INTERNAL
2013/1/16
LTE MIMO Techniques
eRAN2.2 (MIMO and Beamforming)
HUAWEI TECHNOLOGIES CO., LTD.
Huawei Confidential
www.huawei.com
Training Objectives
After completing this course, you will be able to:
Understand the concepts relevant to the E-UTRAN.
Understand basic LTE signaling procedures.
Master typical signaling procedures such as access procedure, dedicated bearer setup procedure, and handover procedure.
References:
3GPP TS 36.211: Physical Channels and Modulation
3GPP TS 36.213: Physical layer procedures
3GPP TS 36.306: User Equipment (UE) radio access capabilities
Training Objectives
After completing this course, you will be able to:
Understand the concepts relevant to the E-UTRAN.
Understand basic LTE signaling procedures.
Master typical signaling procedures such as access procedure, dedicated bearer setup procedure, and handover procedure.
References:
3GPP TS 36.211: Physical Channels and Modulation
3GPP TS 36.213: Physical layer procedures
3GPP TS 36.306: User Equipment (UE) radio access capabilities
Contents Background
and Overview of the LTE MIMO
Techniques Principles
and Application of the MIMO
Techniques Principles
and Application of Beamforming
Background of Multi-Antenna Techniques
Fifty years ago, Shannon gave the maximum efficiency that a time and frequency communication system can achieve.
S C B log 2 1 bit / s N
The rapid development of wireless communications poses increasingly higher requirement for system capacity and spectral efficiency. Various algorithms are invented, such as spreading the system bandwidth, optimizing the modulation scheme, or using complex CDMA. These methods are limited: Bandwidth cannot be expanded indefinitely; modulation orders cannot increase indefinitely; channels between a CDMA system are not ideally orthogonal. Another dimension, that is, MIMO, is invented to better use the spatial resource. As expressed in the following equation, if multiple antennas are used, the capacity is increased by a multiplication of the number of antennas used. S
C B log 2 1
bit / s M N
Advantages of Multi-Antenna Techniques
The LTE system improves system performance for cell edge users (CEUs) and brings stable and reliable service experience for users. Therefore, multiantenna techniques can make use of the spatial resource and increase the wireless transmission capacity many folds without increasing the transmit power and bandwidth.
Array
gain
Improved
system coverage Diversity
gain Improved
Spatial
multiplexing
system capacity
gain Co-channel interference
reduction
Increased
rate
peak
Increased spectral efficiency
Contents Background
and Overview of the LTE MIMO
Techniques Principles
and Application of the MIMO
Techniques Principles
and Application of Beamforming
Principles and Application of the MIMO Techniques
MIMO is an important technique in the LTE system. MIMO means use of multiple antennas at both the transmitter and receiver. MIMO can better utilize the spatial resource and increase spectral efficiency, achieving array gain, diversity gain, multiplexing gain, and interference rejection gain, providing higher system capacity, wider coverage, and higher user rate. data1
data2 data1
data2
data1 data2 data3
data1 data2 data3
data3 data1 data2 data3
High-speed code stream
data3 data1 data2 data3
Multiple low-speed code streams
Multiple crossed code stream
data1
data2
data3
Recovered high-speed code stream
Classification of MIMO Techniques
Depending on whether the spatial channel information is used, MIMO techniques are classified into open-loop MIMO and closed-loop MIMO. Open-loop MIMO: The UE does not feed back information, the eNodeB is not inf ormed of the UE situation. The protocols support single-stream (TM2) or multi-stream (TM3). Closed-loop MIMO: The UE feeds back information. The gain has a positive correlation with the accuracy of the feedback information. The protocols support single-stream (TM4) or multi-stream (TM6). At present, the feedback granularity supported by the reference signal in port 2 is large and closed-loop MIMO can hardly achieve gains. Closed-loop MIMO requires low UE mobility. At present, the eNodeB cannot accurately estimate the UE movement speed with an error of more than 30 km/h. Depending on the number of simultaneously transmitted spatial data streams, MIMO techniques are classified into spatial diversity and spatial multiplexing. Huawei eNodeBs in eRAN2.2 support the following MIMO modes. These modes are described in detail in the following pages. •
•
MIMO Technique
Multiantenna receive
MIMO Mode
Receive diversity
MU-MIMO
Multiantenna transmit
Open-loop transmit diversity Closed-loop transmit diversity Open-loop spatial multiplexing Closed-loop spatial multiplexing
Feature List in FDD UL 2-Antenna Receive Diversity UL 4-Antenna Receive Diversity UL Interference Rejection Combining UL 2x2 MU-MIMO UL 2x4 MU-MIMO 2x2 MIMO 4x2 MIMO DL 4x4 MIMO
Feature List in TDD UL 2-Antenna Receive Diversity UL 4-Antenna Receive Diversity UL Interference Rejection Combining UL 8-Antenna Receive Diversity UL 2x2 MU-MIMO UL 2x4 MU-MIMO 2x2 MIMO 4x2 MIMO
Multi-Antenna Receive MIMO
eRAN2.2 supports UL 2-Antenna Receive Diversity and optional UL 4-Antenna Receive Diversity and UL 8-Antenna Receive Diversity. The following figure shows the block diagram of receive diversity. The UE uses one antenna to transmit signals; different UEs use different time and frequency resources. The eNodeB uses multiple antennas to receive signals and combine the received signals to maximize SINR, therefore obtaining diversity gain and array gain, increasing the cell coverage and improving single-user capacity.
