Professional Documents
Culture Documents
Issue Draft A
Date 2020-12-29
and other Huawei trademarks are trademarks of Huawei Technologies Co., Ltd.
All other trademarks and trade names mentioned in this document are the property of their respective
holders.
Notice
The purchased products, services and features are stipulated by the contract made between Huawei and
the customer. All or part of the products, services and features described in this document may not be
within the purchase scope or the usage scope. Unless otherwise specified in the contract, all statements,
information, and recommendations in this document are provided "AS IS" without warranties, guarantees
or representations of any kind, either express or implied.
The information in this document is subject to change without notice. Every effort has been made in the
preparation of this document to ensure accuracy of the contents, but all statements, information, and
recommendations in this document do not constitute a warranty of any kind, express or implied.
Website: https://www.huawei.com
Email: support@huawei.com
Contents
1 Change History.........................................................................................................................1
1.1 eRAN17.1 Draft A (2020-12-29)........................................................................................................................................ 1
3 Related Concepts..................................................................................................................... 4
3.1 TA and TAL................................................................................................................................................................................. 4
3.2 MME Pool and MME Pool Area..........................................................................................................................................5
3.3 MME Identities......................................................................................................................................................................... 6
3.4 AS and NAS States.................................................................................................................................................................. 6
3.4.1 AS States................................................................................................................................................................................. 7
3.4.2 NAS States.............................................................................................................................................................................. 8
4 S1-flex...................................................................................................................................... 11
4.1 Principles.................................................................................................................................................................................. 11
4.1.1 MME Selection Based on NNSF Policies.................................................................................................................... 11
4.1.2 MME Selection Based on the High Rank Policy......................................................................................................18
4.1.2.1 MME Pool Selection...................................................................................................................................................... 20
4.1.2.2 MME Selection from an MME Pool......................................................................................................................... 23
4.1.2.3 MME Selection Based on Cell Configurations...................................................................................................... 27
4.1.2.4 MME Selection Based on DCN Range.....................................................................................................................27
4.1.3 Load Rebalancing Among MMEs................................................................................................................................. 31
4.1.4 MME Overload Handling................................................................................................................................................ 32
4.1.5 S1 Interface Failure Handling........................................................................................................................................36
4.2 Network Analysis.................................................................................................................................................................. 37
4.2.1 Benefits................................................................................................................................................................................. 37
4.2.2 Impacts.................................................................................................................................................................................. 37
4.3 Requirements......................................................................................................................................................................... 38
4.3.1 Licenses................................................................................................................................................................................. 38
4.3.2 Software................................................................................................................................................................................38
4.3.3 Hardware.............................................................................................................................................................................. 38
4.3.4 Networking.......................................................................................................................................................................... 38
4.3.5 Others.................................................................................................................................................................................... 40
4.4 Operation and Maintenance............................................................................................................................................. 40
4.4.1 Data Configuration........................................................................................................................................................... 40
4.4.1.1 Data Preparation............................................................................................................................................................ 40
4.4.1.2 Using MML Commands............................................................................................................................................... 45
4.4.1.3 Using the MAE-Deployment...................................................................................................................................... 46
4.4.2 Activation Verification..................................................................................................................................................... 46
4.4.3 Network Monitoring......................................................................................................................................................... 50
5 Parameters.............................................................................................................................. 51
6 Counters.................................................................................................................................. 52
7 Glossary................................................................................................................................... 53
8 Reference Documents...........................................................................................................54
1 Change History
Technical Changes
Change Description Parameter RAT Base Station Model
Change
Editorial Changes
Revised descriptions in this document.
This document only provides guidance for feature activation. Feature deployment and
feature gains depend on the specifics of the network scenario where the feature is
deployed. To achieve the desired gains, contact Huawei professional service engineers.
Software Interfaces
Any parameters, alarms, counters, or managed objects (MOs) described in this
document apply only to the corresponding software release. For future software
releases, refer to the corresponding updated product documentation.
3 Related Concepts
Figure 3-1 TA
NOTE
An MME allocates a TAL to a UE when the UE registers with the network or when
the UE updates its location. All TAs in the TAL are served by the same MME.
The UE initiates a location update when it moves out of the area covered by the
allocated TAL, not when it moves within the TAs in the TAL.
