RAN Sharing - HQoS Guidelines - Ver. 3.1
RAN Sharing - HQoS Guidelines - Ver. 3.1
RAN Sharing - HQoS Guidelines - Ver. 3.1
RAN17.0 INTERNAL
Huawei Technologies Co., Ltd. Product Name Pages
Prepared
renliang 00116371 Date 2014-7-21
by:
Reviewed
Xu Xiaoze 56296 Date 2014-8-10
by:
Approved
Fang Xiang 00160457 Date 2014-8-15
by:
Revision Record
2009-12-01 1.0 Updated this document based on RAN10.0 and Xu Xiaoze 00059296
RAN11.0.
2010-12-07 2.0 Added the MOCN enhancement feature introduced Xu Xiaoze 00059296
in RAN13.0.
2012-04-13 2.1 Updated the contents related to the feature solution Ji Di 00127658
and service design in RAN13.0.
2012-05-30 2.2 Based on review opinion, achieve the final Ji Di 00127658
document
2012-07-06 2.3 Updated the contents related to description of Ji Di 00127658
MOCN Cell Demarcation feature in RAN14.0.
2012-07-28 2.4 Based on review opinion, achieve the final Ji Di 00127658
document
2013-05-17 3.0 Updated the contents related to the feature solution Lu Weifeng 61540
and service design in RAN15.0.
2014-7-21 3.1 Add MOCN Mobility Management renliang 00116371
Add MOCN Independent Iub Transmission
Resource Allocation
Contents
1.1 Purpose
This document provides guidelines for Huawei service delivery personnel in delivering
Universal Mobile Telecommunications System (UMTS) Radio Access Network (RAN)
sharing services. It describes the principles and designs of the RAN Sharing feature. This
document is for internal use only.
1.4 Dependencies
Applicable RAN Versions
This document is developed in compliance with RAN10.0 to RAN17.0.
Information Accuracy
Collect relevant information before RAN design and maintain full contact with operators
during services delivery to ensure information accuracy.
If the collected information is changed, locate the cause of the change and update the design
output promptly.
At the beginning of RAN Sharing delivery, delivery personnel must learn operators'
requirements through full communication and collect necessary information.
If the RAN Sharing feature is included in the network design service package, collect the
necessary information according to network design documents.
Operators have invested a lot of money in obtaining 3G licenses and building 3G networks
and now are struggling under a heavy financial burden. If they are willing to cooperate for a
3G network and share network infrastructures, they may be rewarded with less investment,
faster network construction, and shorter time to market. In addition, operators are merged and
restructured to survive or grow rapidly in the context of financial crisis and booming mobile
Internet. Under these circumstances, the UMTS network sharing solution witnesses a growth
in global application in recent years.
In addition to the market and technological background of the network sharing, this document
also describes the main network sharing applications. The UMTS RAN sharing solution is
classified into the RAN Sharing solution and the MOCN solution. In the RAN Sharing
solution, operators have their own spectrum resources; in the MOCN solution, operators share
spectrum resources. This document describes the principles, impacts on hardware and
software, and deployment scenarios of the two solutions. The combination of RAN sharing
and MOCN emerges as hybrid sharing, in which exclusive use and sharing of spectrum
resources coexist.
Hybrid Sharing is an application of the MOCN private cells.
In RAN Sharing, each operator uses independent carriers, RNCs and NodeBs are shared
physically, and a RAN is divided into multiple logical RANs. In this way, multiple operators
can use different frequencies to provide coverage for UEs in one area. RAN Sharing is
independent from the CN and UEs but only requires operators to upgrade RAN equipment.
RAN Sharing allows operators to deploy features independently in the shared RAN and to
implement cell-level independent operation and maintenance (OM). In this way, it helps
reduce the cost on network construction and separate services from network OM, ensuring
that RAN sharing operators use different operating strategies.
In MOCN scenarios, all operators use common carrier resources so that all RAN equipment,
such as the RNC and NodeB, are fully shared. MOCN requires cooperation from the CN and
UEs. UEs of release 6 and later versions (that is, supporting UEs) support the MOCN. UEs
earlier than release 6 do not support MOCN. However, Huawei MOCN supports UEs earlier
than release 6, assigns the UEs to correct CNs, and achieves the load balancing among
roaming UEs.
Operator A Operator B
CN CN
shared RNC
As shown in Figure 3-1, the dedicated NodeB indicates the private NodeB of operator A.
Like that on the traditional network, all UEs under the private NodeB access only to the
CN of operator A. All cells under the private NodeB broadcast only the Public Land
Mobile Network (PLMN) of operator A. UEs under the private NodeB can be handed
over to the cells of operator A in the shared NodeB. If dedicated NodeB need to be added
or changed frequency, firstly, logic cell should be deployed in RNC, then M2000
distribute license. Otherwise, equipment will alarm.
Provides flexible Iub transmission sharing ( Including IP or ATM transmission).
In Huawei RAN Sharing solution, Iub transmission resources can be provided for
operators in exclusive and shared manners. The following transmission resource
management modes are available:
Full sharing
In full sharing mode, admission control and flow control are shared, maximizing the
usage of transmission resources.
Full independence
In full independence mode, both admission control and flow control are independent.
Transmission resources of operators are for exclusive use of operators and do not
interfere with each other. Different logical ports are configured for operators. The
operators perform independent admission control and flow control and configure
different Virtual Local Area Networks (VLANs) based on the logical ports.
Partial independence
In partial independence mode, admission control is independent but flow control is
shared, meeting the independence requirements and improving resource usage.
Operators share a logical port and use transmission resources by configuring the
admission bandwidth ratio.
Iub transmission sharing is recommended to increase resources usage.
Implements differentiated quality of service (QoS) configuration.
Operators configure different QoS policies, such as Allocation Pretention Priority (ARP),
Traffic Class (TC), and Traffic Handling Priority (THP), and map the policies to different
scheduling priority indicator (SPI) weights and guaranteed bit rates (GBRs). In this way,
the differentiated QoS requirements of operators are met. Operators implement UE
admission control, flow control, and scheduling according to the relevant mapping.
Regardless of whether RAN Sharing is adopted, a CN node such as MSC server and SGSN belongs to
only one operator.
Figure 3-3 Handover between a RAN Sharing cell and a private cell (intra-operator)
The RAN Sharing cell on F2 is adjacent to the private cell of OpA. It is not allowed to hand
UEs of OpB over to this private cell. Therefore, Intra-frequency interference occurs on UEs of
both OpB and OpA. This problem may occur on a radio network where RAN Sharing cells
are networked with private cells. To solve the problem, the following methods are available:
1. Isolate the RAN Sharing cells from the private intra-frequency cells during network
planning.
2. Locate the border between a RAN Sharing cell and an intra-frequency private cell in an
area with low traffic and few UEs to reduce interference.
3. Add inter-frequency or inter-RAT neighboring cells for a RAN Sharing cell and enable
the coverage-based handover to hand over interfered UEs.
4. Enable the uplink QoS-based handover on the RNC to hand over the UEs with poor
uplink signal quality out of the border cell. However, the switch and threshold are of
RNC-level, which affect other cells.
Reselections and Handovers Between RAN Sharing Cells and 2G Roaming Cells
In 3G RAN Sharing, operators may adopt the 2G roaming sharing mode to share the 3G RAN
and 2G network at the same time. For example, UEs of an operator can roam to the 2G
network of another operator. In this scenario, the performance of 3G RAN Sharing may be
affected.
1. 2G-to-3G reselection and handover
Reselection
When a roaming UE of an operator (OpB, for example, as shown in) is reselected
from a 2G cell to a 3G cell, it takes more time than the reselection between 2G cells
or 3G cells.
For UEs of OpA, the 3G-to-2G reselection is the same as that in a common scenario.
To accelerate the reselection of UEs of OpB, the CN of OpB needs to change the
PLMN of OpA to the EPLMN of OpB in the location update message delivered to
UEs of OpB. In this way, when a UE of OpB performs an inter-RAT reselection, the
PLMN of OpA has the same priority as the PLMN of OpB, shortening the reselection
process. As a result, the CN of OpA rejects the UEs of OpB that try to access the 3G
network of OpA. However, as the rejection cause value is changed to that of less
severity (detailed in the preceding section), the process of reselecting UEs of OpB to
the 2G network of OpA is not affected.