Mechanism of Signal Combination
An MMSE receiver uses receive beamforming targeted at a UE. The receiver adjusts the combined weight and changes the direction of the major lobe and side lobe to maximize the SINR of the received signals. There are two combination algorithms for UL receive diversity. Maximum ratio combining (MRC) and interference rejection combining (IRC) can both obtain diversity gain and array gain, improving system performance. MRC and IRC are suitable for environments with different interference characteristics. MRC receivers and IRC receivers are implementation of MMSE receivers in different scenarios.
Differences Between MRC and IRC
Assuming that the interference and noise are both white in the space, MRC receivers use MRC algorithm to achieve MMSE. Assuming that there is colored interference, IRC receivers use IRC algorithm to achieve MMSE. The interference rejection performance of IRC algorithm depends on the interference characteristics. Only separable spatial colored interference can be rejected by IRC algorithm. The performance of IRC algorithm depends on the accuracy of estimating the interference characteristics by the algorithm. In the following scenarios, IRC algorithm provides no advantage.
Difference between white noise and colored noise
If the interference to the antenna channels is strongly correlated to the signals to the antenna channels, the interference and signals are inseparable. In this case, IRC performance is worse than MRC performance. If the interference is white or weak, theoretically IRC algorithm is equivalent to MRC algorithm; their performance is the same. In practice, there is an error in estimating the interference characteristics. Without interference, IRC performance is slightly worse than MRC performance.
The eNodeB measures the spatial color of the interference to determine whether a user is under white interference or colored interference.
Adaptive Switchover Between MRC and IRC
For eNodeBs of V1.5, IRC is optional. If IRC is not selected, an eNodeB uses MRC. If IRC is selected, an eNodeB adaptively selects IRC or MRC depending on the current radio channel quality.
If there is separable strong colored interference, the system automatically
uses IRC algorithm.
If there is no separable strong colored interference, the system automatically rolls back to MRC algorithm.
In UL 2x2 MU-MIMO mode, the eNodeB does not support UL Interference
Rejection Combining or UL 2-Antenna Receive Diversity.
In UL 4-Antenna Receive Diversity mode, the eNodeB supports UL Interference Rejection Combining.
Multi-User MIMO (MU-MIMO)
Theoretically, the number of virtual MIMO users in the same RB cannot exceed the number of receive antennas of the eNodeB. eNodeBs of V1.5 support MU-MIM O 2x2. The following figure shows MU-MIMO 2x2.
The protocols support a maximum of MU-MIMO 4x4.
MU-MIMO Configurations
If the value of the UlSchSwitch parameter is UlVmimoSwitch, the system adaptively switches between LBFD-00202001 UL 2-Antenna Receive Diversity and LOFD-001002 UL 2x2 MU-MIMO depending on the channel quality. If the value of the UlSchSwitch parameter is not UlVmimoSwitch, the system supports LBFD-00202001 UL 2-Antenna Receive Diversity only. In UL 2x2 MU-MIMO mode, the system throughput is increased. This mode is not suitable for highspeed mobility at 120 km/h or 350 km/h and frequency hoping.
Multi-Antenna Transmit MIMO
The eNodeB supports multi-antenna transmission and the UE does not. DL 2x2 MIMO, DL 4x2 MIMO, and DL 4X4 MIMO are described. R9 defines nine multi-antenna transmission modes (TMs). The eNodeB adaptively selects one TM according to the channel condition and service requirement. Supported No.