To learn the concepts of MME pool and MME pool area, see:
Figure 3-2 shows the relationships among MME pools, MME pool areas, and
eNodeBs.
Figure 3-2 Relationships between MME pools and MME pool areas
● Each MME in an MME pool connects to all eNodeBs in the MME pool area.
● Each eNodeB in the MME pool area connects to all MMEs in the MME pool.
● All eNodeBs in the same TA connect to MMEs in the same way.
● An MME pool area consists of one or multiple complete TAs.
● A UE does not need to change its serving MME when moving within the MME
pool area.
● MME pool areas may overlap each other, and the overlapping area must be
one or more complete TAs.
● When MME pool areas are overlapped, the TAs in the overlapping areas are
served by multiple MME pools. Each eNodeB in these TAs is connected to all
the MMEs that serve the area.
NOTE
Overlapping areas reduce the number of tracking area updates (TAUs) when a UE
moves in and out of an MME pool area repeatedly. For example, a central business
district (CBD) can be planned as an overlapping area to reduce the number of TAUs
when UEs move between the CBD and other areas.
MMEC
An MME code (MMEC) uniquely identifies an MME within an MME pool. Different
schemes for planning MMECs are available based on whether MME pool areas are
overlapped. For details, see MMEC Planning.
NOTE
MMEI
An MME identifier (MMEI) uniquely identifies an MME within a PLMN. It is
constructed from an MME group identifier (MMEGI, which uniquely identifies an
MME pool within a PLMN) and an MMEC; that is, MMEI = MMEGI + MMEC.
GUMMEI
A globally unique MME identifier (GUMMEI) identifies an MME globally. It is
constructed from a PLMN ID plus an MMEI; that is, GUMMEI = PLMN ID + MMEI.
NOTE
RRC-IDLE ● EMM- No No
DEREGISTERED
● ECM-IDLE
3.4.1 AS States
AS states are the connection states between a UE and an eNodeB. Based on
whether an RRC connection is set up, there are two AS states:
● RRC_IDLE
In this state, the eNodeB does not have the UE context. It can send system
information and paging messages to the UE.
● RRC_CONNECTED
In this state, the eNodeB has the UE context. It can instruct the UE to perform
data transmission and handovers, notify the UE of scheduling information,
receive information such as the channel quality from the UE, and send system
information to the UE.
For details, see 4.6 "EPS Mobility Management and Connection Management
states" in 3GPP TS 23.401 V10.7.0.
EMM
Based on the UE mobility management states on the EPC, NAS states are
classified as follows:
● EMM-DEREGISTERED
In the EMM-DEREGISTERED state, the EMM context in an MME holds no valid
location or routing information for the UE. The UE is unreachable by the MME
because the UE location is unknown.
● EMM-REGISTERED
In the EMM-REGISTERED state, the MME sets up and stores the UE context
and provides services for the UE.
Figure 3-4 shows the transition between the EMM-DEREGISTERED and EMM-
REGISTERED states.
ECM
Based on the UE signaling connectivity states managed by the EPC, NAS states are
classified as follows:
● ECM-IDLE
A UE is in the ECM-IDLE state when there is no NAS signaling connection
between the UE and the EPC. In this state,
– A UE performs cell selection or reselection.
– eNodeBs do not have contexts for the UE.
– There is no dedicated S1-MME connection for the UE between MMEs and
eNodeBs.
– The MME and UE maintain the TAL for the UE if the UE is still in the
EMM-REGISTERED state.
● ECM-CONNECTED
When a UE needs to transmit or receive data, it must access a cell and enter
the ECM-CONNECTED state. In this state:
– The eNodeB sets up and saves the UE context.
– There is a signaling connection between the UE and the MME.
– The UE is also in the EMM-REGISTERED state when the UE is in the ECM-
CONNECTED state.
Figure 3-5 and Figure 3-6 show transitions between the ECM-IDLE and ECM-
CONNECTED states.
● When switching from the ECM-IDLE state to the ECM-CONNECTED state, the
UE attempts to access the current cell. After receiving the access request from
the UE, the eNodeB sends the UE's NAS information to the MME and sets up
a dedicated S1-MME connection for the UE.
