WO - NO2001 - E01 - 1 UMTS Radio Network Optimization P70
WO - NO2001 - E01 - 1 UMTS Radio Network Optimization P70
Optimization Process
Contents
Operator
revenue
Quality of
¥ Operation
cost
service
Q
© ZTE Corporation. All rights reserved
Cause of Network Optimization
Engineering optimization
Engineering optimization refers to the network optimization conducted after
the completion of network construction and before the number allocation. Its
major purpose to enable the network to work normally and ensure that the
network can fulfill the planned coverage and interference objectives.
The major work contents of Engineering Optimization include:
Eliminating the hardware fault of the system
Checking whether the cell configuration is consistent with the planned objective of
the network
Making the coverage and interference reach a satisfactory level
Aligning test
Problem positioning
Optimizing
Making optimization plan adjustment plan
Network
performance meets
the requirements?
Network optimization
Optimization project acceptance report
Data archive
End
© ZTE Corporation. All rights reserved
Preparations
Spectrum scanning
With the authorization from the customer, scan and confirm the frequencies
used by the network at present in the optimization area to ensure that the
frequencies are clean and available.
Calibration test
Calibration test of vehicle-mounted antenna
Calibration test of the external antenna of the test mobile phone
Test of the average penetration loss of the vehicle body
Test of building loss
In terms of call duration, the call modes can be divided into continuous
long call and periodic call.
The difference between this two call modes is the duration. The call
duration for the continuous long call is as long as possible, while that
for the periodic call is fixed (the duration is determined by the real
situation.
The times of periodic call reflects the processing capability and the
result is closer to the user’s usage, while the continuous long time
call reflects the system’s switching capability.
Both scanner and test mobile phone can be used for network
data collection. However they are different from each other in
some way.
Test mobile phone is used to understand the use situation of the users
and collect the downlink information on the network. It implements the
following functions:
Call class event and performance statistics: including the statistics of call
drop rate, blocking rate, call success rate, handover success rate, voice
service quality and data service rate.
The same test tool and parameter setting shall be adopted before and after the
optimization whenever possible.
The same test antenna and feeder shall be adopted before and after the
optimization.
The same analyzers shall be sent to process the data.
The test path shall be selected before and after the optimization.
To guarantee the consistency of the UE movement speed, the distance-based
sampling mode instead of time-based sampling mode is adopted for sampling
the data. The DT tool cannot sample data in the distance-based mode, data
collected can be paused when the vehicle stops in the case of red light.
Check whether the loaded test is conducted in the test area. Ensure that the test is
conducted at the same time segment in each day to obtain the basically the same
network load conditions.
Conduct the test in the same period.
The counter value required for calculating the network KPI can
be extracted flexibly. Besides, the statistics can be conducted
Extraction
flexibly according to different statistic range, or performance
statistics table can be made according to the customer’s
requirement.
As the end user of the network service, ordinary user has more
direct experience of the network performance.
The user complaint information most directly reflects the disadvantage of the
network.
Users are most concerned with the problems included in the complaints,
which need be solved as soon as possible.
The complaint includes geographical location information.
Generally the problems complained include poor signal coverage quality,
difficulty for call completion and frequent call drop.
Graphical Analysis
CQT data analysis be used to obtain the call success rate, call drop
rate, call delay, conversation quality and average rate of data service.
Base on this, we can know the indoor coverage and interference of
the network in the selected area.
As the end user of the network service, ordinary users can feel the
network performance most directly. Due to the diversity and individuality
of each problem described by the user, the user complaint data involve
not only the base station side, but also the billing system and
transmission system. Therefore, the complain data need be carefully
identified to find the data that can actually reflect the network status.
The user complaint can directly reflect the problem phenomena and
geographical information. From the complaint the following information
can be collated out:
Area of poor signal coverage
Area of low call success rate
Area of high call drop rate
Area of poor conversation quality
It indices to compare different data sets from the same angle and
find the difference. If necessary, we can further find cause of the
difference. This method is generally used in signaling flow
analysis.
Find the Top N and Bottom N data from a large amount of data
according to a certain classification method. These data deserve
special attention, for example, the common worst cell method.
If the network scale is large, the network need be divided into different base
stations clusters to locate and solve the network problems in different areas.
After the all base station clusters have been optimized, the global network can be
optimized to solve global network problems and problems across different
clusters. Then assess the performance of the global network after the
optimization and verify whether the network performance indices have meet the
acceptance standard and reached optimization objectives.
The cluster division principles shall be adjusted according to the actual situation,
generally according to the topography of the local area. neighboring areas that
have special requirements for data or voice service shall be grouped in the same
cluster to facilitate optimization debugging. Besides, cluster division can also be
conducted according to the problems found in the network assessment before
the optimization. There shall be overlap between two neighboring clusters.
The optimization of different clusters can be implemented either parallel or
serially depending on the resources and time requirement.