Professional Documents
Culture Documents
Contents
1- FAQs
2- Reselection Frame Structure//Throughput Calculations etc
3- SIBs
3- Registration
4-Paging IDLE Mode
5-Handover
6-DL Power
Control Connected Mode
7-DL Scheduling
8-ANR
9-ICIC Self Optimization
10-MRO Network
FAQs
Frame Structure (FDD)
Related Concept
1- Radio Frame
2-Subframe(TTI)
3-Slot
4- Subcarrier
5- Resource Block (Scheduling Minimum Unit)
6- Resource Element
Channel
RB Subcarrier
BW
Number Number
(MHz)
1.4 6 72
3 15 180
5 25 300
10 50 600
15 75 900
20 100 1200
Downlink Calculation
Downlink maximum throughput = Number of RB × 12 (Number of Sub-carrier with
one RB) × 14 (Number of Symbols with a Sub-frame) × [ 1 - (RS overhead and PDCCH
overhead) ] × Modulation symbols efficiency × MIMO × 1000 (Number of Sub-frame
in one second) × Coding rate
Example:
Calculate the FDD LTE system 10M, 2 * 2 MIMO, 64QAM, the Coding rate is 1.
The single cell downlink physical layer theory rate = 50*12*14*(1-(2/21+1/21))*6*2*1000*1 =82.4Mbps
50 50 RB
12 One RB includes 12 sub-carrier
14 A sub-frame 14 symbol
6 64QAM each symbol represents 6 bits
2 2*2 MIMO
1000 1s=1000ms
2/21 RS overhead (total symbol of one RB=12*14=168, RS symbol number=16, 16/168=2/21)
1/21 PDCCH overhead (If downlink sub-frame PDCCH accounted for only a symbol, and the PDCCH
symbol is the first
symbol of the sub-frame, this is the minimal overhead in PDCCH, a downlink sub-frame occupies 8
subcarriers, so the
minimal PDCCH overhead is 8 symbols, 8 / (14 * 12) =8/168= 1/21.
82.4Mbps this is an ideal value, because the SCH, BCH also take up some of the resources, and consider
the coding rate,
the actual Downlink peak rate around 70Mbps
Uplink Calculation
Uplink maximum throughput = Number of RB × 12 (Number of Sub-carrier with one RB) ×
14 (Number of Symbols with a Sub-frame) × ( 1 - RS overhead ) × Modulation symbols efficiency ×
1000 (Number of Sub-frame in one second) × Coding rate
Example:
Calculate the FDD LTE system 10M, None MIMO, 16QAM, the Coding rate is 1.
The UE uplink physical layer theory rate = 46*12*14*(1-1/7)*4*1000*1=26.5Mbps
46 46 RB
12 One RB includes 12 sub-carrier
14 A sub-frame 14 symbol
4 16QAM each symbol represents 4 bits
1 Coding rate
1/7Pilot overhead
1000 1s=1000ms
UE cat4 does not support 64QAM and MIMO in uplink, and consider the PUCCH occupied 4RB, the pilot
overhead is 1/7,
the uplink can reach the peak rate 25.6Mbps, in fact should also consider the impact of sounding and
PRACH, the uplink
peak rate around 25Mpbs
Carrier Frequency EARFCN Calculation(3GPP : 36.104)
Channel raster
Table 5.7.3-1 E-UTRA channel numbers
Downlink Uplink
The channel raster is 100 kHz for all bands, which E-UTRA FDL_low [MHz] NOffs-DL Range of NDL FUL_low [MHz] NOffs-UL Range of NUL
Operating
means that the carrier centre frequency must be an Band
integer multiple of 100 kHz. 1 2110 0 0 – 599 1920 18000 18000 – 18599
2 1930 600 600 - 1199 1850 18600 18600 – 19199
3 1805 1200 1200 – 1949 1710 19200 19200 – 19949
The relation between EARFCN and the carrier frequency 21 1495.9 6450 6450 – 6599 1447.9 24450 24450 – 24599
…
in MHz for the uplink is given by the following equation 33 1900 36000 36000 – 36199 1900 36000 36000 – 36199
where FUL_low and NOffs-UL are given in table 5.7.3-1 and 34 2010 36200 36200 – 36349 2010 36200 36200 – 36349
NUL is the uplink EARFCN. 35 1850 36350 36350 – 36949 1850 36350 36350 – 36949
36 1930 36950 36950 – 37549 1930 36950 36950 – 37549
37 1910 37550 37550 – 37749 1910 37550 37550 – 37749
38 2570 37750 37750 – 38249 2570 37750 37750 – 38249
FUL = FUL_low + 0.1(NUL – NOffs-UL) 39 1880 38250 38250 – 38649 1880 38250 38250 – 38649
40 2300 38650 38650 – 39649 2300 38650 38650 – 39649
NOTE: The channel numbers that designate carrier frequencies so close to the operating band edges that the carrier
extends beyond the operating band edge shall not be used. This implies that the first 7, 15, 25, 50, 75 and 100
channel numbers at the lower operating band edge and the last 6, 14, 24, 49, 74 and 99 channel numbers at
the upper operating band edge shall not be used for channel bandwidths of 1.4, 3, 5, 10, 15 and 20 MHz
respectively.