U s e d b y F D D / T D D
Applicable Scenario
by Current eNodeB
1 2 3 4 5 6 7 8
U s e d b y T D D
Name
9
Single antenna (port 0) Open-loop transmit diversity Open-loop spatial multiplexing Closed-loop spatial multiplexing MU-MIMO Closed-loop transmit diversity Single antenna (port5) Adaptive singlestream and dualstream beamforming Adaptive singlestream, dualstream, and 4stream beamforming
Single-antenna transmission.
Yes
Suitable for cell edge where the channel condition is complex and interference is large, or high-mobility or low SNR situations.
Yes
Suitable for high UE mobility and complex reflection environment.
Yes
Suitable for good channel condition. Provides high data transmission rate.
Yes
Suitable for two orthogonal UEs. Used to increase cell capacity.
No
Suitable for cell edge, low mobility, and low SINR.
Yes
Suitable for cell edge to reject interference.
No
Suitable for cell edge, low mobility, and high SNR.
Yes
A new mode in LTE-A. Supports a maximum of eight layers. Increases data transmission rate. Suitable for low mobility and high SNR.
No
Concepts •
•
•
•
•
•
•
Port A port is a logical port and does not necessarily correspond to an antenna. There can be multiple ports. The LTE protocols support a maximum of eight physical antennas. Ports correspond to pilot formats, whereas the number of physical antennas has not direct relationship with the pilot formats. Port 0 to port 3: Ports for transmitting common pilots. Usually the number of ports for physical broadcast channels and downlink control channels is the same as that for common pilots. Port 5: A port defined in the LTE for supporting single-stream beamforming. The data of a single port can be weighted and mapped to multiple physical antennas. Port 6: A port for locating the pilot. Port 7 to port 14: Similar to port 5. Supports a maximum of 8 layers. The data of 8 ports can be weighted and mapped to 8 physical antennas. Used for dualstream beamforming. Port 15 to port 22: CSI-RS port. Maximum number of streams = Number of logical antenna ports [2 ports, 4 ports, or 8 ports]
Concepts
Pilots in the LTE system
Cell-specific reference signal (CRS): CRS is known as common pilot. CRS is used by the control channels for channel estimation and demodulation. CRS is used for demodulation of TM1 to TM6 and RSRQ measurement.
UE-specific reference signal at port 5: It is used for demodulating TM7.
DM RS at ports 7 to 14: It is used for demodulating TM8 to TM9 and is the reference signal in R9 and R10. It supports MU-MIMO and demodulation of a maximum of eight layers.
Reference signal at port 6: It is used for locating the UE.
Channel status information measurement RS (CSI-RS): It is used for measuring the channel quality indication, precoding matrix indication, and RI. CSI-RS supports measurement of eight ports.
Sounding reference signal (SRS): It is used for measuring the uplink channels and supports uplink scheduling.
Open-Loop Transmit Diversity
In open-loop transmit diversity (TM2), space-frequency block coding (SFBC) is used if the number of transmit antennas is 2; SFBC and frequency switched
transmit diversity (FSTD) are used if the number of transmit antennas is 4.
SFBC: For two-way transmit (LOFD-001001 DL 2x2 MIMO), the transmit diversity
uses SFBC, where X1 and x2 are the information to be transmitted before SFBC, * indicates conjugate operation, f 1 and f 2 are different subcarriers, and Tx 1 and Tx 2 are different transmit antennas. SFBC codes x 1 and x 2 to different antennas and subcarriers for transmission: x 1 over Tx1 f 1, x2 over Tx1 f 2, -x2* over Tx2 f 1, and x1*
over Tx2 f 2. Therefore, by transmitting copies of x 1 and x2 over different antennas and frequencies, SFBC achieves diversity gain.
SFBC+FSTD
For 4-way transmit (DL 4x2 MIMO or DL 4X4 MIMO), SFBC and FSTD are used together. In FSTD, some of the transmit antennas are selected sequentially in frequency for transmission.
The transport format of SFBC+FSTD is as follows: x 1, x2, x3, and x 4 are information
to be transmitted before coding; f 1 to f 4 are different subcarriers; Tx1 and Tx4 are different transmit antennas; * indicates conjugate operation; 0 indicates no information transmitted. In SFBC+FSTD, x 1 to x4 are coded to different antennas and subcarriers for transmission; the transmit antennas are selected. Like SFBC,
SFBC+FSTD achieves diversity gain by transmitting copies over different antennas and frequencies.