NOTE
4 S1-flex
4.1 Principles
With S1-flex, an eNodeB in the LTE/SAE network sets up an S1-MME connection
with each MME in an MME pool. When a UE accesses the network through the
eNodeB, the eNodeB selects an MME from the pool for the UE and sets up a
dedicated S1-MME connection with the MME. S1-flex involves the following
techniques:
● The eNodeB selects an MME based on the following policies:
– If a UE provides an MME identifier, the eNodeB selects the MME
identified by the MME identifier preferentially. This selection policy is NAS
node selection function (NNSF). For details, see 4.1.1 MME Selection
Based on NNSF Policies.
– If a UE does not provide an MME identifier or the MME identified by the
UE-provided MME identifier is unavailable, the eNodeB selects an MME
for the UE based on the priorities, capacities, and loads of MMEs. For
details, see 4.1.2 MME Selection Based on the High Rank Policy.
● MME load rebalancing balances the loads among all the MMEs in an MME
pool. For details, see 4.1.3 Load Rebalancing Among MMEs.
● eNodeBs handle MME overload according to the instructions in an Overload
Start message sent from the MME to the eNodeBs. For details, see 4.1.4 MME
Overload Handling.
● If the status of an S1 interface between an eNodeB and an MME is abnormal,
the eNodeB releases or redirects UEs in the RRC_CONNECTED state over this
S1 interface. For details, see 4.1.5 S1 Interface Failure Handling.
With S1-flex, the eNodeB selects MMEs for the following two types of UEs:
● LTE UEs
● UEs transferred back from a GERAN or UTRAN to E-UTRAN by cell reselection
the MME. If the eNodeB is connected to multiple MMEs, the eNodeB must select
one of these MMEs to set up a dedicated S1-MME connection. The function that
the eNodeB uses to select an MME based on the MME identifier reported by the
UE is called NNSF.
When a UE accesses the network through an eNodeB, the eNodeB must select one
MME from the connected MMEs and set up a dedicated S1-MME connection for
the UE. The dotted box in Figure 4-1 shows the NNSF procedure.
If... Then...
If... Then...
3. The eNodeB decides whether it can find the MME that is identified by the
MME identifier provided by the UE.
If... Then...
The eNodeB does not find the MME The eNodeB goes to 5.
based on NNSF policies
4. Based on the MME status and access reason of the UE, the eNodeB
determines whether the UE can continue to use the MME.
If... Then...
5. The eNodeB selects an MME based on the high rank policy. For details, see
4.1.2 MME Selection Based on the High Rank Policy.
6. The eNodeB sets up a dedicated S1-MME connection to the MME.
NNSF Policies
Figure 4-2 shows an example of the eNodeB searching for the required MME for
the UE.
Figure 4-2 Example of the eNodeB searching for the required MME for the UE
NOTE
● Strict match: The eNodeB selects the MME whose MMEGI and MMEC match
those sent by the UE. This policy is controlled by the
MME_STRICT_SELECTION_SWITCH option of the
CnOperator.OperatorFunSwitch parameter.
● Loose match: The eNodeB selects the MME whose MMEC matches that sent
by the UE.
If MME selection based on NNSF policies fails, the high rank policy is used. For
details, see 4.1.2 MME Selection Based on the High Rank Policy.
Under certain parameter settings, the eNodeB preferentially uses specific policies
to select MMEs for the two types of UEs. For details, see Table 4-1.
An LTE UE Strict
match >
High rank
An LTE UE Strict
match >
High rank
An LTE UE Strict
match >
High rank
An LTE UE Strict
match >
Loose
match >
High rank
An LTE UE Strict
match >
High rank
An LTE UE Strict
match >
Loose
match >
High rank
An LTE UE Strict
match >
High rank
NOTE
In the preceding table, ">" indicates that the policy on its left has a higher priority than that
on its right. For example, "Strict match > Loose match > High rank" indicates that the strict
match policy is preferentially used; if the strict match policy fails, the loose match policy is
used; if the loose match also fails, the high rank policy is used.
UE Type Determination
● If the RRCConnectionSetupComplete message reported by the UE contains the
GUMMEI TYPE IE, the eNodeB determines the UE type based on the value of
the IE, as described in Figure 4-2. For details, see 3GPP TS 36.300.