Handover
For UEs of OpA, the 3G-to-2G handover is the same as that in a common scenario.
For UEs of OpB, configure 2G cells of OpA as the neighboring cells of 3G cells of
OpA and enable inter-RAT and inter-operator handovers. This type of handover
requires handover relationships between CNs of OpA and OpB.
Operator Identities
Both PLMN identity and CnOpIndex are used to identify an operator within different scopes.
An operator globally corresponds to a unique PLMN identity, which consists of mobile
country code (MCC) and mobile network code (MNC). A new parameter CnOpIndex is
introduced to identify a unique operator in a RAN.
Operator Types
There are two types of operators, primary operator and secondary operator, in RAN Sharing.
A primary operator takes precedence over a secondary operator in authority management and
license management. The primary operator can be anyone, for example, a third party in charge
of the shared RAN or an operator in possession of physical RAN equipment.
When adding an operator to the RNC, OperatorType must be used to check whether the
operator is a primary one or a secondary one. The primary operator must be specified in an
RNC, and the primary operator is only configured one. The number of operators is not more
than four in an RNC.
The RNC selects correct CN nodes for UEs that are under the same carrier but served by
different operators.
Shares network features among operators.
Operators share carriers. Therefore, all the network features are shared among operators.
Realizes independent performance statistics.
Operators share carriers. Therefore, Huawei MOCN does not allow operators to perform
independent FM(Fault Management) and CM(Configuration Management). However,
operators can perform PM(Performance Management) independently. For example,
operators can independently measure the radio access bearer (RAB) setup success rate,
the number of RAB releases with different causes, the circuit switched (CS) and packet
switched (PS) traffic volumes, and the consumption of resources, such as code, power,
and channel elements (CEs).
Achieves flexible configurations of shared and private NodeBs.
Huawei MOCN enables operators to deploy private NodeBs under a shared RNC, as
shown in Figure 3-6.
As shown in Figure 3-6, the dedicated NodeB indicates the private NodeB of operator A.
Like that in a common scenario, all UEs in the private NodeB access only to the CN of
OpA.
Cells under the private NodeB broadcast only PLMN IDs of the operator that the private
NodeB belongs to. UEs in the cells under the private NodeB can be handed over to
MOCN cells. The private cells of OpA are not available to the UEs of OpB and OpC.
The RNC filters the neighboring cell list according to the PLMN information of the
operator serving the current UEs. For details about mobility management, see 3.2.12
"MOCN Mobility Management."
The flexible configurations suit the MOCN in which partial frequencies are shared.
Supports full sharing of Iub transmission resources.
The MOCN allows operators to fully share the Iub transmission resources only.
Ensures the CS/PS coordination.
For the non-supporting roaming UEs (roaming UEs that do not support the MOCN),
their CS and PS services may register on the CNs of different operators. In this case,
CS/PS coordination must be ensured in the MOCN solution. For details, see 3.2.9
"CS/PS Coordination."
Pool Area
A RAN node service area contains all the cells controlled by the RAN node. A pool area
comprises one or more RAN node service areas. In a pool area, one or more CN nodes that
work in load sharing mode provide services in parallel.
UEs are served by one dedicated CN node in a pool area as long as the UEs are covered by
the pool area. Pool areas can overlap each other. If several overlapping pool areas cover the
same RAN node service area, the RAN node service area belongs to these pool areas. The CS
pool area and PS pool area are configured on a per-RAN node service area basis.
In a pool area, UEs can roam freely without changing the serving CN node. The RAN node
service area of a pool area can belong to one pool area or several pool areas. If several
overlapping pool areas cover the same RAN node service area, the RAN node service area
belongs to these pool areas.
Figure 3-7 shows an example of the pool area configuration.
As shown in Figure 3-7, the pool areas that do not support the Iu Flex, such as RAN node areas 4 and 8,
can coexist with the pool areas that support the Iu Flex.
The CS pool area and PS pool area can be configured as identical pool areas, such as CS pool
area 2 and PS pool area 2, or different pool areas, such as CS pool area 1 and PS pool area 1.
Each pool area is configured with unique CN node capability or number.
NRI
The Network Resource Identifier (NRI) uniquely identifies a CN node that serves a pool area.
The NRI ranges from 0 bit to 10 bits in length. In one pool area, all CN nodes use the same
NRI length. In areas where pool areas overlap, the NRI identifies the CN node that serves
these overlapping pool areas. For overlapping pool areas, all CN nodes serving specific pool
areas use the same NRI length.
The CN nodes in the CS domain and PS domain are addressed separately, so the NRIs of the
CS and PS domains are independent of each other.
One or more NRIs can be assigned to one CN node.
Parameters related to NRIs are as follows:
Length of PS NRI in bits
Length of CS NRI in bits
Network resource identifier
The NRI is part of the Temporary Mobile Station Identity (TMSI) for the CS domain, or
part of the Packet TMSI (P-TMSI) for the PS domain (from bit 14 to 23). TMSIs or
P-TMSIs are assigned to UEs by the serving CN node, with NRIs determined by using
certain bits. The NRI length can be configured on the RNC. If an NRI is less than 10 bits,
the NRI should count from bit 23. If an NRI is 0 bits, the NRI feature is not used. The
mapping between the NRIs and the CN nodes in the CN pool area is configured on RAN
nodes.
NullNRI
NullNRI is a special value of the NRI, which is configured on the RNC and is used to
identify an operator. Assume that the state of a CN node is OFFLOAD (which can be
configured on the RNC). When the NRI that the TMSI or P-TMSI of a UE carries is
NullNRI, the UE cannot access the CN node. The RNC connects the UE to a CN node in
the Normal state instead. If the NRI is valid and is not NullNRI, the UE can access the
CN node.
If an NRI is invalid (an NRI which is not configured on the RNC), the UE cannot access the CN node in
the OFFLOAD state.
NullNRI applies when Iu Flex is used, which can offload a CN node in the Iu Flex only
or Iu Flex + MOCN scenario. In MOCN only scenarios, NullNRI is not configured
because each operator has only one CN node.
IDNNS
In the Wideband Code Division Multiple Access (WCDMA) system, the UE provides an Intra
Domain NAS Node Selector (IDNNS) in the access stratum (AS) part of the Initial Direct
Transfer message to the RAN node. The IDNNS contains a 10-bit routing parameter, which
specifies the NRI value. In addition, the IDNNS contains the identity that indicates the source
of the routing parameters. The identity can be TMSI/P-TMSI, international mobile subscriber
identity (IMSI), international mobile equipment identity (IMEI), or others. The RAN node
masks the significant bits out of the routing parameter part of the IDNNS to determine the
NRI that is used to identify the relevant CN node. The most significant bit of the NRI
corresponds to the most significant bit of the routing parameter in the IDNNS.
If the IDNNS is obtained from IMSI, the value ranges from 0 to 999, with each value
corresponding to a CN node. RAN nodes can route messages to the correct CN nodes
according to the values.
NNSF
The NAS node selection function (NNSF) performs the selection of a CN node and the
processing of the IMSI paging message in both CS and PS domains. The MOCN introduces
the NNSF for selecting CN nodes regardless of whether Iu Flex is employed.
In a RAN node (namely the RNC), the NNSF selects a specific CN node (MSC server or
SGSN) and routes the initial NAS signaling message to the selected CN node.
If a CN node address configured for the NRI can be obtained from the initial NAS signaling
message, the NNSF routes the message to the CN node. If no CN node address is configured
for the obtained NRI or no NRI can be obtained (for example, the UE indicates an identity
corresponding to a UE contains no NRI), the NNSF selects an available CN node based on
load balancing and routes the message to the selected CN node.
No matter whether the IMSI or TMSI is contained in the paging request message that the CN sends to
page a UE, the UE responds with a paging response message containing the TMSI. The RNC determines
the CN node that the UE can access according to the NRI carried in the TMSI paging response message.
UEs regard that a location update fails and repeat the location update if the following
three types of PLMNs are inconsistent:
PLMNs of the RNC broadcast information decoded by the UE
PLMNs carried in the location area identity (LAI) and service area identifier (SAI) of
the Initial UE message (namely location registration request) that the RNC transmits
to the CN
PLMNs carried in the LAI and SAI of the location registration reception message that
the CN transmits to UEs
Independent CNs indicate independent signaling points and PLMNs.
}
}
}
If Iu Flex is not used, there is only one CN node for the operator in each domain.