Example
FDL (center Freq) = FDL_low + 0.1(NDL
(EARFCN) – NOffs-DL)
Or
NDL (EARFCN)=10*(1815-1805)+1200
NDL (EARFCN)=1300
RS Mapping on different AxA MIMO
It is worth noting that the position of
the reference signals is dependent on
the value of the
Physical Cell ID. As such, the system
performs a calculation (Physical Cell ID
mod 6) to
determine the correct offset. Figure A
illustrates two cells, each producing a
different
offset.
Figure A
QCI stands for QoS Class Identifier. This is a special indentifier defining the quality of packet
communication provided by LTE.
The range of the class is from 1 to 9. Each of this class is defined as in the following table (TS
23.203).
IDLE Mode Behavior
Idle Mode Overview
PLMN Selection
Cell selection & cell reselection
System Information reception
Tracking area registration
Paging monitoring procedure
Idle Mode Overview
A UE that is powered on but does not have an RRC connection to the radio network
is defined as being in idle mode. In the case of idle mode management, the eNodeB
sends configurations by broadcasting system information, and accordingly, UEs select
suitable cells to camp on. Idle mode management can increase the access success rate,
improve the quality of service, and ensure that UEs camp on cells with good signal
quality.
PLMN Selection
• Qrxlevmeas is the measured RX level in the cell (RSRP), expressed in decibels with reference to one milliwatt
(dBm).
• Qrxlevmin is the minimum required RX level (set in the eNodeB) in the cell, expressed in units of dBm.
• Qrxlevminoffset is the offset to Qrxlevmin. This offset is taken into account when the UE attempts to camp on a
cell in a higher-priority PLMN. That is, when camped on a cell in a VPLMN, the UE considers this offset
parameter, which was signaled from the associated cell in the higher-priority PLMN, in the Srxlev evaluation.
• Pcompensation is generated according to the function max(PMax - UE Maximum Output Power, 0). The value is
expressed in decibels (dB).
• PMax is the maximum allowed transmit power of the UE in the cell, expressed in units of dBm. It is used in
uplink transmission.
Cell Reselection
• The signal strength of both serving cell and
neighboring cells varies with the movement of UE
and so the UE need to select the most suitable
cell to camp on. This process is called cell
reselection.
Note: If the highest cell is unsuitable because is part of list of forbidden Tac for
roaming or it does not belong to registered PLMN or an EPLMN, the UE does
not consider this cell as candidate for reselection for a maximum of 300
seconds.
Inter-RAT/Inter-Frequency low Priority Cell Reselection
Decision
The MIB is transmitted over the BCH. The scheduling period of the MIB is 40 ms. The first
transmission of the MIB is scheduled in subframe 0 of radio frames for which the SFN mod 4 = 0,
and repetitions are scheduled in subframe 0 of the last three radio frames in each period.
System Information Type-1
Site ID : 0x12519=75033
Cell ID: 02 ( sec-C)
SIB3 Common cell reselection parameters for all the cells and intra-frequency cell
reselection parameters
System Information(Sib-4//Sib-6)
• TAC in SIB1:
For Subframe PO
The subframe number i_s of a PO is derived from the
following formula:
i_s =Floor (UE_ID/N) mod Ns
T=DRX Cycle
N=N is min(T,NB). The NB parameter specifies the number of PO subframes in a DRX cycle.
Based on the actual
configuration on the eNodeB, NB can be set to 4T, 2T, T, T/2, T/4, T/8, T/16, or T/32.
Ns =max(1,NB/T).
UE_ID is IMSI mod 1024.