NOTE
This enhanced algorithm for deciding a UE type applies only to 3GPP Release 10 UEs.
In addition, the MME release as indicated by S1Interface.MmeRelease must be
Release_R10(Release 10) or later and the S1 interface must support the GUMMEI
TYPE IE.
If... Then...
If... Then...
NOTE
● When the eNodeB uses the most significant bit of the MMEGI to decide a UE type
and the operator has set the most significant bit of the LAC of the GSM or UMTS
network to 1, the eNodeB decides a UE as an LTE UE if the UE has fallen back from
the LTE network to the GSM or UMTS network and then reselected to the LTE
network.
● The value of the GUMMEI TYPE IE in the message reported by a UE reselecting to
the LTE network from a higher RAT (such as NR) is native. The eNodeB decides
this UE as an LTE UE.
● The eNodeB selects an MME pool based on the priorities, network topologies,
and average loads of MME pools. For details, see 4.1.2.1 MME Pool
Selection.
● The eNodeB selects an MME based on the priorities, capacities, and average
loads of MMEs in the MME pool. For details, see 4.1.2.2 MME Selection from
an MME Pool.
● The eNodeB selects an MME pool based on the priorities, network topologies,
and capacity proportions of MME pools. For details, see 4.1.2.1 MME Pool
Selection.
● The eNodeB selects an MME based on the priorities and capacity proportions
of MMEs in the MME pool. For details, see 4.1.2.2 MME Selection from an
MME Pool.
NOTE
If a frequency division duplex (FDD) cell and a time division duplex (TDD) cell under the
same eNodeB are served by different MMEs, the eNodeB selects an MME for a UE accessing
the FDD or TDD cell only from MMEs that serve the FDD or TDD cell. In this scenario, the
eNodeB selects an MME for a UE based only on MME capacities and loads but not the
MME pool and MME priorities. For details, see 4.1.2.3 MME Selection Based on Cell
Configurations.
The following describes the priorities, network topologies, average loads, and
capacity proportions of MME pools.
Network Topology
● Deciding whether network topologies are the same
Through signaling exchange over X2 interfaces, each eNodeB obtains the
information about the MME pools to which its neighboring eNodeBs are
connected. If an eNodeB is connected to the same MME pool as a
neighboring eNodeB, the two eNodeBs belong to the same MME pool area.
eNodeB topologies in MME pool areas 1 and 2 are different, because the numbers of
eNodeB5's neighboring eNodeBs in the two MME pool areas are different.
If the UE accesses the network through eNodeB 5, the probability that the UE
will move to the coverage area of eNodeB 1, 2, 3, 4, or 6 is much higher than
the probability of moving to the coverage area of other eNodeBs. Therefore,
selecting MME pool 1 can reduce the probability that the serving MME for the
UE will be changed.
Average Load
● Deciding whether the average loads of MME pools are the same
Average load of an MME pool = Sum of loads of all MMEs in the MME pool/
Number of MMEs. To learn the method of calculating the load of an MME,
see 4.1.2.2 MME Selection from an MME Pool.
● Selecting an MME pool based on the average loads
The eNodeB selects an MME pool with a lighter load.
NOTE
● If all MMEs in an MME pool are overloaded and the overload indication matches the
RRC connection setup cause, the eNodeB excludes this MME pool from the MME pool
candidate list. For details about the mapping between overload indications and RRC
connection setup causes, see 4.1.4 MME Overload Handling.
● If S1 interfaces between the eNodeB and all MMEs in an MME pool are faulty, the
eNodeB excludes this MME pool from the MME pool candidate list. For details about S1
interface fault handling, see 4.1.5 S1 Interface Failure Handling.
Capacity Proportion
● Deciding whether the capacity proportions are the same
Capacity proportion of an MME pool = Capacity of the MME pool/Total
capacity of all MME pools
NOTE
The capacity of an MME pool is sent to the eNodeB by the EPC and is indicated by the
value of MME Relative Capacity in the command output of DSP S1INTERFACE.
● Selecting an MME pool based on the capacity proportions
The eNodeB converts the capacity proportion of an MME pool to a probability
and uses this probability to select an MME pool for a UE. A higher capacity
proportion indicates a higher probability for an MME pool to be selected.