Therefore, the initial NAS message is directly routed to this CN node according to the
CN domain.
If Iu Flex is used, the NNSF procedure is performed to select a specific CN node from
the CN nodes of the operator.
Regardless of whether MOCN is adopted, a CN node such as MSC server and SGSN belongs to only
one operator.
2.Reroute Command
6.Reroute Command
Supporting
MSC A MSC B RNC UE
1.System information
3.Network selection
6. If UE roaming is allowed, MSC B delivers the Location Update Accept message to the
UE. If UE roaming is not allowed, MSC B delivers the Location Update Reject message
to the UE.
Registration Phase
1. If the Gs interface exists between the CS domain and the PS domain of a CN, the
registration information in the CS domain is forwarded by the PS domain. Therefore, the
CN ensures CS/PS coordination.
2. If the Gs interface does not exist, the RNC ensures CS/PS coordination.
When the CN detects that the CS and PS domains register on different operators, the CN
carries "Redirection Indication" in the downlink Direct Transfer message with the cause value
"CS/PS coordination required". In addition, the CN sends the IMSI to the RNC and requests
for ensuring CS/PS coordination. Upon receiving the cause value "CS/PS coordination
required", the RNC checks the IMSI. If the IMSI belongs to the primary or secondary operator,
the RNC forwards a Direct Transfer message to the relevant CN. If the IMSI does not belong
to the primary or secondary operator, the RNC sends an Initial Direct Transfer message
carrying the Permanent NAS UE Identity IE (namely, IMSI) and NAS Sequence Number IE
again back to the CN to retry CS/PS coordination.
Existing application cases show that sites with the Gs interface are rare in the commercial network.
Generally, the RNC instead of the CN ensures CS/PS coordination.
The previous CS/PS coordination targets at non-supporting roaming UEs. For non-supporting
non-roaming UEs, CS and PS domains will not register on two operators. As specified in 3GPP
protocols, supporting roaming UEs ensure CS/PS coordination.
Routing roaming UEs in proportion is a new feature added in RAN13.0. The RNC routes
non-supporting roaming UEs by predefined proportion and ensure that the number of roaming
UEs routed to different operators conforms to the proportion. This feature ensures
MOCN-enabled operators to benefit from roaming UEs (roaming fees) by predefined
proportion in the shared network.
This feature does not conflict with the default operator. That is, if a default operator is
configured, the RNC still distributes roaming UEs among operators by proportion.
Handover Management
When UEs are connected, handover execution depends on the measurement control messages
delivered by the RNC and the UE measurement result.
1. Handovers between MOCN cells (intra-operator)
This type of handovers requires that the RNC determine the operator attributes of
neighboring cells before delivering measurement control messages. Only the information
about neighboring cells contains the operator to which the UE belongs can be delivered
to the UE through the measurement control message. In this case, the UE can be handed
over to the correct neighboring cell. Handovers between intra-operator MOCN cells are
classified into intra-frequency handovers and inter-frequency handovers. The handover
can be triggered by factors such as coverage, load, and QoS, as shown in Figure 3-10:
At the radio resource control (RRC) phase, the RNC cannot confirm the PLMN that the
UE accesses. In this case, the RNC performs the handover first before confirming the
PLMN to prevent call drop. Then, the RNC delivers the information about all
neighboring cells to the UE through the measurement control message. After the new cell
admits the UE, the RNC determines the PLMN for admitting the UE before sending the
Initial Direct Transfer message, which is the same as the process in which the original
cell directly admits the UE.
2. Handovers between MOCN cells and private cells (intra-operator)
Neighboring cells of MOCN cells are private cells of an operator. The RNC delivers
measurement control messages by operator attributes. Only the UEs that belong to this
operator (for example, OpA in Figure 3-11) can be handed over.
Figure 3-11 Handovers between MOCN cells and private cells (intra-operator)
The MOCN cell on F1 is adjacent to the private cell of OpA. UE handovers are not
allowed between OpB and OpC. Therefore, intra-frequency interference occurs on UEs
of both OpB and OpC. This problem often occurs on a radio network where MOCN cells
are mixed with private cells. To solve the problem, refer to the following methods:
Isolate the MOCN cells from the private intra-frequency cells during network
planning.
Locate the border between an MOCN cell and an intra-frequency private cell in an
area with low traffic and few UEs to reduce interference.
Add inter-frequency or inter-RAT neighboring cells for an MOCN cell and enable the
coverage-based handover to hand over interfered UEs.
Enable the uplink QoS-based handover on the RNC to hand over the UEs with poor
uplink signal quality out of the border cell. However, the switch and threshold are of
RNC-level, which affect other cells.
3. Handovers between MOCN cells and 2G cells (intra-operator)
Neighboring 2G cells of 3G MOCN cells belong to one operator. The RNC delivers
measurement control messages by operator attributes. Only the UEs that belong to this
operator can be handed over. If the 2G-to-3G handover function is enabled, the UEs can
also be handed over to 3G MOCN cells (for example, OpB in Figure 3-12).
4. Inter-operator handovers
In 3G MOCN, the inter-operator handover must be enabled in special application
scenarios. Intra-RAT inter-operator handovers and inter-RAT inter-operator handovers
can be controlled on the RNC. If they are enabled, the RNC does not filter the
neighboring cells of an operator when delivering a measurement control message to UEs
of the operator but deliver the message to all neighboring cells. In this case, the UE may
be handed over to a cell of another operator (for example, OpB in Figure 3-13).
Currently, sites support inter-RAT inter-operator handovers but do not support intra-RAT
inter-operator handovers. Network sharing is implemented at such sites by using 3G
MOCN + 2G roaming. For example, in the TMV shared network of country T, UE 00
needs to be handed over from the MOCN cell to the GSM R99 cell. For details, see
section 5.2.3 "Coexistence of the MOCN RNC and Dedicated RNC."
5. Inter-RNC handovers
Same as the inter-RNC handovers in common scenarios, the inter-RNC handovers in
MOCN scenarios also require the neighboring combination switch.
If the neighboring combination switch for serving radio network controller (SRNC) is
turned off:
When the optimum cell is in the SRNC, only the neighboring cells that the SRNC
configures in the optimum cell will be added to the neighboring cell set of the UE.
The neighboring cell set includes intra-frequency, inter-frequency, and inter-RAT
neighboring cells.
When the optimum cell is in the drift RNC (DRNC), only the neighboring cells that
the DRNC configures in the optimum cell will be added to the neighboring cell set of
the UE. The neighboring cell set includes intra-frequency, inter-frequency, and
inter-RAT neighboring cells.
If the neighboring combination switch for SRNC is turned on:
No matter whether the optimum cell is in the SRNC or DRNC, the target neighboring
cells configured in the SRNC and DRNC will be added to the neighboring cell set of
the UE.
Currently, the neighboring cell switch is turned on by default.
The operators to which target neighboring cells belong must also be considered in
inter-RNC handovers in MOCN scenarios. The Multiple PLMN List IE is added to
the Neighboring FDD Cell Information IE of the inter-RNC Radio Link Setup
Response message specified in 3GPP R7 or later. The Multiple PLMN List IE
informs the SRNC of the operator to which neighboring cells belong.
If inter-operator handover is disabled, the information about neighboring cells will be
delivered to the UE through the measurement control message only when the
information contains the operator to which the UE belongs. The PLMN information
delivered by the DRNC prevails even if neighboring cells have already been
configured in the SRNC.
Figure 3-14 shows the intra-operator inter-RNC handover.
Operator
information about
neighboring cells
6. Relocation
Relocation in MOCN scenarios is the same as relocation in common scenarios.
Relocation includes static relocation, hard handover-triggered relocation, and
cell/UTRAN registration area (URA) update-triggered relocation. The difference lies in
that operators to be selected must be specified for relocation in MOCN scenarios.
After determining the CN operator of the destination network based on the available
access information, current PLMN, or similar information in a node of the shared
network during relocation, the SRNC informs the destination network of the selected CN
operator. The SRNC fills in the PLMN ID of the selected operator in Target RNC ID of
the Relocation Required message. In this way, the CN does not need to select the PLMN.
In static relocation, the target RNC (TRNC) notifies the UE of the PLMN of the new cell
by sending a Utran Mobility Information message, in which the PLMN identity is set to
the common PLMN and the primary PLMN identity to the PLMN of the selected
operator. Non-supporting UEs obtain the new PLMN ID based on the PLMN identity,
whereas supporting UEs obtain the new PLMN ID based on the primary PLMN identity.