Closed-Loop Transmit Diversity
Spatial Multiplexing
Spatial multiplexing means transmission of multiple spatial data streams over different antennas in the same RB. The dimension of spatial channels is increased compared with the single-antenna technique. Therefore, spatial multiplexing increases system capacity and achieves spatial multiplexing gain. Spatial multiplexing includes two operations: layer mapping and precoding. Depending on whether the precoding matrix is obtained based on the feedback information of the UE, spatial multiplexing is classified into open-loop spatial multiplexing (TM3) and closed-loop spatial multiplexing (TM4). The following figure shows the 2x2 s patial multiplexing:
Engineering Guidelines of MIMO
The RRU models in LTE TDD that support MIMO are RRU3232
and RRU3235.
Contents Background
and Overview of the LTE MIMO
Techniques Principles
and Application of the MIMO
Techniques Principles
and Application of Beamforming
Principles and Application of Beamforming
Beamforming is a downlink multi-antenna technique. The transmitter of an eNodeB weights the data before transmission, forming narrow beams and aiming the energy at the target user, as shown in the following figure.
Beamforming does not require the UE to feed back information or use multiple antennas to transmit data. The direction of incoming wave and the path loss information are obtained by measuring the uplink received signal.
The benefits of beamforming are as follows: Increased SINR in the direction of incoming wave from the UE. Increased system capacity and coverage.
Classification of Beamforming Techniques
DOA beamforming and MIMO beamforming:
Direction of Arrival (DOA) beamforming: The eNodeB estimates the direction of arrival of the s ignal, uses the DOA information to calculate the transmit weight, and targets the major lobe of the transmit beam at the best direction. MIMO beamforming: The eNodeB uses the channel information to calculate the transmit weight, forming a beam. Open-loop beamforming and closed-loop beamforming:
Open-loop beamforming: The eNodeB uses the unlink channel information to weigh the transmit signal and does not require the UE to feed back the channel information. The protocols support single-stream (TM2) or multi-stream (TM3). Open-loop beamforming can increase the CEU throughput and coverage. Closed-loop Beamforming: The eNodeB requires the UE to feed back channel information, such as codebook to the eNodeB and uses the feedback information to weigh the transmit signal. Due to the feedback delay, closed-loop beamforming is suitable for low mobility scenarios. Due to the influence of the feedback granularity, the performance of closed-loop beamforming is slightly worse than that of open-loop beamforming. The protocols support single-stream (TM4) or multistream (TM6). The eNodeB cannot estimate the UE movement speed; the error is more than 30 km/h.
In the industry, the TDD system uses open-loop beamforming and the FDD system uses closed-loop beamforming. Huawei eNodeB supports openloop beamforming.
Classification of Beamforming (Single-Stream) Single-stream
beamforming means transmission of a single data stream in the
same OFDM resource block. It is suitable for situations of poor channel quality. Single-stream
beamforming achieves diversity gain by 1 dB by increasing the
SNR. Take
4-antenna as an example. The following figure shows single-stream
beamforming. The data stream S is weighted by w 1 to w4 and is sent to the four antenna ports for transmission.
Classification of Beamforming (Dual-Stream) Dual-stream
beamforming means transmission of two data streams in the same OFDM resource block, leading to spatial multiplexing. It is suitable for situations of good channel quality. Take 4-antenna as an example. The following figure shows dual-stream beamforming. There are two data streams S1 and S2; each antenna has two weights w i1 and wi2. S1 is weighted by four weights: w 11 to w41; S2 is weighted by another four weights w 12 to w42. The weighted streams are summed and sent to the four antenna ports for transmission.
Engineering Guidelines of Beamforming
Before configuring beamforming antennas, you need to understand the correspondence between the port No. and the co-polarization of cross-polarized antennas. The following figure shows the connection between RRU ports and antenna element of the four or eight antennas. At present, the RRU models in LTE TDD that support beamforming are RRU3232, RRU3233, and RRU3235.
4-antenna cross polarization mapping
4-antenna linear polarization mapping
4-antenna circular polarization mapping
8-antenna cross polarization mapping
Key Configuration Points in Adding a Beamforming Cell
Add an LBBP by running the ADD BRD command with Mode set to TDD_ENHANCE.