Relative load of the MME for non-NB-IoT UEs = Number of dedicated S1-
MME connections for non-NB-IoT UEs x 255/Relative capacity of the MME
Relative load of the MME for NB-IoT UEs = Number of dedicated S1-MME
connections for NB-IoT UEs x 255/Relative capacity of the MME
NOTE
For FDD, the eNodeB calculates the relative load based on the number of NB-IoT UEs and
selects an MME with the minimum relative load for NB-IoT UEs when the
UE_TYPE_BASED_MME_LOAD_BAL_SW option of the
GlobalProcSwitch.MmeSelectProcSwitch parameter is selected. In this case, the relative
load of MMEs on the EPC may change.
In some situations, the relative capacity for an MME can be set not based on the
actual processing capability. For example:
After an MME is added to an MME pool or the capacity of an existing MME is
expanded in the MME pool, the relative capacity of the MME can be set to a value
that is much greater than its actual processing capability. In this scenario, the load of
this MME increases quickly. When the load reaches a specified value, the operator can
update the MME's relative capacity based on its actual processing capability.
To remove an MME from an MME pool, the eNodeB is informed that the MME is
overloaded, preventing UEs from selecting this MME. After the UEs or MME releases
all calls, you can disable the MME and remove it from the MME pool.
For details, see 4.1.3 Load Rebalancing Among MMEs.
The eNodeB needs to set up an S1-MME connection to each MME in an MME
pool and maintains the relative capacity of each MME as follows:
If... Then...
The eNodeB The MME notifies the eNodeB of its processing capability
is setting up relative to the other MMEs in the same pool (relative
an S1-MME capacity) through the Relative MME Capacity IE.
connection
to an MME
The eNodeB The MME informs the eNodeB of its capability by using the
has set up an Relative MME Capacity IE in the Configuration Update
S1-MME message if the MME capability is updated.
connection
to an MME
NOTE
● If an MME is overloaded and the overload indication matches the RRC connection
setup cause, the eNodeB excludes the MME from the MME candidate list. For
details about the mapping between overload indications and RRC connection setup
causes, see 4.1.4 MME Overload Handling.
● If the S1 interface between the eNodeB and an MME is faulty, the eNodeB also
excludes the MME from the MME candidate list. For details about S1 interface
fault handling, see 4.1.5 S1 Interface Failure Handling.
The numbers of configured MMEGIs and MMECs must be equal to the number of
configured MMEs.
This function can work with the procedure described in 4.1.2.4 MME Selection
Based on DCN Range. For details, see descriptions of MME selection in 4.1.2.4
MME Selection Based on DCN Range for FDD and TDD cells that are served by
the same eNodeB but different MMEs.
MME selection based on DCN range is a procedure where the serving MME
designates a DCN for a UE during its access based on the UE usage type specified
upon UE subscription, and the eNodeB selects an MME for the UE based on the
DCN information provided by the serving MME.
Huawei eNodeBs support the DCN function introduced in 3GPP Release 13. This
function has no requirements for UE releases.
The DCN function is controlled by the DCN_SUPPORT_SWITCH option of the
CnOperator.OperatorFunSwitch parameter, and it can be deployed by PLMN.
During an attach or tracking area update (TAU) procedure, the eNodeB selects an
MME for the UE based on the access information provided by the UE. If the
selected MME does not support the service required by the UE, the MME triggers
an NAS message redirection procedure, as shown in Figure 4-10.
NOTICE
● eMTC UEs that the eNodeB has identified through PRACH time-frequency
resources can access only eMTC MMEs (for which
MmeCapInfo.MmeSupportEmtcDedEpcCap is set to a value other than
NOT_SUPPORT). The EPC does not have the MME range that is configured on
the eNodeB. Therefore, UE access fails if the MME group identified by the
MMEGI selected during NAS message redirection does not contain these MMEs.
Considering this, the eMTC MMEs configured on the eNodeB must be
consistent with those on the EPC that serve the eNodeB. Operators can spare
the task of configuring eMTC MMEs on the eNodeB and let the EPC assign
appropriate MMEs for UEs.