In hard handover-triggered relocation, the SRNC notifies the UE of the PLMN identity
and primary PLMN identity by sending a Radio Bear Reconfiguration message, in which
the PLMN identity is set to the common PLMN and the primary PLMN identity to the
PLMN of the selected operator.
In cell/URA update-triggered relocation, the TRNC notifies the UE of the PLMN
identity and primary PLMN identity by sending a Cell Update Confirm message, in
which the PLMN identity is set to the common PLMN and the primary PLMN identity to
the PLMN of the selected operator.
If the SRNC cannot confirm the operator to which the UE will be relocated, the SRNC
sets the PLMN in Target RNC ID of the Relocation Required message to the common
PLMN. The source CN selects a CN, and the target CN sends the selected PLMN in the
Relocation Request message to the TRNC.
7. Cell/URA update
The cell/URA update in MOCN scenarios is the same as that in common scenarios.
After receiving a location update request, the RNC checks the PLMN ID list contained in
the target cell. If the list contains the PLMN that the UE connects to, the RNC allows the
cell/URA update. Otherwise, the RNC determines whether to enable inter-PLMN
handovers. If the switch is turned on, the RNC allows the cell/URA update; if not, the
RNC does not allow the cell/URA update.
Cell Reselection
Cell reselection is a proactive behavior of UEs. It is free from the control by the network but
affected by network parameters.
Cell reselection in MOCN scenarios is more complicated than that in common scenarios,
because non-supporting UEs and supporting UEs have different understandings on the PLMN
and cell reselection is affected by the mixing of MOCN cells, private cells, and 2G cells. The
following details the difference between cell reselection in MOCN scenarios and cell
reselection in common scenarios.
1. Reselections between MOCN cells
If adjacent MOCN cells are served by the same operator, non-supporting UEs and
supporting UEs of all operators can perform cell reselection like that in a common
scenario.
If adjacent MOCN cells are served by different operators, cell reselection is more
complicated. As shown in Figure 3-16, when cell 3 is not served by OpC, UEs of OpC
will not be reselected to cell 3.
However, cell reselections are different for non-supporting UEs and supporting UEs,
which is described as follows:
Non-supporting UEs
Cell 1 and cell 2 are configured with the same common PLMN, and cell 2 and cell 3
are configured with different common PLMNs. In this way, non-supporting UEs of
OpC do not regard cell 3 to be served by the same operator and therefore will not be
reselected to the cell. Cell 2 and cell 3 can be configured with the same LAC. In this
way, UEs of OpA and OpB can move randomly, without affecting the non-supporting
UEs of OpC. OpA CN and OpB CN need to configure the common PLMNs of cell 2
and cell 3 as mutual EPLMNs and deliver the EPLMNs to UEs of OpA and OpB
through the location update message.
In addition, common PLMNs of cell 2 and cell 3 correspond to logos of OpA and
OpB.
Supporting UEs
A supporting UE of OpC does not read its own PLMN from cell 3. Therefore, the
supporting UE will not be reselected to cell 3.
2. Reselections between MOCN cells and private cells
If an MOCN cell is adjacent to a private cell, UEs of non-private cells cannot be
reselected to the private cell. As shown in Figure 3-17, both non-supporting UEs and
supporting UEs of OpB and OpC cannot find the HPLMNs from cell 2 and cell 3.
Therefore, the UEs will not be reselected to the cells.
Cell reselections are different for non-supporting UEs and supporting UEs of OpA,
which is described as follows:
Non-supporting UEs
A non-supporting UE of OpA can only read the common PLMN from an MOCN cell
(for example, cell 1) and read the HPLMN of OpA from a private cell (for example,
cell 2 or cell 3). Therefore, during the reselection from an MOCN cell to a private cell,
the UE first camps on an MOCN cell (RPLMN is preferred). When no MOCN cell is
available, the UE will be reselected to its own private cell using periodic HPLMN
reselection.
The HPLMN reselection period is the multiples of six minutes, ranging from six minutes to eight hours.
The period is written into the SIM card. If the period is not written, it is 60 minutes by default.
During the reselection from a private cell to an MOCN cell, the UE first camps on the
private cell, not the MOCN cell. To speed up the reselection, the OpA CN needs to
set the common PLMN to the EPLMN of the HPLMN in the private cell. In this way,
the UE regards the MOCN cell as a cell in the EPLMN.
The CN delivers the EPLMN to the UE in the location update message. The EPLMN is configured
according to the LAC and IMSI. The EPLMN is updated each time the UE performs a location update.
CS and PS domain of the core network needs configure the same EPLMN, otherwise the last location
area or routing area updated EPLMN list overwrites the previous one.
Supporting UEs
A supporting UE of OpA reads the HPLMN of the operator from both an MOCN cell
and a private cell. Therefore, the reselection is the same as a common reselection,
which does not require special configurations and is not affected by the
configurations of the previous non-supporting UEs.
3. Reselections between MOCN cells and 2G cells
If an MOCN cell is adjacent to a 2G cell, UEs of non-2G cells cannot be reselected to the
2G cell. As shown in Figure 3-18, both non-supporting UEs and supporting UEs of OpB
and OpC cannot find HPLMNs from cell 4. Therefore, the UEs will not be reselected to
the cell.
Like a private cell, cell reselections are different for non-supporting UEs and supporting
UEs of OpA, which are described as follows:
Non-supporting UEs
A non-supporting UE of OpA can only read the common PLMN from an MOCN cell
(for example, cell 1) and read the HPLMN of OpA from a 2G cell (for example, cell
4). Therefore, during the reselection from an MOCN cell to a 2G cell, the UE camps
on its own 2G cell using the periodic HPLMN reselection.
During the reselection from a 2G cell to an MOCN cell, the UE first camps on the
private cell, not the MOCN cell because the RPLMN is preferred. To speed up the
reselection, the OpA CN needs to set the common PLMN to the EPLMN of the
HPLMN in the 2G cell. In this way, the UE regards the MOCN cell as a cell in the
EPLMN.
The 2G network provides seamless coverage. Therefore, this scenario is common.
However, the previous configurations may result in ping-pong reselections of a
non-supporting UE between a 2G cell and a 3G cell. That is, the UE will be
reselected from the 3G cell to the 2G cell based on the HPLMN reselection period
and then reselected back from the 2G cell to the 3G cell based on the EPLMN and 3G
network. Use the following methods to solve the problem:
Set the common PLMN as the EHPLMN of the HPLMN and write the common
PLMN into the SIM card (EFEHPLMN) of OpA. As a result, UEs of OpA will be
reselected to the common PLMN (namely the 3G MOCN cell) during each
HPLMN reselection.
Prolong the HPLMN reselection period in the SIM card to avoid ping-pong
reselection.
However, these operations require that relevant information be written in the SIM
card before sales, which is difficult to implement on commercial networks. For
details, contact operators.
Supporting UEs
A supporting UE of OpA reads the HPLMN of the operator from both an MOCN cell
and a 2G cell. Therefore, the reselection is the same as a common inter-RAT
reselection, which does not require special configurations and is not affected by the
configurations of the previous non-supporting UEs.
The reselection between an MOCN cell and a OpA 2G cell is the same as the
reselection between an MOCN cell and a OpB/OpC 2G cell.
4. Reselections between MOCN cells and 2G roaming cells
When roaming is enabled on 2G cells, reselections between MOCN cells and 2G
roaming cells are more complicated than reselections between MOCN cells and 2G cells.
For example, in the offices of country T and country PH, operators use 2G roaming + 3G
MOCN to implement network sharing. Figure 3-19 shows that the OpA GSM allows
UEs of OpB and OpC to roam in certain areas, such as LAC 5 and cell 5. OpA CN
configures roaming in different areas and determines whether to allow roaming UEs of
other operators according to different LACs.