SIB-2
Understanding of NB
SFN for PF
For Subframe PO
IMSI: IMSI(448835805669362) The subframe number i_s of a PO is derived from the
following formula:
N=N is min(T,NB) N=min(T,T) T=128 i_s =Floor (UE_ID/N) mod Ns
Ns =max(1,NB/T) Ns=max(1,NB/T)
Ns=max(1,T/T) 1
UE_ID is IMSI mod 1024 (448835805669362)
mod 1024=1010
SFN mod T=(128 div 128) x (1010 mod 128)= 114
i_s=Floor(UE_ID/N) mod Ns= Floor(1010/128) mod 1=
Floor(7.890625) mod 1=7 mod 1= 0
From calcuation:
11 12 12 12 12 12
0 1 2 3 4 5 6 7 8 9 10 11 12 13 … …. …. 4 … … 3 4 5 6 7 Ns=1
P
PF PF PF PF PF PF PF PF PF PF PF PF PF PF PF PF PF PF F PF PF PF PF PF PF I_S=0
0 1 2 3 4 5 6 7 8 9
P
O
Connected Mode
• Handover
• Power Control (DL)
• Scheduling (DL)
Handover Procedure
mobility robust
optimization (MRO)
Measurement Triggering
Only voice
Handover Events
Key Concept
Step Direction Message Target Cell Memo
RSRP=68-140=-
72dbm
RSRQ=(32-40)/2=-
4db
Data Forwarding
Example
PRS(dBm) PA(dB) PB
Recommendation
1 15.2 0 0
default
value ( Huawei
suggestion ) 18.2 -3 1
recommendation
2 19.9 -4.77 2
recommendation3 21.2 -6 3
• Max C/I
• l Round robin (RR)
• l Proportional fair (PF)
• l Enhanced proportional fair (EPF)
• Semi Persistent
• Dynamic
Semi-Persistent Scheduling
Semi-persistent scheduling is introduced to reduce the overhead of control signaling. Semi-
persistent scheduling is a process where one user uses the same time-and-frequency
resources in a specified semi-persistent scheduling period (20 ms in Huawei eNodeB) until
they are released. Semi-persistent scheduling is mainly used for processing services with a
constant rate, regular packet arrival, and low delay requirements, such as the Voice over IP
(VoIP). By adopting semi-persistent scheduling, VoIP services can save the overhead of
control signaling and increase the VoIP capacity.
Dynamic Scheduling
In dynamic scheduling, scheduling is performed every Transmission Time Interval (TTI) of 1
ms and all the UEs to be scheduled are notified with the scheduling information through
control signaling within this TTI. Dynamic scheduling has no requirements on the size and
arrival time of data packets. Therefore, dynamic scheduling is applicable for all services.
DL Scheduler
Downlink scheduling allocates time-and-
frequency resources at the Physical Downlink
Shared Channel (PDSCH) for transmission of
system messages and downlink data. Downlink
scheduling described in this chapter is based
on the EPF scheduling strategy.
The scheduling priority of control-plane data is only lower than that of VoIP services.
Control-plane data is subject to dynamic scheduling. Control-plane data consists of
common control messages and UE-level control messages. The scheduling of IMS
signaling is the same as that of UE-level control messages. Handover and Power
control is also UE-Level Control messages.
HAQR Retransmission Data
CQI
The service with higher spectral efficiency of the corresponding wideband CQI has a higher
priority.
UE differentiation factor
The UE differentiation factor reflects the priority of UEs of different levels. The UE with a
higher level set by operators has a higher priority in scheduling.
Channel quality
The instantaneous channel quality of the UE is taken into
account. The UE with better instantaneous channel quality
has a higher priority. In the case of the same channel quality,
the GBR service with QCI of 1 has a higher priority than other
GBR services.
Delay
The closer the waiting time of the first packet in the buffer is
to the Packet Delay Budget (PDB), the higher the priority is.
The PDB value depends on the QCI.
Relative priority
The prioritization of GBR services is different from that of
non-GBR services. This factor is added to compare the
priority of GBR services with that of non-GBR services.
MCS Selection & Resource Allocation
• An NCL of a cell contains the information about the neighboring cells of a cell.
Unless otherwise stated, neighboring cells mentioned in this document exclude
intra-eNodeB neighboring cells. NCLs are classified into intra-RAT NCLs and inter-
RAT NCLs. Each cell has one intra-RAT NCL and multiple inter-RAT NCLs.