● If the range of MMEs that can be selected for UEs in a cell is configured
through the CellOp MO, UEs in this cell can access these MMEs only. If these
MMEs do not support all UE Usage Type values or the MME group identified by
the MME Group ID that the EPC selects during NAS message redirection does
not contain these MMEs, certain UEs in this cell will fail to access the MME. In
this case, do not set the range of MMEs on the eNodeB for such a cell and let
the EPC assign appropriate MMEs for UEs.
● If the access information of an NB-IoT UE contains only the RegisterMME IE
but not the S-TMSI IE, and the UE supports user plane CIoT EPS optimization,
the eNodeB preferentially selects an MME that supports user plane CIoT EPS
optimization. If such an MME is unavailable, the eNodeB selects an MME that
supports control plane CIoT EPS optimization. If the eNodeB is configured with
a default DCN, the same method applies.
NOTICE
When load rebalancing is initiated on an MME, the MME's relative capacity can be
changed and delivered to each connected eNodeB to prevent the eNodeBs from
selecting this MME.
During the RRC connection setup, the RRC connection request message does not
carry the S-TMSI, and the RRC connection setup completion message does not
carry the RegisterMME IE. The eNodeB then reselects an MME for the UE.
d. The eNodeB selects an MME for the UE and sets up a dedicated S1-MME
connection with MMEb. For details about how the eNodeB selects an
MME, see 4.1.2 MME Selection Based on the High Rank Policy.
● For an ECM-IDLE UE, the MME first initiates a paging procedure for the UE to
enter the ECM-CONNECTED state. The subsequent procedure is the same as
that for a UE in the ECM-CONNECTED state.
The MME load rebalancing function requires the collaboration of the MME,
eNodeB, and UE.
The eNodeB initiates an RRC connection release with the release cause "Load
Balancing TAU Required" if all the following conditions are met:
● The RRC Connection Request message of the UE contains an S-TMSI, and the
MME identified by the S-TMSI is overloaded.
● The overload indication matches the RRC connection setup cause.
● The UE is not redirected to another cell.
NOTE
When load rebalancing is initiated on a high-priority MME, the eNodeB still preferentially
selects this MME for UEs initiating TAU procedures. Therefore, you must lower the priority
of this MME.
If the MME informs the eNodeB that it is highly loaded and there are lightly loaded MMEs
under the same PLMN as the highly loaded MME, the eNodeB will determine whether the
UE should access the highly loaded MME when the
GlobalProcSwitch.EnbTrigMmeLoadRebalSwitch parameter is set to ON. If the eNodeB
determines that the UE should access the highly loaded MME, the eNodeB will initiate an
RRC connection release with the release cause "Load Balancing TAU Required".
If all MMEs in an MME pool are overloaded and the overload indication matches the
RRC connection setup cause, the eNodeB does not select this MME pool for UE access.
3. The MME and eNodeB take measures to reduce the MME load.
4. When the MME is no longer overloaded, it sends an Overload Stop message
to the eNodeBs.
5. After receiving the Overload Stop message, each eNodeB deletes the MME
overload indication and updates the information about the related MME pool.
Normal processing resumes.
When an MME sends an Overload Start message to an eNodeB, the eNodeB will not
allocate the newly accessed UEs that match the overload cause to the MME. UEs that
are already served by the MME, however, may still access the MME but will increase
the MME load.
The following are examples of eNodeB actions in the case of MME overload.
NOTE
For indications in the Overload Start message from the MME, see the Overload Action IE in
section 9.2.3.20 in 3GPP TS 36.413 V10.5.0 (2012-03). For RRC connection setup causes
contained in the RRCConnectionRequest message, see the EstablishmentCause IE of the
RRCConnectionRequest message described in section 6.2.2 in 3GPP TS 36.331 V10.5.0
(2012-03).
In FDD:
● After receiving an RRCConnectionRequest message from the UE in non–RAN
sharing scenarios, if all MMEs connecting to the eNodeB are overloaded and the
serving cell of the UE has no inter-RAT neighboring cells, the eNodeB directly
rejects the UE's RRC connection setup request by sending an RRCConnectionReject
message to the UE and specifies the period after which the UE can attempt to
access the eNodeB again.
● In RAN sharing scenarios, the eNodeB sends an RRCConnectionReject message to
the UE only when all MMEs are overloaded and the serving cell of the UE has no
inter-RAT neighboring cells.
