ease of parameters
typically deployment in mixed-
executed within vendor scenarios
coordinated Enhanced robustness
This document describes use cases of activities involving SON features, and uses the
term ‘SON function’ to describe the logical entity which provides the algorithms, data
storage and interactions with other nodes to perform SON activities. This term is
deliberately non-specific about where this SON function is located, and thus
encompasses a number of options. That is, this entity could be using the centralized
SON model, the distributed SON model, the hybrid SON model or the localized SON
model, as described in Self-Organizing Networks: Concepts and Requirements [4] and
Section 2. In some of the use cases, it is clear that there must be some distributed
element to the SON function (for example, when SON-related information is to be
exchanged over X2 with other small cells) for the use case to be as described, and in
other use cases it is clear that there must be some centralized SON function (for
example, for holding centralised data and applying central control of multiple small
cells) is inferred.
The use of the term ‘SON function’ is also not specific about how much influence the
small cell SON features may have on the operation of the macrocells. The macrocells
may be totally autonomous, taking no account of the small cell network in their
management of load and interference (and possibly including separate autonomous
macrocell SON functions), or there may be SON function that integrates the SON
features of the small cells with SON features of the macrocells — for example,
balancing the load and interference between all the cells within the network. This level
of interaction clearly affects the complexity and ability of the SON function, but does
not really affect the use cases described here, as each small cell interacts with the
SON function. What that SON function does (except in very broad terms) to arrive at
the configuration or commands given to the small cells (and, where applicable,
macrocells) is outside the scope of this document.
The overall SON function may be considered as a set of smaller features, some self-
contained, others interacting with each other. This section describes some of those
constituent SON features. The SON features are broadly classified into three groups:
self-configuration, self-optimization and self-healing. The specific set of features that
goes into each group is captured in detail in the use cases section.
Self-optimization features encapsulate all the processes, actions and steps wherein UE
and eNB measurements, event reports and performance measurements are used to
auto-tune the network. These features are deployed in the operational state. The
operational state is understood as the state where the RF interface is switched on and
the cell is providing service to users.
Processes, actions and steps wherein the SON function attempts to diagnose/repair
malfunctioning small cells or optimize the network around a failed small cell are
categorized as self-healing features.
a- 1 : configuration of IP address
(A ) Basic Setup and detection of ACS
Configuration
b- 2 : coverage/ capacity related
parameter configuration
Self- Optimisation
( operational state )
(C ) Optimization/ c- 1 : neighbour list optimisation
Adaptation
c- 2 : coverage and capacity control
The steps in the figure are representative and not exhaustive. The detailed steps are
brought out in the use cases in the next section.
In the following sections the use cases are enumerated into specific categories.
Specifically the topics covered are configuration, planning, deployment and operation,
optimization and maintenance. Care is taken to point out whether the use case
specifically applies to UMTS or LTE or both. Where necessary, inter-frequency issues
but within the same RAT are also highlighted.
The urban and enterprise SON use cases have several scenarios in common. Table 4–1
provides a comparison between urban and enterprise SON use cases where the key
differences are noted.
4.1 Configuration
Overview Many new small cells are added to the network at the same time as a
group
Actors New small cells, ACS, SON function, optional small cell gateway, NLM, an IT or
operator technician.
Applicability LTE and UMTS
Preconditions A precise or approximate plan for the deployment of multiple small cells in the
given location. Available power and backhaul for the small cells (such as DSL or
P-MP radio), a gateway and ACS/EMS server accessible via the backhaul. An
installer and/or a technician working for, or under the direction of, the mobile
network operator
Flow 1. The installer/technician places/installs multiple small cells at the positions
identified in the plan for that site
2. The small cells are provisioned with the ACS and SON function details
3. The small cells connect to the ACS and SON function(s), are authenticated,
and obtain their configuration from the ACS
4. As part of configuration, the small cells are configured to belong to a
common cluster and initialized to be in a training mode (see notes 1 and 2)
5. Optionally, NLM is switched on, neighbor network parameters (DL/UL
frequencies, PCI (PSC) etc.) are discovered and NLM data sent to SON
function(s)
6. The SON function(s) can reside in individual small cells, one small cell
and/or an ACS-like server
7. The SON function(s) assigns temporary or permanent radio parameters —
for example, PCI (list) (PSC (range/list)) or RSI, with consideration also to
any co-channel macro (e)NodeBs that are visible to the small cell (taking
into consideration any policies provisioned into the SON function).
8. The small cell connects to the MME (MSC and SGSN) in the EPC (CN)
(through the gateway or directly)
9. Optionally, LTE small cells perform TNL address resolution for NLM-
discovered and ACS-provisioned neighbors or are provided by their SON
functions with neighbor address information
10. Optionally, the small cells establish X2 (Iurh) links to any neighbors for
which TNL addresses are available (whether discovered or provisioned)
11. The small cell transmitters are turned on and small cells begin to operate in
training mode.