• An NCL includes the ECGIs (for E-UTRAN cells) or CGIs (for inter-RAT cells), PCIs,
and EARFCNs of the neighboring cells.
• The eNodeB adds newly detected neighboring cells to the NCL. The NCL is used
as a basis for creating neighbor relations. Neighboring cells in the NCL can be
automatically managed (for example, added, deleted, or modified) by ANR. They
can also be managed manually.
An NRT of a cell contains the information about the neighbor relations
between a cell and its neighboring cells.
NRTs are classified into intra-RAT NRTs and inter-RAT NRTs. Each cell has
one intra-RAT intra-frequency NRT,
one intra-RAT inter-frequency NRT, and multiple inter-RAT NRTs. The
intra-RAT intra-frequency NRT and intra-RAT
intra-frequency NRT are referred to as the intra-RAT NRT in this
document.
shows an example of the NRT. The information in this table is for
reference only.
Table 3-1 An example of the NRT
SN LCI Local Cell TCI No Remove No HO
PLMN
A TempNRT is a temporary NRT. It has the same data structure as the NRT. Each cell has an
intra-RAT intra-frequency TempNRT and an intra-RAT inter-frequency TempNRT but does not
have an inter-RAT TempNRT. The Intra-RAT intra-frequency TempNRT and intra-RAT intra-
frequency TempNRT are referred to as the intra-RAT TempNRT in this document. After
detecting a new intra-RAT neighbor relation, the eNodeB adds it to the intra-RAT TempNRT.
Then, the eNodeB regularly maintains the neighbor relation in the TempNRT. If the new
neighbor relation is normal, the eNodeB adds it to the intra-RAT NRT.
Blacklist
HO Blacklist
An HO blacklist contains the information about neighbor relations that cannot be used for a
handover or removed automatically from the NRT by ANR. The neighbor relations in the HO
blacklist must meet the following conditions:
NO Remove = TRUE
NO HO = TRUE
A neighbor relation can be added to the HO blacklist manually.
X2 Blacklist
An X2 blacklist contains the information about an eNodeB and its neighboring eNodeBs. X2
interfaces cannot be set up automatically between the eNodeB and the neighboring eNodeBs. If
an X2 interface has been set up, it will be removed automatically.
Whitelist
HO Whitelist
An HO whitelist [1] contains the information about neighbor relations that can be used for a
handover but cannot be removed automatically from the NRT by ANR. The neighbor relations in
the HO whitelist must meet the following conditions:
NO Remove = TRUE
NO HO = FALSE
A neighbor relation can be added to the HO whitelist manually.
X2 Whitelist
An X2 whitelist contains the information about an eNodeB and its neighboring eNodeBs. The X2
interfaces established between the eNodeB and the neighboring eNodeBs cannot be removed
automatically
PCI
A PCI is the identifier of a physical cell. A maximum of 504 PCIs are supported, according to
reference document. Therefore, PCI collisions occur inevitably. PCI collisions negatively affect
handover performance and the handover success rate. For details about PCI collision handling,
• Intra-RAT ANR
• Intra-RAT Fast ANR
• Inter-RAT ANR
• Inter-RAT Fast ANR
Intra-RAT ANR
1. The source eNodeB delivers the intra/inter-frequency
measurement configuration to the UE and requests the
UE to measure intra-frequency neighboring cells that
meet the measurement configuration.
2. The UE detects that the PCI of cell B meets the measurement Source Neighbor
configuration and reports it to the source eNodeB. Then, the source
eNodeB checks whether the intra/inter-RAT NCL of cell A includes the
PCI of cell B. If yes, the procedure ends. If no, the following steps
continue.
3. The source eNodeB instructs the UE, using the newly discovered
PCI as a parameter, to read the ECGI, Tracking Area Code (TAC), and
PLMN ID list of cell B.
5. The UE reports the detected ECGI, TAC, and PLMN ID list of cell B to
the source eNodeB.
The source eNodeB adds the newly detected neighboring cell of cell B
to the intra-RAT NCL of cell A and adds the neighbor relation to the
intra-RAT TempNRT
Intra-RAT Fast ANR
Before a UE performs handovers, the eNodeB can obtain the information about all neighboring
cells with the signal quality reaching or exceeding 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.