2. The eNodeB selects an MME for the UE.
The eNodeB selects an MME based on the MME identifier provided by the UE,
MME priorities, MME capacities, MME loads, and the S1 interface status.
3. The eNodeB checks whether the MME is overloaded.
The eNodeB compares the RRC connection setup cause and the overload
indication to determine whether the UE can access the overloaded MME.
If... Then...
If... Then...
A fault in the S1 interface between an eNodeB and an MME affects MME pool
selection and MME selection in an MME pool. If a fault occurs on the S1 control
plane, the eNodeB cannot select the MME of the S1 interface for UEs that attempt
to access the network.
If the S1 interface status in the control plane is abnormal, the eNodeB handles the
UEs in the RRC_CONNECTED state over this interface as described in the following
table. The S1 interface status can be queried by running the DSP S1INTERFACE
command. For details, see 4.4.2 Activation Verification.
If... Then...
The eNodeB is connected The eNodeB releases all UEs that are carried on the
to other MMEs that faulty S1 interface. When UEs re-access the network,
belong to the same the eNodeB selects other available S1 interfaces for
serving operator the UEs.
If... Then...
The eNodeB is not The eNodeB initiates redirection for all UEs that are
connected to any other carried on this S1 interface.
MME that belongs to the
same serving operator
4.2.1 Benefits
S1-flex provides the following benefits:
● Signaling overheads for location updates are reduced because the serving
MME does not need to change when the UE moves in an MME pool area (the
area covered by an MME pool).
● Load balancing is implemented among MMEs in the same MME pool,
improving resource utilization.
● MME nodes are easy to add or remove with minimal impact on ongoing
services.
● MME selection is based on priorities, capacities and loads, which enhances
network reliability because each MME in an MME pool is a backup for the
other MMEs in the pool.
● MME selection based on priorities enables location-based MME backup,
improving network reliability.
4.2.2 Impacts
Network Impacts
When MME selection based on cell configurations is enabled by setting the
CellOp.MMECfgNum parameter to a non-zero value:
● The UE access delay may increase if the MME triggers a NAS message
redirection when a UE initiates an attach procedure or moves from one MME
to another.
● During an inter-DCN handover, the serving MME may proactively release the
UE after the inter-DCN handover is completed. In this case, UE services will be
interrupted. The MME releases the UE for it to initiate a TAU procedure, and
then the MME can initiate a NAS message redirection towards the
corresponding MME.
Function Impacts
None
4.3 Requirements
4.3.1 Licenses
None
4.3.2 Software
Prerequisite Functions
None
4.3.3 Hardware
Base Station Models
No requirements
Boards
No requirements
RF Modules
No requirements
4.3.4 Networking
MMEC Planning
Before deploying S1-flex, make an MMEC plan based on the following rules:
● If MME pool areas do not overlap, MMECs can be reused among the MME
pools.
B on the eNodeBs in area B, then UEs in area B will also access the backup
MMEs in normal situations.
● If some MMEs in area A need to be backup MMEs for eNodeBs in both area A
and area B, the backup MMEs cannot belong to the same MME pool as other
MMEs in area A.
4.3.5 Others
The EPC must support the S1-flex feature. It is not a Huawei proprietary feature
and can be used with EPC equipment from Huawei or other vendors. If the EPC
equipment is provided by Huawei, the version must be PS9.1 or later. If the EPC
equipment is provided by another vendor, check whether the vendor's equipment
supports this feature.
MMEs must support the pooling function. During S1-MME connection setup, these
MMEs must send the IEs MME-group-id, MME-Code, and MME-capacity that are
required for supporting S1-flex to eNodeBs. S1-flex imposes restrictions on the
planning of MME pools and TAs. For details about the planning rules, see 3
Related Concepts.
Operators must determine a list of MMEs to which each eNodeB needs to connect
based on the networking requirements, and configure the corresponding SCTP
links, S1 interfaces, transport-layer managed objects (MOs), and routes to the
MMEs based on this list.
Operators must plan one or more IP addresses and SCTP port numbers for S1
interfaces based on the networking requirements, and ensure that the
transmission links between eNodeBs and MMEs in MME pools are working
properly.
The functions "MME pool selection based on the network topology and average
loads" and "MME selection based on MME capacities and loads" depend on
whether the given MMEs support the MME pooling function.