(Roaming allowed)
Reselection between an MOCN cell and a 2G roaming cell does not affect the operator
serving roaming UEs, such as OpA, but affects the UEs of roaming operators, such as
OpB and OpC. Cell reselections are different for non-supporting UEs and supporting
UEs, which is described as follows:
Non-Supporting UEs
Reselection from an MOCN cell to a 2G roaming cell
Non-supporting UEs of OpB and OpC can only read the common PLMN from an
MOCN cell (for example, cell 1). A 2G cell (for example, cell 5) of OpA
broadcasts the PLMN of OpA, which will not be selected by UEs of OpB and
OpC. Therefore, OpB CN and OpC CN need to configure the PLMN of OpA as
the EPLMN of the common PLMN in the MOCN cell and deliver the EPLMN to
UEs of OpB and OpC through the location update message. In this way, UEs of
OpB and OpC regard the 2G cell of OpA as a cell in the EPLMN and are
reselected to the 2G cell where roaming is permitted. The reselection is the same
as a common inter-RAT reselection.
However, a private 3G cell (cell 3) or a 2G cell (cell 4) of OpA where roaming is
forbidden exists near the MOCN cell (cell 1). The PLMNs broadcast by cell 3 and
cell 4 are the same as that broadcast by cell 5. According to the preceding
configurations, UEs of OpB and OpC try to connect to cell 3 and cell 5 but are
rejected by OpA CN. To ensure that UEs roam in 2G cells of OpA, the rejection
cause value is of less severity, for example, "No Suitable Cell in LA." In this way,
UEs do not attempt to be reselected to the cell of the LAC after the rejection but
attempt to be reselected to other LACs of OpA.
Reselection from a 2G roaming cell to an MOCN cell
During the reselection from a 2G roaming cell of OpA to an MOCN cell,
non-supporting UEs of OpB and OpC first camp on the private cell, not the
MOCN cell because the RPLMN is preferred. To speed up the reselection, the
OpA CN needs to configure the common PLMN as the EPLMN of the HPLMN in
the 2G roaming cell and send the EPLMN to UEs of OpB and OpC (also UEs of
OpA in the preceding section) through the location update message. In this way,
the UE regards the MOCN cell as a cell in the EPLMN.
The NRI obtained by the UE belongs to OpA because the UE is reselected from
the cell of OpA. In this way, after the UE is reselected to the MOCN cell and
initiates the location update (registration), the RNC first routes the UE to OpA CN.
After being rejected, the UE re-attempts to be reselected to OpA CN. Several
attempts later, the UE may be reselected back to OpB CN or OpC CN, prolonging
the network connection time of the UE. Therefore, improvements are made in
RAN12.0, that is, the Redirect Attempt flag can be carried in inter-RAT
reselection. When the CN that the UE originally camps on rejects the UE, the UE
attempts to be reselected to its own CN, speeding up the reselection.
Supporting UEs
Reselection from an MOCN cell to a 2G roaming cell
Supporting UEs of OpB and OpC can read their own PLMNs (namely, HPLMNs)
from an MOCN cell (cell 1). A 2G cell of OpA (cell 5) broadcasts the PLMN of
OpA, which will not be selected by UEs of OpB and OpC. Therefore, OpB CN
and OpC CN need to configure the PLMN of OpA as the EPLMN of their own
HPLMN in the MOCN cell and deliver the EPLMN to UEs of OpB and OpC
through the location update message. In this way, UEs of OpB and OpC regard the
2G cell of OpA as a cell in the EPLMN and are reselected to the 2G cell that OpA
allows for roaming. The reselection is the same as a common inter-RAT
reselection.
Similarly, UEs may be reselected to a private 3G cell of OpA (cell 3) or a 2G cell
that does not allow roaming (cell 4). The rejection cause value is of less severity,
for example, "No Suitable Cell in LA."
Reselection from a 2G roaming cell to an MOCN cell
Supporting UEs of OpB and OpC can be reselected from the 2G roaming cell of
OpA to the MOCN cell by the following methods:
Method 1: OpA CN configures the PLMNs of OpB and OpC as the EPLMN of
the OpA PLMN in the 2G roaming cell and delivers the EPLMN to UEs of OpB
and OpC but not UEs of OpA through the location update message. In this way,
UEs of OpB and OpC regard the 2G roaming cell of OpA as a cell in the EPLMN.
However, supporting UEs of OpB and OpC can also read the PLMN of OpA in
the MOCN cell. Some UEs may attempt to be reselected to the PLMN of OpA
because OpA belongs to the RPLMN and are rejected. Then, the UEs attempt to
be reselected to the HPLMN, prolonging the UE reselection time.
The rejection cause value shall be of greater severity. In this way, the UE can be triggered to initiate the
PLMN reselection, for example, "Roaming not allowed in this location area." For details, see 3GPP TS
23.122.
Method 2: Supporting UEs of OpB and OpC can be reselected to the MOCN cell
using periodic HPLMN reselection.
After dividing CE resources into groups and identifying a UE's serving operator, the
NodeB allocates the CE resources in this operator's private groups to the UE. If the CE
resources in the operator's private groups are used up, the CE resources in the common group
can be used. The CE resources in the common group are used on a first come, first served
basis.
The RNC relieves CE congestion for each operator separately, this feature introduces
operator-level congestion relief for uplink CE resources, including: BE service rate
reduction and Switching from 2 ms TTI to 10 ms TTI.
Benefits
This feature prevents a single operator from occupying excess CE resources in MOCN
scenarios and ensures that each operator retains independent CE resources.
This feature slightly decreases CE resource utilization. If CE resources are not properly
allocated (for example, if insufficient CE resources are allocated to an operator), this feature
brings some negative effects. The negative effects include access performance deterioration
(for example, a decrease in the RRC setup success rate and RAB setup success rate) and
HSUPA throughput decrease.
System Capacity
This feature has no impact on system capacity when the CE resources allocated to each
operator are sufficient.
If the resources are not properly allocated among operators, system capacity decreases. For
example, if a total of 100 CEs are equally allocated to operators A and B who require 60 and
40 CEs, respectively, only 90 CEs are utilized, with a 10% decrease in system capacity.
Network Performance
Before you deploy this feature, allocate CE resources to operators based on their traffic
requirements.
For example, allocate the CE resources equally if all operators have almost the same traffic
requirements.
This feature has no impact on network performance when the CE resources allocated to each
operator are sufficient.
If the CE resources are not properly allocated among operators, network performance
deteriorates. For example, if a total of 100 CEs are equally allocated to operators A and B who
require 60 and 40 CEs, respectively, network performance deteriorates, with a 10% decrease
in resource utilization, cell HSUPA throughput, R99 PS throughput, and access success rate.
Related Features
To allocate user-plane bandwidth independently, the RNC is configured with multiple logical
ports, where each logical port corresponds to one operator. After identifying the UE's serving
operator, the RNC sends the UE's user-plane data to the corresponding logical port.
Figure 3-21 Sharing one physical link over the Iub interface
Each operator uses a dedicated physical link over the Iub interface, as shown in
Figure 3-22.
Figure 3-22 Using dedicated physical links over the Iub interface
This feature applies only to RNC downlink transmission resources, not to NodeB uplink
transmission resources.
Benefits
This feature allocates Iub transmission resources to operators in MOCN scenarios, which
prevents one operator from occupying excess Iub transmission resources and ensures that
each operator retains independent Iub transmission resources.
This feature slightly decreases Iub resource utilization. If Iub resources are not properly
allocated (for example, if insufficient Iub bandwidth is allocated to an operator), this feature
brings some negative effects. The negative effects include access performance deterioration
(for example, a decrease in the RRC setup success rate and RAB setup success rate) and
HSPA throughput decrease.
System Capacity
This feature has no impact on system capacity when the Iub transmission resources allocated
to each operator are sufficient.
If the resources are not properly allocated among operators, system capacity decreases. For
example, if a total of 10 Mbit/s bandwidth is equally allocated to operators A and B who
require 6 Mbit/s and 4 Mbit/s, respectively, only 9 Mbit/s bandwidth is utilized, with a 10%
decrease in system capacity.
Network Performance
Before you deploy this feature, allocate Iub transmission resources to operators based on their
traffic requirements. For example, allocate the Iub transmission resources equally if all
operators have almost the same traffic requirements.
This feature has no impact on network performance when the Iub transmission resources
allocated to each operator are sufficient.
If the Iub transmission resources are not properly allocated among operators, network
performance deteriorates. For example, if a total of 10 Mbit/s bandwidth is equally allocated
to operators A and B who require 6 Mbit/s and 4 Mbit/s, respectively, resource utilization and
access success rate decrease by 10%. The throughput of a single HSDPA user decreases as
follows:
100% x (1 Minimum bandwidth allocated to an operator/Min(Bandwidth obtained before
service activation, Maximum bandwidth required by an UE)).