Inter-RAT ANR
1. The source eNodeB delivers the inter-RAT
measurement configuration (including target RATs and
EARFCNs) to the UE, activates the measurement gap
mode, and instructs the UE to measure the neighboring
cells that meet the measurement configuration.
2. The UE detects that the PCI of cell B meets the
measurement configuration and reports it to cell A. If the
source eNodeB detects that its NCL does not include the
PCI of cell B, it proceeds to the following step.
5. The UE reports the source eNodeB the CGI and other parameters
of cell B.
The source eNodeB adds the newly detected neighboring cell to its
inter-RAT NCL and adds the neighbor relation to the inter-RAT NRT.
Inter-RAT Fast ANR
After inter-RAT fast ANR is activated, the eNodeB delivers the inter-RAT measurement
configuration to the UE and instructs the UE to detect neighboring GERAN, UTRAN, and CDMA
cells by using periodic measurements.
The principles of inter-RAT fast ANR are the same as those of intra-RAT fast ANR
PCI Collision Handling
A PCI collision occurs if 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 (also
known as cross-cell coverage). If two intra-frequency neighboring cells have the same PCI,
interference will be caused.
When a PCI collision occurs, the eNodeB cannot determine the target cell for a handover. This
deteriorates the handover performance and reduces the handover success rate. Therefore,
eliminating PCI collisions is an important issue in network optimization.
After a PCI collision is eliminated, the PCI is unique in the coverage area of the cell and unique in
the neighbor relations of the cell.
PCI collision detections are triggered after intra-RAT ANR updates neighboring cells. PCI collision
handling involves automatically detecting PCI collisions and reallocating PCIs.
PCI reallocation is a process of allocating a new PCI to a cell whose PCI collides with the PCI
of another cell. This aims to eliminate PCI collisions.
If Optimization Analysis Mode is set to Immediate or Scheduled, the M2000 triggers PCI
reallocation in the mode specified by the value of Optimization Analysis Mode. The M2000
also provides suggestions on PCI reallocation upon receiving a PCI collision alarm.
ICIC
Overview ICIC ICIC
All physical resource blocks (PRBs) occupied by user
equipment (UEs) in a cell are mutually orthogonal in
the frequency domain; therefore, intra-cell
interference is very low. However, inter-cell DL UL
interference is relatively high because the frequency
reuse factor is 1, in which case every cell can provide
services over the entire system band. For cell edge Static Dynamic Static Dynamic
users (CEUs), the impact of the inter-cell interference
is especially severe. Therefore, to increase the cell
capacity and CEU throughput, inter-cell interference
must be mitigated.
ICIC is a technology that collaborates with power control and media access control
(MAC) scheduling technologies to mitigate inter-cell interference. ICIC divides the entire
system band into three frequency bands and uses different frequency bands at the edge
of neighboring cells. CEUs, which cause high interference or may be sensitive to
interference, are preferentially scheduled in the cell edge bands to mitigate inter-cell
interference. The interference mitigation enhances the network coverage and improves
the CEU throughput
Technical Principles of ICIC
Key Concept:
A3 Event for ICIC
CEU/CCU
Power Control
MAC Scheduling
The relationships between the key techniques
are described as follows:
i) CEU/CCU identification is a technique of
identifying the UE type (CEU or CCU) based on
event A3.
ii) Edge band mode assignment is a technique
of allocating different edge bands to
neighboring cells. Edge band adjustment is a
technique of expanding or shrinking the edge
band of a cell based on inter-cell interference
and the cell load. Edge band mode assignment
and edge band adjustment collaborate to
determine the edge band of each cell.
iii) Power control and MAC scheduling
collaborate to allocate PRBs to UEs based on
cell edge bands and UE types. PRBs in edge
bands are mainly allocated to CEUs, and those
in center bands are mainly allocated to CCUs.
CEU/CCU Identification
Principles
When initially accessing a network, a UE is recognized as a CCU by the serving cell; after a
handover, the UE is recognized as a CEU by the target cell. After a short period following the
initial access or handover, the eNodeB starts to use event A3 for ICIC (referred to as ICIC event
A3 in this document) to determine whether the UEs are CEUs or CCUs.
Handovers in the LTE system are classified into intra-frequency and inter-frequency handovers.
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 incorrectly 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 measured
separately:
shows the ping-pong handover decision. When the ping-pong handover decision is performed,
period 2 is compared with the threshold for the ping-pong handover period. If the value of
period 2 is smaller than the value of the threshold PingpongTimeThd, 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 the handover.