Oeo102090 Lte Eran2.1 Anr Feature Issue 1.01
Oeo102090 Lte Eran2.1 Anr Feature Issue 1.01
Oeo102090 Lte Eran2.1 Anr Feature Issue 1.01
1 ANR Feature
Operation and maintenance (OM) of the radio access network has become increasingly
complex, difficult, and costly because of a huge amount of network elements,
implementation of different system standards, and coexistence of different equipment
vendors and telecom operators. To overcome this situation, the Self-Organization Network
(SON) concept is proposed. The main functions of SON are self-configuration, self-
optimization, and self-healing. ANR is a self-optimization function. It automatically
maintains the integrity and effectiveness of Neighbor Cell Lists (NCLs) to increase handover
success rates and improve network performance. In addition, ANR does not require
manual intervention, which reduces the costs of network planning and optimization.
Neighbor relations are classified into normal and abnormal neighbor relations. Abnormal
neighbor relations exist in the cases of missing neighboring cells, unstable neighbor
relations, PCI collisions, and abnormal neighboring cell coverage. ANR automatically
detects missing neighboring cells, PCI collisions, and abnormal neighboring cell coverage
and analyzes neighbor relations.
Based on neighbor relations, ANR is classified into intra-RAT ANR and inter-RAT ANR.
Based on the methods of measuring neighboring cells, ANR is classified into event-
triggered ANR and fast ANR (also known as periodic ANR).
The NRT contains the following information, which can be updated automatically or
manually:
SN: denotes the serial number of a neighbor relation. For E-UTRAN cell, maximum
to 32 intra-frequency neighbor relations and 64 inter-frequency neighbor relations
can be maintained in the table
LCI: Local cell ID, identifies the source cell of a neighbor relation.
TCI: identifies the target cell of a neighbor relation. This attribute is defined by the
ECGI of the target cell.
NO Remove: indicates whether a neighbor relation can be removed from the NRT
by ANR. By default, the NO Remove attribute of a neighbor relation is set to FALSE.
If the NO Remove attribute of a neighbor relation is TRUE, this neighbor
relation cannot be removed from the NRT.
If the NO Remove attribute of a neighbor relation is FALSE, this neighbor
relation can be removed from the NRT.
NO HO: indicates whether this neighbor relation can be used for a handover. By
default, the NO HO attribute of a neighbor relation is set to FALSE.
If the NO HO attribute of a neighbor relation is TRUE, this neighbor relation
cannot be used for a handover.
If the NO HO attribute of a neighbor relation is FALSE, this neighbor
relation can be used for a handover.
Temporary NRT has the same data structure as the NRT. Each cell has an intra-RAT
TempNRT but does not have an inter-RAT TempNRT. After detecting a new intra-RAT
neighbor relation, the eNodeB adds it to the intra-RAT TempNRT. Then, the eNodeB
regularly analyzes the neighbor relation in the TempNRT. If the new neighbor relation is
normal, the eNodeB adds it to the intra-RAT NRT.
HO blacklist contains the information about neighbor relations that cannot be removed
automatically from the NRT by ANR or used for a handover. The neighbor relations in the
HO blacklist must meet the following conditions:
NO Remove = TRUE
NO HO = TRUE
A neighbor relation is added to the HO blacklist if one of the following conditions is met:
The neighbor relation is unstable or leads to a low handover success rate but does
not reach the threshold for automatically removing the neighbor relation.
The neighbor relation is removed and added frequently, causing a ping-pong
effect.
HO white list contains the information about neighbor relations that cannot be removed
automatically from the NRT by ANR but can be used for a handover. The neighbor
relations in the HO white list must meet the following conditions:
NO Remove = TRUE
NO HO = FALSE
A neighbor relation is added to the HO whitelist in some special cases. For example: one
eNodeB must stop servicing if it needs to be maintained for a long period. As a result, the
NRTs of this eNodeB and its surrounding eNodeBs change significantly. To prevent such
changes and future time-consuming re-establishment of neighbor relations by using ANR,
telecom operators add associated neighbor relations to the HO white list. In this case,
these neighbor relations cannot be removed, and the eNodeB does not need to establish
them after it resumes.