Related Features
Figure 4-1 Coexistence of shared and non-shared RNCs under full coverage
Operator A CN Operator B CN
RNC A
Iur
If an operator shares a carrier in some areas with another operator but does not share the
carrier in other areas, co-channel interference may occur. Figure 4-2 shows that operator B
shares F2 with operator A in some areas. Co-channel interference occurs in the border areas of
operator A and operator B (as shown in the dashed circle in Figure 4-2).
Deploy the borders of intra-frequency inter-operator cells in an area with low traffic and
fewer UEs to reduce interference.
Add inter-frequency or inter-RAT neighboring cells to intra-frequency inter-operator
cells and enable the coverage-based handover to hand over interfered UEs.
Enable the uplink QoS-based handover on the RNC to hand over the UEs with poor
uplink signal quality on intra-frequency inter-operator cells. However, the switch and
threshold are of RNC-level, which affect other cells.
Operator A CN Operator B CN
Shared
BSS A BSS B
RNC1
There is a special scenario, that is, RAN Sharing is enabled on the shared 3G network, but
roaming is enabled on the 2G network for sharing. As shown in Figure 4-4, UEs of operator A
can roam to GSM cells of operator B. Therefore, in addition to the neighboring relationships
between the GSM cell of operator B and the MOCN cell, the neighboring relationships
between the GSM cell of operator A and the MOCN cell and the neighboring relationships
between GSM cells of operator A and operator B must be configured.
Handovers from a shared 3G cell of operator A to a GSM cell of operator B: The
inter-RAT inter-PLMN handover switch must be turned on.
Reselection from a shared 3G cell to a GSM cell of operator B: Configure the PLMN of
operator B to the EPLMN of operator A in the CN of operator A and deliver the EPLMN
to the shared 3G cell of operator A adjacent to the GSM cell of operator B. In this way,
UEs of operator A can be reselected to the GSM cell of operator as the EPLMN,
speeding up the reselection. In addition, the shared 3G cell of operator A does not
configure the shared 3G cell of operator B as the neighboring cell. Therefore, UEs of
operator A do not attempt to be reselected to the shared 3G cell of operator B.
Reselection from a GSM cell to a shared 3G cell of operator A: The reselection relies on
the periodic HPLMN reselection of operator A, prolonging the reselection.
Figure 4-5 Coexistence of full and partial coverage for different operators
Operator A CN Operator B CN
Iur Shared
RNC A
RNC1
Operator A
MSC server 3
MSC server 2
MSC server 1
MGW 4 SGSN 2
MGW 3 SGSN 1
MGW 2
MGW 1
Operator B
SGSN 1
Shared SGSN 2
RNC SGSN 3
Operator C Cell
4.3.1 Iu Interface
An operator can only use the operator's own CN nodes. Therefore, each operator can use one
or more physical ports to connect its own CN nodes. The physical bandwidth related to the
physical ports is operator dedicated.
In the shared RNC, a global CN identity is added to each CN node. The global CN identity
consists of the PLMN ID and the CN ID.
Figure 4-8 shows an example where the RNC is shared by operator A and operator B.
Operator A uses Iu Flex, whereas operator B does not.
Figure 4-8 Operator A and operator B share an RNC with operator A using Iu Flex
The switch to enable or disable Iu Flex is configurable for each operator, that is, the operator
can individually enable or disable Iu Flex. The configuration of Iu Flex in the shared RNC is
similar to that in the non-shared RNC. In the shared RNC, the configuration for the initial
NAS message routing should be performed separately by each operator. The configuration is
as follows:
NRI length for the CS or PS domain
Mapping between NRI and global CN identity
Mapping between IMSI route value and global CN identity
IMSI route value = (IMSI div 10) mod 1000. The value range is 0-999.
IPoA/IP
Logical RNC 2 CBC 2
Operator A
Operator B
Figure 4-11 Separation of transmission resources for the Iub user plane
When Iub user plane is dedicated, the transmission resources for the user planes of operators
are independent physical links, as shown in Figure 4-13.
When ATM/IP dual stack is adopted, there are two cases, as shown in Table 4-1.
Case Method
Two operators engage dual stacks respectively.
Remarks:
1 Generally only consider the downlink (RNC-> Node B) the user plane Dedicated; does
not support user plane dedicated uplink (NodeB-> RNC); main reasons: the uplink
transmission more congestion than the downlink transmission;
2 Downlink user plane dedicated is to configure the logical port;
Figure 4-15 Iur control plane of two RNCs shared by operators A and B
Figure 4-16 Iur user plane of two RNCs shared by operators A and B
Physical RNC 1 Physical RNC 2
Signaling node 1 Signaling node 2
Transport paths
Logical RNC 1a Logical RNC 2a
Operator A
Operator B
Common
Traffic of Operator A
Traffic of Operator B
4.4 OM Design
The OM mode of a shared network is closely related to that negotiated between operators.
OM modes are classified into joint OM mode and rent OM mode, which are described as
follows:
In joint OM mode, operators A and B provide respective devices to establish a joint
company and the joint company operates and maintains the network.
In rent OM mode, operator A shares some devices with operator B and operator A
operates and maintains the network.
In the two OM modes, a network is managed by one element management system (EMS) or
M2000, and then the message and private data are transmitted to the respective system of each
operator through the northbound interface. Currently, it is not allowed for each operator to use
an M2000 to manage an MOCN network.
In joint OM mode, the joint company provides information about each operator to the
upper-level network management system (NMS) of operator A and operator B. The joint
company can be any one of operator A and operator B. In rent OM mode, primary operator A
provides information about operator B to the upper-level NMS of operator B.
In RAN Sharing, M2000 northbound interfaces can distinguish cell-level data. If cell A
belongs to operator A, operator A can receive the private data of cell A and all public data over
the northbound interface(NBI).
Private data
Private data indicates the cell-level FM, PM, and CM data, which are related only to
dedicated operators. Private data are delivered to the NMS of the operator over the
M2000 northbound interface.
Shared data
Shared data indicates all the FM, PM, and CM data not the cell-level data exported over
the northbound interface.
The traffic statistics on RNC and NodeB can be performed independently for operators. The
measurement object is Operator. The traffic statistics can be subscribed and queried on M2000
PM window. Currently, the Operator objects of the northbound interface are reported without
distinguishing the operators.
During the northbound interconnection of different FMs, CMs, and PMs, confirm the
interface format supported by the existing system of the operator. Currently, the M2000
supports the following interfaces.
FM interface: includes Corba alarm interface and American Standard Code for
Information Interchange (ASCII) streaming interface of the M2000.
PM interface: indicates the M2000 performance file interface.
CM interface: indicates the CME file import and export interface.
Private group
Each operator who shares the NodeB has a corresponding private group in the NodeB
license. The private group defines the operator's dedicated resources and functions.
For example, private group A defines resources and functions that can be used by
operator A only.
Common group
The common group is optional in the NodeB license. The group defines common
resources and functions that can be used by all operators.
Resource control items
Resource control items include CEs, frequencies, and Power Amplifier (PA) output
power. The total number of CEs in a NodeB license cannot exceed the supported
capacity of the NodeB hardware. Resource control items could be deployed respectively
by operators on M2000
CE usage principle
The primary operator purchases the CE license from Huawei and obtains the total
number of available CEs in a NodeB. Then, the primary operator negotiates with the
secondary operator to determine the number of CEs that each operator can consume.
That is, determine private groups corresponding to the fixed frequencies (namely, cells).
In addition, some CE resources can be used in public groups and shared by operators.
When a Radio Link (RL) is to be set up, it consumes the CEs in the private group first. If
the required CEs exceed 110% of the CEs in the private group, the RL begins to
consume the CEs in the common group. If the required CEs still exceed 110% of the CEs
in the common group, the RL setup fails.
For example, assume that one NodeB is shared by operators A and B. Operator A has
cells 0, 1, and 2. Operator B has cells 3, 4, and 5. There are three groups in the NodeB
license: private group A, private group B, and common group. Each of the three groups
has 8 UL CEs and 8 DL CEs. If there is a 384 kbit/s call request in cell 0, the call
requires 10 UL CEs. The call uses 9 UL CEs of private group A and 1 UL CE of the
common group. If, at the same time, there is another 384 kbit/s call request in cell 3 and
the call also requires 10 UL CEs, the call uses 9 UL CEs of private group B and 1 UL CE
of the common group. In total, 2 UL CEs of the common group are used.