Configure the MME priority (EP model). The mapping MO is S1. The following
table describes the key parameters.
Load Rebalancing
The following table describes the parameters that must be set in the
GlobalProcSwitch MO to configure the switch of eNodeB-triggered MME load
rebalancing.
Parameter Parameter ID Setting Notes
Name
MME Select GlobalProcS This parameter has the following two options:
Procedure witch.MmeSe MMESELECTALGOSWITCHFORGUUE and
Switch lectProcSwitc NNSFENHANCEMENTSWITCH
h
Set the two options to ON and OFF when the
following conditions are met:
Operators require that an eNodeB select an
MME with a light load for UEs accessing an
LTE network from a GSM or UMTS network,
and the EPC is required to support normal
access of UEs when the eNodeB selects an
MME that does not store the UEs' contexts.
Otherwise, it is not recommended that the
MMESELECTALGOSWITCHFORGUUE option
be selected.
(Optional) The following table describes the parameters that must be set in the
S1Interface MO to configure the release of the MME.
NOTE
If the DCN function is enabled on the EPC but the DCN_SUPPORT_SWITCH option of the
CnOperator.OperatorFunSwitch parameter is deselected, attach or TAU procedures of UEs
fail because the eNodeB does not respond to the NAS message redirection requests from
the EPC.
The following table describes the parameters that must be set in a DefaultDcnCfg
MO to configure a default DCN.
The following table describes the parameters that must be set in the
GlobalProcSwitch MO to configure the MME selection policies.
MME Select GlobalProcS If there are a large number of UEs on the live
Procedure witch.MmeSe network, you are advised to select the
Switch lectProcSwitc MME_OVERLOAD_BASED_SELECT_SW option.
h
Step 2 Use multiple UEs that have not been registered with the MMEs to access the
network and initiate services separately.
Then, check whether the UEs can perform services properly; for example, whether
the UEs can perform uplink and downlink File Transfer Protocol (FTP) services.
Step 3 Run the DSP S1INTERFACE command to query the value of S1 Interface User
Number for each S1 interface.
If the UE access statistics comply with the principles described in Principles, the
eNodeB performs load balancing among multiple MMEs when the UEs access the
network. The NNSF procedures are correct.
----End
MME Priority
Step 1 Run the DSP S1INTERFACE command to query the MME priority configuration, as
shown in Figure 4-16.
Step 2 Enable UEs to access the network and check whether the UE access statistics
comply with the principles described in 4.1.2.2 MME Selection from an MME
Pool.
If the UE access statistics comply with the principles, the priority-based MME
selection procedure is correct.
----End
NNSF
Step 1 Enable a UE to access the network. Then check the mmec IE in the
RRC_CONN_SETUP_CMP message. As shown in Figure 4-17, the value of the
mmec IE is 0F.
Step 2 Run the DSP SERVEDGUMMEIS command to query the GUMMEI of the serving
MME.
----End
If any of these alarms is generated, clear it by referring to 3900 & 5900 Series
Base Station Alarm Reference in 3900 & 5900 Series Base Station Product
Documentation.
5 Parameters
You can find the EXCEL files of parameter reference and used reserved parameter list for
the software version used on the live network from the product documentation delivered
with that version.
Step 2 On the Parameter List sheet, filter the Feature ID column. Click Text Filters and
choose Contains. Enter the feature ID, for example, LOFD-001016 or
TDLOFD-001016.
Step 3 Click OK. All parameters related to the feature are displayed.
----End
Step 1 Open the EXCEL file of the used reserved parameter list.
Step 2 On the Used Reserved Parameter List sheet, use the MO, Parameter ID, and BIT
columns to locate the reserved parameter, which may be only a bit of a parameter.
View its information, including the meaning, values, impacts, and product version
in which it is activated for use.
----End
6 Counters
The following hyperlinked EXCEL files of performance counter reference match the
software version with which this document is released.
● Node Performance Counter Summary: contains device and transport counters.
● eNodeBFunction Performance Counter Summary: contains all counters related
to radio access functions, including air interface management, access control,
mobility control, and radio resource management.
NOTE
You can find the EXCEL files of performance counter reference for the software version used
on the live network from the product documentation delivered with that version.
----End
7 Glossary
8 Reference Documents