Abnormal neighboring cell coverage refers to the coverage of a cell that is not adjacent to
the serving cell but can be detected by a UE in the serving cell. The eNodeB regards this
cell as a neighboring cell of the serving cell and therefore attempts to add the neighbor
relation to the NRT. The signals of an abnormal neighboring cell are generally unstable and
therefore the success rate of handovers to this cell is low.
eNodeB calculates the number of handover attempts and the handover success rate for a
neighbor relation after the neighbor relation is added to the TempNRT.
At the end of a measurement period, the eNodeB checks the statistical result to analyze
the neighbor relation. The neighbor relations that do not meet the requirements are
removed from or kept in the TempNRT for later analysis.
At the end of a measurement period, the eNodeB removes a neighbor relation from the
TempNRT in one of the following cases:
The neighbor relation is removed from the intra-RAT TempNRT and is added to the
intra-RAT NRT if number of handover attempts to a neighboring cell is greater than
or equal to the value specified by the NCellHOStatNum parameter, and the
handover success rate is greater than or equal to the value specified by the
AddCellThd parameter.
The neighbor relation is removed from the intra-RAT TempNRT but is not added to
the intra-RAT NRT if the number of handover attempts to all neighboring cells is
greater than or equal to the value specified by the NCellHOStatNum parameter,
and the number of UE reports for a neighboring cell is smaller than a specified
threshold specified by the DelCellThd parameter.
Otherwise, the eNodeB keeps the neighbor relation in the intra-RAT TempNRT for the
further analysis.
During automatic analysis of neighbor relations in the intra-RAT NRT, the eNodeB
calculates the number of handover attempts and the handover success rate for a neighbor
relation after the neighbor relation is added to the intra-RAT NRT from the intra-RAT
TempNRT. At the end of a measurement period, the eNodeB checks the statistical result to
analyze the neighbor relation. The neighbor relations that meet the requirements are
added to the intra-RAT NRT.
At the end of a measurement period, the eNodeB add a neighbor relation to the intra-RAT
TempNRT from the intra-RAT NRT in one of the following case:
The number of handover attempts to a neighboring cell is greater than or equal to
the value specified by the NCellHOStatNum parameter, and the handover success
rate is greater than or equal to the value specified by the DelCellThd parameter.
The number of handover attempts to all neighboring cells is greater than or equal
to the value specified by the NCellHOStatNum parameter, and the number of UE
reports for a neighboring cell is smaller than a specified threshold.
Otherwise, the eNodeB keeps the neighbor relation in the intra-RAT NRT.
A PCI collision means two cells in an NCL have the same PCI but different ECGIs. PCI
collisions may be caused by improper network planning or abnormal neighboring cell
coverage.If two neighboring cells have the same PCI, interference will be caused.
After adding a neighbor relation to its intra-RAT NRT, the eNodeB compares the PCI of the
new neighboring cell with the PCIs of existing neighboring cells. If the new neighboring
cell and any existing neighboring cell have different ECGIs but the same PCI, the eNodeB
reports PCI collisions to the M2000. The M2000 collects statistics about PCI collisions,
generates a list of information about PCI collisions, and reports PCI collision alarms.
PCI reallocation is a process of reallocating a new PCI to a cell whose PCI collides with the
PCI of another cell. The purpose is to eliminate PCI collisions. The M2000 triggers the PCI
reallocation algorithm to provide suggestions on PCI reallocation.
Abnormal neighboring cell coverage refers to the coverage of a cell that is not adjacent to
the serving cell but can be detected by a UE in the serving cell. Abnormal neighboring cell
coverage decreases the handover success rate because of abnormal neighbor relations it
has introduced. Therefore, detecting and eliminating abnormal neighboring cell coverage
play an important role in network optimization.
If the IntraRatEventAnrSwitch check box is selected, the M2000 triggers the algorithm
for detecting abnormal neighboring cell coverage and listing abnormal neighboring cells
when:
The M2000 receives from the eNodeA a request for adding a new neighbor
relation to the intra-RAT NRT.
Operator request for querying the information about abnormal neighboring cell
coverage.
Assume that cell A and cell B are involved in a handover. The UE is under the coverage of
cell A, and cell B may be an abnormal neighboring cell of cell A. Figure 4-3 shows the
process of detecting abnormal neighboring cell coverage.