NodeB can make statistics on the number of CEs consumed by each operator, and RNC
can implement admission control according to the number of CEs configured for each
operator.
Function control items
NodeB function control items are oriented to NodeB-level features, which are configured
on the M2000. The function control items are shared by operators and cannot be
configured independently for each operator.
items in the following sequence: the primary operator, and then the secondary operators. No
sequence between secondary operators is required.
Resource control items
On the RNC, resource control items can be configured for different operators as required.
Resource control items in the RNC license include the maximum traffic volume of CS,
the maximum throughput of PS, HSDPA, HSUPA, and MBMS, and the maximum
number of activated UEs of the enhanced FastDormancy. In addition, hardware capacity
is included for the primary operator. The maximum number of resources used by each
operator cannot exceed the threshold configured in the License.
Function control items
The RNC license provides more than 60 function control items, which can be classified
into two types: items totally controlled by the primary operator and items controlled by
each operator.
Items totally controlled by the primary operator allow the function to be enabled or
disabled by the RNC not an individual operator. The function is not activated when any
of the following conditions is true:
The function control item is not activated in the RNC license.
The function control item is not selected when the primary operator runs the ACT
license.
Items controlled by each operator allow an operator to determine whether to activate the
function. The function is not activated when any of the following conditions is true:
The function control item is not activated in the RNC license.
The function control item is not selected when the operator sets the RNC license.
Items totally controlled by the primary operator (based on RAN13.0) are as follows:
RANSHARE(RAN sharing).
RAN_SHARING_ENHANCED_PACKAGE(RAN Sharing Enhanced Package).
ATMIPDUALSTACK(IUB ATM/IP Dual Stack Transportation Function).
IPTRANINIUR(IP Transportation in Iur Interface).
FRACTIONALIP(Fractional IP).
FRACTIONALATM(Fractional ATM).
IPIUBOVERBOOKING(IP IUB overbooking Function).
ATMOVERBOOKING(ATM IUB overbooking Function).
IUBHYBRIDIPTRAN(IUB Hybrid IP Transportation Function).
ACRWHENSPUOVER(Access Class Restriction when service processing unit (SPU)
overload).
License control for urgency
Hardware capacity license (165 Mbit/s) (per board)
Hardware capacity license (300 Mbit/s) (per board)
Other function control items in the RNC license can be controlled by each operator.
obtain the total number of hardware resources required by the network, and then design the
network capacity based on the calculation.
4.7.1 BTS3812E/BTS3812AE
The following table lists the carriers that can be used by the two operators independently.
4.7.2 DBS3800
The following table lists the carriers that can be used by the two operators independently.
4.7.3 BTS3900/BTS3900A
The following table lists the carriers that can be used by the two operators independently.
4.7.4 DBS3900
The following table lists the carriers that can be used by the two operators independently.
Only mobility management between MOCN cells is involved in this scenario. For details, see
the related description of MOCN mobility management in section 3.2.12 "MOCN Mobility
Management."
For typical commercial networks, see the networks deployed by operator T and operator B in
country C in section 5.2.1 "Full MOCN."
In this scenario, mobility management between MOCN cells and mobility management
between MOCN cells and private cells are involved. For details, see the related description of
MOCN mobility management in section 3.2.12 "MOCN Mobility Management."
The methods or principles for resource management, operation and maintenance, and function
control comply with the regulations of MOCN, not RAN Sharing. For typical commercial
networks, see TMV network sharing of country T in section 5.2.3 "Coexistence of the MOCN
RNC and Dedicated RNC", network sharing between operator P and operator O of country P
in section 5.2.4 "Coexistence of the MOCN RNC and 2G Network", and network sharing
between operator S and operator D of country PH in section 5.2.5 "Coexistence of Iu Flex and
MOCN."
Figure 5-3 Coexistence of the MOCN RNC and dedicated RNC under full coverage
Operator A CN Operator B CN
Iu
Operator A
Iur
Dedicated RNC MOCN Shared
RNC
Operator B Iur
Dedicated RNC
Operator B Cell
MOCN Cell
Operator A Cell
The MOCN RNC connects to dedicated RNC A and RNC B over the Iur interface, and
neighboring cell relationships are configured. When a UE of operator A in the same MOCN
cell transfers from a shared area to a non-shared area, the RNC enables the UE of operator A
to be handed over to a neighboring cell of operator A. The same situation applies to operator
B. For details about mobility management in this scenario, see the related description of
MOCN mobility management in section 3.2.12 "MOCN Mobility Management."
For typical commercial networks, see network sharing between operator S and operator D of
country PH in section 5.2.5 "Coexistence of Iu Flex and MOCN."
mobility management in this scenario, see the related description of MOCN mobility
management in section 3.2.12 "MOCN Mobility Management."
Operator A CN Operator B CN
There is a special scenario, in which MOCN is enabled on the shared 3G network but roaming
is enabled on the 2G network for sharing. As shown in Figure 5-5, UEs of operator A can
roam to GSM cells of operator B. Therefore, in addition to the neighboring relationships
between the GSM cell of operator B and the MOCN cell, the neighboring relationships
between the GSM cell of operator A and the MOCN cell and the neighboring relationships
between GSM cells of operator A and operator B must be configured. For details about
mobility management in this scenario, see the related description of MOCN mobility
management in section 3.2.12 "MOCN Mobility Management."
For typical commercial networks, see TMV network sharing of country T in section 5.2.3
"Coexistence of the MOCN RNC and Dedicated RNC" and network sharing between operator
S and operator D in country PH in section 5.2.5 "Coexistence of Iu Flex and MOCN."
Operator A
MSC server 3
MSC server 2
MSC server 1
MGW 4 SGSN 2
MGW 3 SGSN 1
MGW 2
MGW 1
Operator B
MOCN SGSN 1
SGSN 2
Shared
RNC SGSN 3
5.3.1 Iu Interface
On the Iu interface, an operator can only use its own CN nodes. Therefore, each operator can
use one or more physical ports to connect its own CN nodes. The physical bandwidth related
to the physical ports is operator dedicated.
In the MOCN RNC, a global CN identity is added to each CN node. The global CN identity
consists of the PLMN ID and the CN ID.
Figure 5-8 shows an example where the RNC is shared by operator A and operator B.
Operator A uses Iu Flex, whereas operator B does not.
Figure 5-8 Operator A and operator B share an RNC with operator A using Iu Flex
SGSN 1
SGSN 2
Iu
Operator A
MOCN Shared RNC
Operator B
The switch to enable or disable Iu Flex is configurable for each operator, that is, the operator
can individually enable or disable Iu Flex. The configuration of Iu Flex in the MOCN RNC is
similar to that in the non-shared RNC. In the MOCN RNC, the configuration for the initial
NAS message routing should be performed separately by each operator. The configuration is
as follows:
NRI length for the CS or PS domain
Mapping between NRI and global CN identity
Mapping between IMSI route value and global CN identity
IMSI route value = (IMSI div 10) mod 1000. The value range is 0-999.
CCP1 MOCN
2 CCP2 Shared
RNC
ALCAP
3 OM path
Operator A
Operator B
Common
RNC NodeB
Signaling node 1 Signaling node 1
Transport paths
MOCN MOCN
Shared Shared
RNC NodeB
Operator A
Operator B
Common
Traffic of Operator A
Traffic of Operator B
In RAN15.0 and later version, the introduction of 3.2.15 MOCN Independent Iub
Transmission Resource Allocation Featrue, to support the user plane transmission resource of
Iub interface independent, the Iub signal bandwidth shared by each operator, wherein MOCN
share a common channel of the cell bandwidth must be shared. This feature is only for the
RNC downlink transmission resources to be divided, does not support the NodeB side uplink
transmission resource partitioning.
MOCN Independent Iub Transmission Resource Allocation feature applies to the following
scenarios:
Operators share one physical link over the Iub interface, as shown in Figure 5-11.
Figure 5-11 Sharing one physical link over the Iub interface
Each operator uses a dedicated physical link over the Iub interface, as shown in
Figure 5-12.
Figure 5-12 Using dedicated physical links over the Iub interface
In MOCN scene, the customer requirements that the Iub bandwidth must assign by the
operator, you can deploy this feature.