After adding the cell, run the following commands to turn on the beamforming measurement switch and algorithm switch:
MOD MEASURESWITCH: UlintfMeasSwitch=SW_BfNValidMeas1&SW_BfNRankMeas-1&SW_BfSrsMeas-1; MOD CELLALGOSWITCH: LocalCellId=0, BfAlgoSwitch=BfSwitch-1;
KPI of Beamforming Leading 4x2 Beamforming Enhanced the Capacity
Always Leading in Beamforming
3GPP R8 single-stream beamforming
1st to launch Single-stream Beamforming
+15%
3GPP R9 dual-stream beamforming 1st to support Dual-stream Beamforming
Test Result in Japan S
3GPP R10 Multi-User Beamforming
+10%
+15% Hisilcon Balong710 Chipset is the first to support dual-stream beamforming
Hisilcon Balong700 Chipset is the first to support single-stream beamforming
2011H1
2011H2
2012H1
>2Mbps
>4
TM7
91.50%
7
TM2
82.80%
61
KPI of Beamforming Relevant features Single-stream beamforming must be enabled before dual-stream beamforming. Influence on the KPI Single-stream or dual-stream beamforming has the following influence on the KPI: Cell average throughput If the single-stream and dual-stream beamforming is enabled, the signal energy received by the UE is increased, the MCS is increased at the same UE position, beamforming achieves higher cell average throughput than transmit diversity. In comparison with no beamforming, single-stream beamforming increases the cell average throughput by 15% to 25%. In comparison with single-stream beamforming, adaptive single-stream and dual-stream beamforming increases the cell average throughput by more than 10%. •
Beamforming compared with 2R diversity (UL) •
~ 30% gain in cell average throughput
•
~ 50% gain in cell edge user throughput
Beamforming compared with 2x2 MIMO (DL) •
~ 15% gain in cell average throughput
23%~90% increasing in edge ~ 40% gain in celluser edgethroughput user throughput •
Adaptive MIMO and Beamforming
With adaptive beamforming and MIMO, the UE always uses TM of high spect ral efficiency under the same channel condition. In comparison with non -adaptive MIMO or beamforming, adaptive MIMO and beamforming significantly increases average cell throughput. If beamforming is used, due to the overhead of UE-specific reference signal, the number of resource blocks is reduced. Therefore, in case of good channel quality, beamforming throughput is slightly lower than MIMO throughput. At high UE mobility (higher than 120 km/h), the eNodeB cannot track the channel change accurately according to the sounding reference signal. In this situation, beamforming is not suitable. Adaptive beamforming and MIMO (low mobility)
Adaptive beamforming and MIMO (high mobility)
Adaptive MIMO and Beamforming The BFMIMOADAPTIVESWITCH parameter is used to select adaptive beamforming or MIMO. The eNodeB selects beamforming or MIMO according to the value of the parameter, the UE movement speed, and SINR.
If the value of the parameter is NO_ADAPTIVE, the eNodeB does not support adaptive Beamforming and MIMO. If the value of the parameter is TxD_BF_ADAPTIVE, the eNodeB supports adaptive TM2 (transmit diversity) and beamforming. There are two scenarios: low UE mobility and high UE mobility. Low UE mobility: For UEs that do not support R9, single-stream beamforming (TM7) is used; for UEs that support R9, single-stream beamforming (TM7 or TM8) is used at low S INR and dual-stream beamforming (TM8) is used at high SINR. High UE mobility: Transmit diversity is used.
If the value of the parameter is MIMO_BF_ADAPTIVE, the eNodeB supports adaptive transmit diversity, dual-stream MIMO (TM3), and beamforming. There are two scenarios: low UE mobility and high UE mobility. Low UE mobility: For UEs that do not support R9, single-stream beamforming (TM7) is used at low SINR and dual-stream MIMO (TM3) is used at high SINR; for UEs that support R9, single-stream beamforming is used at low SINR and dual-stream beamforming (TM8) is used at high S INR. High UE mobility: Transmit diversity is used at low SINR and dual-stream MIMO (TM3) is used at high SINR.
Comparison Between Beamforming and Other Techniques
Though a space diversity system or intelligent antenna system
has multiple transmit or receive antennas, they can transmit only single-stream data. A MIMO system can transmit single stream or multiple streams depending on the channel quality.
MIMO requires that the number of receive antennas is not less than the number of transmit antennas. Space diversity and intelligent antennas do not have this requirement.