12. The installer/technician walks around the site with a UE in connected mode
or with some other tool to collect information about RF environment at the
site (such as signal quality of the small cells or surrounding macros)
13. The collected information is either passed autonomously or with
installer/technician assistance to the SON function(s)
14. At the end of the walk-around, SON function(s) is instructed to determine
transmit power for the small cells
15. The SON function determines TPC parameters using obtained RF
information considering any co-channel macros and UL interference issues
to provide good coverage and capacity at the site
16. The small cells start radiating with the updated Tx power and begin normal
operation
17. Optionally, installer/technician performs additional walk-arounds to verify
coverage and mobility at the site. Based on this second walk-around, SON
function(s) may raise alarms or show error codes suggesting required
changes in the deployment. In such a case, the installer/technician may re-
do the calibration after taking remedial measures
Uses
Notes 1. Training mode is meant to indicate vendor-specific extensions that
(optional) help the equipment (small cell and/or UE) to be put in states that
Overview A small cell in a network along with ACS/SON server determines that
another small cell (or macro) is no longer a neighbor
Actors Small cell, SON function, optional NLM
Applicability LTE and UMTS
Preconditions A properly situated, connected, authenticated, provisioned and functional small
cell in an operator’s network. A SON function that is provisioned with the
proper policies and is in operation. Optionally, a functional NLM
Flow 1. A small cell in operation does not receive UE measurements regarding a
PCI (PSC)/(E)CGI for a configurable period of time
2. Alternately the ACS server or SON function decommissions a small cell
3. If it is a measurements/timeout-based determination, then the SON
function determines that another cell is no longer a neighbor
4. The small cell is commanded by the ACS server or SON function to remove
the neighbor
5. The small cell tears down X2 (Iurh) connections if necessary
6. The PCI (PSC) / (E)CGI are removed from the local and global NRT
Uses
Notes As a result of the NRT update, the small cell SIB messages may need to be
(optional) updated by RRM
Table 4–10 Deleting a neighbor in LTE and UMTS
4.3 Optimization
Overview An existing small cell is assigned a new PCI (PSC) due to discovery of a
PCI (PSC) conflict
Actors Small cell, SON function, optional NLM
Applicability LTE and UMTS
Preconditions A properly situated, connected, authenticated, provisioned and functional small
cell in an operator’s network. A SON function that is provisioned with the
proper policies and is in operation. Optionally, a functional NLM
Flow 1. An existing small cell discovers a new neighbor and vice-versa (for
example, through A4 reporting, X2, or NLM)
2. The SON function determines a PCI (PSC) conflict or confusion after these
events
3. Based on the network topology and PCI (PSC) policies (such as the pool of
PCIs (PSCs) available for the small cells), the SON function computes or
selects a new PCI (PSC) value for one or more small cells to avoid the
conflict
This use case addresses load balancing across macro and small cell layers, in
particular using some of the methods available in LTE. Based on the load metrics
available from both layers, the SON function could allow UEs to camp on small cells
preferentially for macro offload. However this needs to be coordinated with
appropriate resource reservation/assignment of protected resources on small cells to
manage interference and enable reliable connections to these UEs. To achieve this, a
SON function can determine several parameters including RSRP bias to small cells,
time domain (ABS) and possibly frequency domain resource partitions.
Overview Small cell and macro partition resources to help balance load and
manage co-channel interference
Actors Small cell, macro, SON function, optional NLM
Applicability LTE
Preconditions A properly situated, connected, authenticated, provisioned and functional small
cell in an operator’s network. A SON function that is provisioned with the
proper policies and is in operation. Optionally, a functional NLM. Small cell(s) is
deployed in the same frequency as macros. For time domain partitioning,
Flow 1. The small cells in operation measure their own traffic load and interference
levels and report to the SON function. The load reports and triggers may be
separately defined for inter-RAT load balancing functionality.
2. The SON function reviews RAT/frequency-specific load information of the
small cells and determines whether certain small cells are overloaded
and/or heavily interfered based on available air interface or backhaul
resources
3. The SON function conveys its decision/configuration to small cells to
perform SON activities
Adjust the related inter-RAT handover parameters on small cells to offload
some traffic from busy small cells to less busy small cell(s) on a different RAT
Notes 1. SON function may also aggregate load metrics from a macro layer and use
(optional) these to determine handover parameters for the small cells
2. The architectural and procedural support for inter-RAT connected and idle-
mode load balancing are discussed in detail in sections 3.1.2 and 3.3.1.1 of
SCF document 088 [6]
Table 4–18 Inter-RAT load balancing
4.4 Maintenance
Urban deployments will be a key part of small cell deployments and therefore a key
part of small cell adoption and their success. SON functions are especially important in
making the planning, deployment, optimization and management of these urban
networks streamlined, so that they function the best they can — more so, in fact, than
in the enterprise case. If solutions meet the baseline use cases in this document, this
goal can be met.