01-01-2013, 11:21 AM
MIMO for LTE
SIMO and MISO
SIMO
A single transmit antenna and NRreceive antennas
Receive Spatial (Antenna) Diversity
MISO
NTtransmit antennas and a single receive antenna
Transmit Spatial (Antenna) Diversity
MIMO
NTtransmit antennas and NRreceive antennas
Diversity gain ~ NTNRCapacity ~ min(NT, NR)
Performance-oriented MIMOData-rate-oriented MIMO
Receive Antenna Diversity
Selection Diversity
Antenna element with the highest SNR (or other metric) is selected
Greatest SNR improvement when
Desired signal subject to independent (uncorrelated) fading and signal receiver with the same average power at each element
Background noise is AWGN, equal power and uncorrelated across elements
Equal Gain Combining (EGC)
Co-phasing and summing the decision statistics at each element
LOS signal (no fading): equivalent to finding a beamformingweight vector w
Rayleigh fading signals: no longer optimum, but tends to achieve more diversity than selection diversity
Maximal Ratio Combining (MRC)
Each sensor output is co-phased and weighted by its SNR before combining
MRC require design in receiver circuitry to achieve correct weighting factors
Optimalin the sense that provides best statistical reduction of fading of any linear combiner
Shows little improvement over EGC for Gaussian noise channel
MIMO Spatial Multiplexing (SM)
Multiple Input Multiple Output (MIMO)
Multiple antennas at both transmitter and receiver
MIMO uses multipath to advantageto “multiply data rate”
Transmits different data along different paths (simplified view)
MxN MIMO can multiply data rate by M or N (whichever is less) if there is enough multipath.
Best in urban high-multipath environment (and indoors)
Less effective in suburban and rural low-multipath environments
MU-MIMO in theory
Unlike SU-MIMO which is the point-to-point multiple-antenna channels, MU-MIMO is the multiple-antenna broadcast channels.
The multiuser capacity of multiple-antenna broadcast channels depends heavily on whether the transmitter knows the channel coefficients to each user.
For example, in a Gaussian broadcast channel with Mtransmit antennas and nsingle-antenna users, the sum rate capacity scales like Mlog log nfor large nif perfect channel state information (CSI) is available at the transmitter, yet only logarithmically with Mif it is not.
The MU-MIMO with perfect CSI at the transmitter is well known capacity achieving dirty paper coding (DPC) [1].
On the other hand, other techniques such as random beamforming (RBF) [2] and opportunistic beamforming [3] have been studied as the sub-optimum scheme where there is only partial or very little side information available at the transmitter.
SIMO and MISO
SIMO
A single transmit antenna and NRreceive antennas
Receive Spatial (Antenna) Diversity
MISO
NTtransmit antennas and a single receive antenna
Transmit Spatial (Antenna) Diversity
MIMO
NTtransmit antennas and NRreceive antennas
Diversity gain ~ NTNRCapacity ~ min(NT, NR)
Performance-oriented MIMOData-rate-oriented MIMO
Receive Antenna Diversity
Selection Diversity
Antenna element with the highest SNR (or other metric) is selected
Greatest SNR improvement when
Desired signal subject to independent (uncorrelated) fading and signal receiver with the same average power at each element
Background noise is AWGN, equal power and uncorrelated across elements
Equal Gain Combining (EGC)
Co-phasing and summing the decision statistics at each element
LOS signal (no fading): equivalent to finding a beamformingweight vector w
Rayleigh fading signals: no longer optimum, but tends to achieve more diversity than selection diversity
Maximal Ratio Combining (MRC)
Each sensor output is co-phased and weighted by its SNR before combining
MRC require design in receiver circuitry to achieve correct weighting factors
Optimalin the sense that provides best statistical reduction of fading of any linear combiner
Shows little improvement over EGC for Gaussian noise channel
MIMO Spatial Multiplexing (SM)
Multiple Input Multiple Output (MIMO)
Multiple antennas at both transmitter and receiver
MIMO uses multipath to advantageto “multiply data rate”
Transmits different data along different paths (simplified view)
MxN MIMO can multiply data rate by M or N (whichever is less) if there is enough multipath.
Best in urban high-multipath environment (and indoors)
Less effective in suburban and rural low-multipath environments
MU-MIMO in theory
Unlike SU-MIMO which is the point-to-point multiple-antenna channels, MU-MIMO is the multiple-antenna broadcast channels.
The multiuser capacity of multiple-antenna broadcast channels depends heavily on whether the transmitter knows the channel coefficients to each user.
For example, in a Gaussian broadcast channel with Mtransmit antennas and nsingle-antenna users, the sum rate capacity scales like Mlog log nfor large nif perfect channel state information (CSI) is available at the transmitter, yet only logarithmically with Mif it is not.
The MU-MIMO with perfect CSI at the transmitter is well known capacity achieving dirty paper coding (DPC) [1].
On the other hand, other techniques such as random beamforming (RBF) [2] and opportunistic beamforming [3] have been studied as the sub-optimum scheme where there is only partial or very little side information available at the transmitter.