The process of detecting abnormal neighboring cell coverage is as follows:
Abnormal cell detection algorithm triggered
It queries the longitudes and latitudes of cell A (serving cell) and cell B (target cell).
It calculates the distance between cell A and cell B.
It checks whether the distance between cell A and cell B exceeds the maximum
distance between two normal neighboring cells. If yes, go to step 5. If no, the
coverage of cell B is considered as normal.
It calculates the number of layers between cell A and cell B.
Note: The number of layers between cell A and cell B refers to the number
of non-intra-eNodeB neighboring cells dropping in the circle whose
diameter is the line connecting the centers of the two cells.
It checks whether the number of layers exceeds the maximum value. If yes, the
coverage of cell B is considered abnormal. If no, the coverage of cell B is
considered as normal.
Before a UE performs handovers, the eNodeB can obtain the information about all
neighboring cells with certain RSRP (it is specified by the FastAnrRsrpThd parameter)
based on the reporting of periodic UE measurements. This reduces the impact of event-
triggered UE measurements on handover performance when the UE performs handovers.
After intra-RAT fast ANR is activated, the eNodeB automatically detects missing
neighboring cells by using periodic UE measurements of intra-RAT intra-frequency or inter-
frequency cells. The UE involved in periodic measurements reports all detected neighboring
cells to the eNodeB at a regular interval, which is specified by the FastAnrRprtInterval
parameter. Then, the eNodeB adds the missing neighboring cells to its intra-RAT NCL. The
procedure for detecting missing neighboring cells by using periodic UE measurements is
the same as that by using event-triggered UE measurements.
When the number of involved UEs reaches the upper limit, the eNodeB does not select a
new UE for periodic measurements until a UE stops periodic measurements. The upper
limit is specified by the FastAnrIntraRatMeasUeNum parameter.
Periodic UE measurements also increase the power consumption of a UE. Therefore, intra-
RAT fast ANR restricts the number of periodic measurement reports by each UE. When the
number of periodic measurement reports by a UE reaches the upper limit, the UE stops
periodic measurements so that the eNodeB can select another UE for periodic
measurements until the eNodeB deactivates intra-RAT fast ANR. The upper limit is
specified by the FastAnrRprtAmount parameter.
The total number of neighboring cells that meet the RSRP requirement is limited, and
periodic UE measurements have a negative impact on the uplink throughput of the
network. Therefore, intra-RAT fast ANR restricts the total number of UEs involved in intra-
RAT periodic measurements. The eNodeB checks whether the total number of involved
UEs reaches the upper limit at a regular interval, which is specified by the
FastAnrCheckPeriod parameter. The upper limit is specified by the
FastAnrIntraRatUeNumThd parameter. If yes, the eNodeB automatically deactivates
intra-RAT fast ANR. If no, periodic UE measurements continue.
Parameter description
Parameter
MO Description
ID
Cdma200
Indicates whether handovers of UEs to the associated
0HrpdNcel NoHoFlag
neighboring cell are prohibited.
l
Cdma200
Indicates whether automatic removal of the neighboring relation
0HrpdNcel NoRmvFlag
through the ANR algorithm is prohibited.
l
Cdma200
Indicates whether handovers of UEs to the associated
01XRTTN NoHoFlag
neighboring cell are prohibited.
cell
Cdma200
Indicates whether automatic removal of the neighboring relation
01XRTTN NoRmvFlag
through the ANR algorithm is prohibited.
cell
EutranIntr
Indicates whether handovers of UEs to the neighboring cell are
aFreqNCel NoHoFlag
prohibited.
l
EutranIntr
Indicates whether to allow automatic removal of the neighboring
aFreqNCel NoRmvFlag
relation through the ANR algorithm.
l
Parameter description
Parameter
MO Description
ID
FastAnrChe Indicates the fast ANR checking timer. When the timer is expired,
ANR
ckPeriod the eNodeB automatically checks whether to disable fast ANR.
Indicates the RSRP threshold for fast ANR. If the signal quality in
FastAnrRsr
ANR a neighboring cell reported by the UE is lower than the threshold,
pThd
the cell is not automatically added to the NRT.