OEO102100 LTE eRAN2.1 MRO Feature ISSUE 1.00 PDF
OEO102100 LTE eRAN2.1 MRO Feature ISSUE 1.00 PDF
1 MRO Feature
By identifying handover scenarios and counting the number of handovers in the scenarios,
the MRO feature optimizes the handover-related parameter settings based on the counting
results, minimizing handover failures, service drops, and undesirable handovers. The MRO
feature optimizes the handover-related parameter settings through the following process:
eNodeB identification: For intra-RAT MRO, it request the X2 interface which need
setup by the same vedor’s eNodeB. Firstly, eNode should detect whether the peer
eNodeB belong to the same vendor.
Handover scenario identification: The characteristics of abnormal handovers are
analyzed, and the scenarios for the premature handover, delayed handover, and
ping-pong handover are defined. During a handover, the MRO feature identifies
the handover scenario.
Handover scenario handling: Within an MRO period, the eNodeB counts the
number of abnormal handovers that are identified in the handover scenario
identification process. When an MRO period arrives, the eNodeB determines how
to change the values of parameters based on the number of abnormal handovers
and the threshold for the number of abnormal handovers.
Result monitoring: After the values of parameter are changed, the handover-
related counters are monitored. If the counters improve, the parameter settings are
not rolled back during the next MRO period. If the counters deteriorate, the
parameter settings are rolled back during the next MRO period.
Intra-RAT MRO is a process where the LTE parameters related to intra-frequency or inter-
frequency handover are optimized. An intra-frequency handover is triggered by event A3,
and an inter-frequency handover is triggered by events A2 and A4. Therefore, the
parameters whose values need to be changed are the Cell Individual Offset (CIO)
parameter of event A3, the CIO parameter of event A4, and the threshold for event A2.
In the case of Huawei eNodeB, an RLF occurs in the source cell and the RRC connection is
reestablished in a non-source cell. In this situation, you can infer that the handover is delayed
because the UE is moved out of the coverage of the source cell.
Handovers in the LTE system are classified into the intra-frequency handover and inter-frequency
handover. An intra-frequency handover is triggered by event A3, and an inter-frequency handover is
triggered by events A2 and A4. Event A2 determines whether to deliver inter-frequency
measurement reports. If the parameters related to event A2 or A4 are improperly set, the delivery of
inter-frequency measurement reports is delayed, resulting in a delayed handover. Delayed inter-
frequency handovers are classified into the following types, which are counted separately:
Delayed handover unrelated to event A2: If a measurement report for event A2 is available
and the eNodeB successfully delivers the inter-frequency measurement report and activates
gap patterns, a delayed handover occurs because the UE is moved out of the coverage of
the source cell.
Delayed handover related to event A2: If no measurement report for event A2 is available,
or if the eNodeB fails to deliver the inter-frequency measurement report or activate gap
patterns, a delayed handover occurs because the UE is moved out of the coverage of the
source cell.
Delayed handover related to event A4: The source eNodeB receives the measurement
report for event A4, and then requests an inter-frequency handover. However, the UE is
moved out of the coverage of the source cell without receiving the inter-frequency
handover command. Therefore, an RLF occurs and the RRC connection is reestablishment in
a non-source cell with better signal quality.
Ping-Pong handover is to define handover is performed between the source and target
back and forth.When the ping-pong handover decision is performed, camping time is
compared with the threshold for the ping-pong handover period. If the value of camping
time is smaller than the value of the threshold PingpongTimeThd, you can infer that a
ping-pong handover occurs. A ping-pong handover indicates that cell B has poorer signal
quality than cell A and therefore is not qualified as the target cell for handover.
When the preceding conditions are met, the number of premature handovers and the
number of delayed handovers are compared. If the number of premature handovers is
greater than the number of delayed handovers, one step is decreased for the CIO
parameter of event A3. That is, the radius of the neighboring cell is reduced, which puts
back the premature handovers. If the number of premature handovers is smaller than the
number of delayed handovers, one step is increased for the CIO parameter of event A3.
That is, the radius of the neighboring cell is enlarged, which brings forward the delayed
handovers.
Note that the number of delayed handovers is multiplied by a weighing factor before the
comparison because the probability of service drops caused by delayed handovers is higher
than the probability of service drops caused by premature handovers. This ensures that
parameter optimization is performed properly.
If the proportion of RLFs in the current period is greater than the proportion of RLFs in the
previous period, the parameter settings should be rolled back.
Inter-RAT MRO is a process where the parameters related to the handover between an LTE
cell and a non-LTE cell are optimized. Although non-LTE cells are involved in inter-RAT
MRO, only LTE parameters can be optimized. Therefore, the decision on whether the peer
eNodeB is from Huawei is not performed during inter-RAT MRO, eliminating the need for
eNodeB identification.
To trigger an inter-RAT handover, event A2 triggers an inter-RAT handover measurement
of the UE, and then the UE requests an inter-RAT handover if the conditions for triggering
event B1 are met. Therefore, the threshold for event A2 or the hysteresis, threshold, or
time-to-trigger for event B1 can be modified if inter-RAT handovers are required at an
earlier or later time.
It is recommended that the time to trigger measurement not be changed because the
trigger time has a major impact on measurement results. The threshold for event B1 is an
absolute value for the signal quality in the case of handover. Modifying the threshold for
event B1 to determine an earlier or later handover is risky. For example, if the threshold is
modified to a lower level for an earlier handover to another system, a handover may occur
when the signal quality of the cell in the other system is not good enough, and service
drops may also occur.
It can be concluded on the basis of the preceding analysis that the hysteresis for event B1
can be modified within a certain range for a later handover if a premature inter-RAT
handover occurs. In case of a delayed inter-RAT handover, the threshold for A2 should be
modified if the handover is related to event A2. In case of a delayed inter-RAT handover,
the hysteresis for event B1 should be modified if the handover is related to event B1.