Pre-deployment need to evaluate to prevent the traffic operators assigned Iub bandwidth is
less than the traffic demand.
If the inter-operators independence requirements of resources, you can deploy this feature. To
protect operator bandwidth usage completely independent, the total bandwidth of
configuration of the all logical ports bandwidth do not recommend more than the physical
port bandwidth.
Figure 5-13 Iur control plane of two RNCs shared by operators A and B
RNC 1 SPC 1 RNC 2 SPC 2
Control links
MOCN MOCN
Shared ... Shared
RNC 1 RNC 2
Load sharing
Operator A
Operator B
Common
Figure 5-14 Iur user plane of two RNCs shared by operators A and B
RNC 1 RNC 2
Signaling node 1 Signaling node 2
Transport paths
MOCN MOCN
Shared Shared
RNC 1 RNC 2
Operator A
Operator B
Common
Traffic of Operator A
Traffic of Operator B
5.4 OM Design
The OM mode of an MOCN network is closely related to that negotiated between operators.
OM modes are classified into joint OM mode and rent OM mode, which are described as
follows:
In joint OM mode, operators A and B provide respective devices to establish a joint
company and the joint company operates and maintains the network.
In rent OM mode, operator A shares some devices with operator B and operator A
operates and maintains the network.
In either of the two OM modes, a network is managed by one EMS or M2000, and then the
public and private data are transmitted to the respective system of each operator through the
northbound interface(NBI). Currently, an MOCN network cannot be managed on operators'
respective M2000s.
An MOCN shares FM and CM. Each operator can manage and monitor all shared cells to
obtain alarms and configurations of these cells. A configuration change in shared cells made
by an operator affects other operators. Therefore, it is recommended that the joint OM team or
the primary operator team change the configuration. PM can be partially dedicated to each
operator. That is, each operator can obtain the statistics about independent RNC-level traffic
volume, for example, CS traffic (unit: Erlang) and PS throughput. However, each operator
still needs to share the performance statistics of shared cells. Operator-specific statistics can
be subscribed to or queried on the M2000. Currently, the northbound interface cannot report
operator-specific statistics.
If each shared operator does not have its own NMSs, one set of M2000 is required for
managing the network, and it is recommended that the joint OM team or the primary operator
team operate and maintain the network.
When MOCN Cell Resource Demarcation is active, handover switch between different
operators cannot be opened.It is not recommended that this feature be used with the feature
WRFD-010696 DC-HSDPA because enabling Dual-Carrier High Speed Downlink Packet
Access (DC-HSDPA) results in inaccurate statistics on HSDPA power. DC-HSDPA takes
statistics of HSDPA power resources based on two cells whereas MOCN takes statistics of
HSDPA power resources based on one cell.
RAN15.0 introduces MOCN CE separate allocation of CE resources (details please refer to
3.2.14 MOCN Independent CE Resource Allocation), support the operators to allocate the CE
License resource by NodeB level. if the customer have this requirement, you can deploy this
feature.
1) Before you deploy this feature, allocate CE resources to operators based on their
traffic:If among the operators the independent request to resources is high, then the
sharing group may not configure CE, only configures in the private group; But the total
CE utilization of NodeB resources will decline.
2) If the operators consider the balance between independence and resource utilization, then
the sharing group may configure more CE, in the sharing group CE is more, the
utilization is higher, the independence is lower.
Sum of Various operators' private group CE and sharing group CE cannot exceed the purchase
of the CE license document total.
Outer operator: In mobility management, UEs need to be handed over from the operator
of the SRNC to the operator of the TRNC. If an operator is not a primary operator or a
secondary operator of the SRNC, the operator is defined as an outer operator to support
mobility management.
Common operator: The PLMN ID is indicated in the system broadcast information. UEs
that do not support HSDPA consider the corresponding operator as the serving operator,
as in conventional networks. Common operators support non-supporting UEs.
Originally, the primary operator and secondary operators are configured. To support the
MOCN, common operators are added. Considering that a cell under the neighboring RNC
may be an MOCN cell, outer operators are added for the purpose of convenient configuration.
Roaming Capacity 10 20 10
Then, the roaming capacity ratio among B, C, and D is 10:20:10, that is, 1:2:1. That is,
when UEs of operator A roam to the shared RNC, one fourth of the UEs will be routed to
operator B's CN nodes, two fourths to operator C's CN nodes, and one fourth to operator
D's CN nodes.
To share the roaming relationship table between operators, the delay and number of
initial access attempts can be decreased. As specified in the feature, each operator can be
configured with a maximum of 512 PLMNs with roaming relationships. The
configuration of the roaming relationship table increases OM workload of operators
because roaming relationships may change frequently or be added and there is a large
amount of data.
2. Inventory evolution
In the current MOCN market, UEs select MOCNs using the default policy, that is,
roaming UEs are routed geographically. In the specified geographical area, roaming UEs
are routed to the default operator. To route roaming UEs in proportion, use slow
evolution or fast evolution.
Slow evolution
The RNC software is upgraded, and new roaming UEs that access a shared area are
routed in proportion. However, the roaming UEs that have registered before the RNC
software upgrade still camp on the previous operators' CN nodes, because NRI
mapping exists. In this way, roaming UEs in the entire network are not routed in
proportion, that is, this feature takes effect slowly. Roaming UEs move frequently.
Therefore, roaming UEs will roam out of the previous area after a short time. When
roaming UEs move to the MOCN cells, the UEs are routed in proportion. Therefore,
in the entire network, this feature takes effect gradually.
Fast evolution
The RNC software is upgraded, UE registration information is removed from CNs,
and all UEs reregister and reselect networks within one or two days. Roaming UEs
register for the first time, the CN delivers consistency requirement, and the RNC
enables roaming UEs to be routed in proportion. Subscription UEs register back to
the home operators' CN nodes. This feature takes effect fast. However, the removal of
UE registration information prolongs registration time of UEs because the data needs
to be obtained from the HLR. If the registration information of only roaming UEs is
removed, the registration of only roaming UEs is prolonged.
3. Feature verification
Assume that operator A and operator B deploy an MOCN network and 75% roaming
UEs are expected to be routed to operator A's CN nodes and 25% roaming UEs to
operator B's CN nodes.
When UEs that have a roaming agreement with both operators register in the network,
the RNC obtains their IMSIs and performs a mod 4 operator. The UEs of the IMSIs with
the remainders of 0, 1, and 2 will be routed to operator A's CN nodes and the UEs of the
IMSIs with the remainder of 3 will be routed to operator B's CN nodes, routing roaming
UEs in the ratio of 3: 1. After IMSI mod is used, CS and PS domains can register in the
same operator even though they register separately.
Within a short time, roaming UEs are routed not exactly in a specified proportion. In the
long term, roaming UEs are routed in proportion. If required, feature verification can be
performed in a lab. However, IMSIs of UEs need to be selected properly.
4. License control
No license is used for this feature.
-1&SPARE_2_SWITCH-0&SPARE_3_SWITCH-0&SPARE_4_SWITCH-0&SPARE_5
_SWITCH-0&SPARE_6_SWITCH-1;
If inter-operator handovers are allowed in the MOCN, the InterPlmnHoAllowedIntraRat or
InterPlmnHoAllowedInterRat parameter must be set to ON.
If DefaultCnOp is set to a value ranging from 1 to 3, the default load balancing mode is used.
If DefaultCnOp is set to 255, the round robin selection mode is used for load balancing.
The resource management mode is specified for Iub bandwidth allocation, which corresponds
to the RscMngMode parameter in the ADD NODEB command.
A dedicated NodeB belongs to only one operator. Therefore, resource management for
multiple operators does not exist. During the dedicated NodeB configuration, RscMngMode
is set to SHARE by default.
A dedicated NodeB and a shared NodeB can coexist under an RNC, but the RAN Sharing
NodeB and MOCN NodeB cannot.
To prevent handovers between different operators, it is recommended that all NodeBs under
an RNC be shared unless otherwise specified.
Currently, the number of commercial RAN Sharing offices is smaller than the number of
commercial UMTS offices. However, RAN Sharing offices are facing a growth as operators
pay more attention to the Capital Expenditure (CAPEX) and Operational Expenditure
(OPEX).
The following table lists the commercial RAN Sharing offices of Huawei.
9 Appendix
9.1.2 MOCN
Script for the RAN11.0 commercial office: