Section 3 6 SDMA v1

Section 3 6 SDMA v1

Introduction Precoding Scheduling (user selection)

Chapter 3: Physical-layer transmission techniques

Instructor: Nguyen Le Hung Email: nlhung@dut.udn.vn; nnguyenlehung@yahoo.com Department of Electronics & Telecommunications Engineering

Danang University of Technology, University of Danang

Introduction Precoding Scheduling (user selection)

SDMA and OFDM

Multiuser transmission

Precoding classification

An example of linear precoding

Power allocation in ZF precoding

Possible research problems

Exhaustive selection

Greedy selection

Introduction Precoding Scheduling (user selection)

SDMA and OFDM

Multiuser transmission SDMA with OFDM

The integration of multi-antenna and OFDM techniques has

provided remarkable diversity and capacity gains in broadband

wireless communications

In multiuser (MU) transmissions, the use of multiantenna array at the base station (BS) enables simultaneous transmission of multiple data streams to multiple users by exploiting spatial separations

among users

IFFT SU-MIMO precoder ABS/eNB

IFFT MU-MIMO precoder

Introduction Precoding Scheduling (user selection)

SDMA and OFDM

Multiuser transmission

A simple example of multiuser (MU) transmission

1 , 1

h

2 , 1

h

M

h1 ,

Base Station

1

s

Modulation

Coded bits

of user 1

2

s

Modulation

Coded bits

of user 2

1 , 2

h

2 , 2

h

M

h2,

Antenna 1

Antenna M

De-mod

Channel estimator

User 2

De-mod

Channel estimator

User 1

1

y

2

y

𝑦1=𝑠1

𝑀

∑

𝑚=1

ℎ1,𝑚+𝑠2

𝑀

∑

𝑚=1

𝑀

∑

𝑚=1

𝑀

∑

𝑚=1

ℎ2,𝑚+𝑧2

Introduction Precoding Scheduling (user selection)

An example of linear precoding Power allocation in ZF precoding Possible research problems Precoding classification

In the so-called space division multiple access (SDMA), multiuser diversity is the primary factor that increases significantly the system sum-rate (throughput)

As a result, an appropriate multiuser encoding technique (at the BS)

is indispensable to attain the considerable sum-rate gain in SDMA

It is well-known that dirty paper coding (DPC) is an optimal

multiuser encoding strategy that achieves the capacity limit of MU broadcast (BC) channels but at the cost of extremely high

computation burden as the number of users is large

Recent studies have introduced several suboptimal multiuser

encoding techniques with lower complexity (relative to DPC) that can be categorized into:

nonlinear precoding such as: vector perturbation, Tomlinson

Harashima techniques

linear precoding such as: minimum mean squared error (MMSE),

zero-forcing

Introduction Precoding Scheduling (user selection)

An example of linear precoding Power allocation in ZF precoding Possible research problems Multiuser transmission techniques

Broadband communications LTE (4G) system

Broadband communications

(high data rate and reliability)

Diversity

Multipath channel Modeling

CSI feedback Analog Digital

Vector quantization

g

Quasi-staticTime-variant

LBG Grassmannian Random

Scheduling Precoding

Exhaustive

search

Greed or iterative search

Linear methods

Non-linear methods

Codebook-based ones

Random

user selection

VP

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding Possible research problems

An example of linear precoding

1 , 1

h

2 , 1

h

M

h1,

Base Station

Feedback link of channel state information (CSI)

1

s

X

X

X

1 , 1

w

Modulation

Coded bits

of user 1

2 , 1

w

M

w1,

2

s

X

X

X

1 , 2

w

Modulation

Coded bits

of user 2

2 , 2

w

M

w2 ,

1 , 2

h

2 , 2

h

M

h2,

Antenna 1

Antenna M

De-mod

Channel estimator

User 2

De-mod

Channel estimator

User 1

1

y

2

y

𝑦1 = 𝑠1

𝑀

∑

𝑚=1

𝑀

∑ 𝑚=1 𝑤2,𝑚 ℎ1,𝑚 + 𝑧1, and 𝑦2 = 𝑠2

𝑀

∑ 𝑚=1 𝑤2,𝑚 ℎ2,𝑚+𝑠1

𝑀

∑ 𝑚=1 𝑤1,𝑚ℎ2,𝑚 +𝑧2

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding Possible research problems Inter-user interference

The received signals at user-𝑢 can be determined by

𝑀

∑

𝑚=1

𝑤𝑢,𝑚ℎ𝑢,𝑚+𝑠𝑢′

𝑀

∑

𝑚=1

𝑤𝑢′ ,𝑚ℎ𝑢,𝑚+ 𝑧𝑢, 𝑢, 𝑢′

∈ {1, 2}, (1)

∑𝑀

that would significantly degrade the performance of the system

that satisfy the following condition

𝑀

∑

𝑚=1

𝑢

The above technique is called as zero-forcing (ZF) precoding

easily solved by expressing received signals in a vector form

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding Possible research problems Zero forcing (ZF) precoding formulation

In the presence of two users, the previous equations become

[

𝑦1

𝑦2

]

= [

]

⎡

⎢

⎤

⎥ [

𝑠1

𝑠2

] + [

𝑧1

𝑧2

]

In the presence of 𝑈 users, the received signal can be expressed by:

where y =

⎡

⎢

⎤

⎡

⎢

⎤

⎡

⎢

⎤

⎥

⎡

⎢

⎤

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding Possible research problems Zero-forcing precoding formulation (cont.)

To eliminate inter-user interference, precoding matrix W can be

determined by

so that

With precoding, the received signal can be written by

vector form at 𝑀 antennas in the base station

𝔼

∑

𝑚=1

∣𝑥𝑚∣2

]

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding

Possible research problems Power allocation in ZF precoding

The power constraint (7) is equivalent to

𝑈

∑

𝑢=1

𝑃𝑢𝑠𝑢

After ZF precoding, the received signals at 𝑈 users are given by

⎡

⎢

𝑦1

𝑦𝑈

⎤

⎡

⎢

√

𝑃1𝑠1

√

𝑃𝑈𝑠𝑈

⎤

⎡

⎢

𝑧1

𝑧𝑈

⎤

Hence, the resultant sum-rate of the multiuser system is

𝑃𝑢: ∑ 𝑈 𝑢=1 𝜆𝑢𝑃𝑢≤𝑃 max

𝑈

∑

𝑢=1

Introduction Precoding Scheduling (user selection)

An example of linear precoding

Power allocation in ZF precoding

Possible research problems Power allocation in ZF precoding (cont.)

easily determined by the following waterfilling process

satisfy

𝑈

∑

𝑢=1

process attempts to eliminate the inter-user interference and

maximize the system sum-rate

The problem of how to perform user selection (finding the set

system sum-rate will be addressed in the next section

Introduction Precoding Scheduling (user selection)

An example of linear precoding Power allocation in ZF precoding

Possible research problems

Precoding in LTE downlink transmissions

Data bits

of user 1

Channel

encoder Interleaver

Layer mapper

MQAM mapper MQAM mapper Precoding

OFDMA modulator

OFDMA modulator

Precoding matrix generator

Recovered data bits Channel decoder

Channel Estimator

OFDMA Demodulator

BER evaluator

of user 1

OFDMA Demodulator

Channel State Information (CSI)

MIMO demapper

Limited feedback link

User 1

Base Station (BS)

Data bits

of user N

Channel

encoder Interleaver mapper Layer

MQAM mapper MQAM mapper

W

X

Y = W *X

BER evaluator

of user N

Multipath fading channel

User N

Introduction Precoding Scheduling (user selection)

Exhaustive selection

Greedy selection Exhaustive selection

Given a precoding technique, scheduling (user selection) is to find a set of users among all active users to maximize the system sum-rate Obviously, the simple optimal method for user selection is exhaustive search but its complexity is impractically high as the number of users

is large

Introduction Precoding Scheduling (user selection)

Exhaustive selection

Greedy selection

Greedy selection

Greedy user selection algorithm

indices

𝜂 = 0 stands for the number of selected users, initially set to zero

maximizes the resulting sum-rate of the system called 𝐶max

𝐶𝜂= 𝐶max

Ω𝜂= Ω𝜂−1∪

{𝑢} (select one more user)

Θ𝜂= Θ𝜂−1∖{𝑢} (ignore user-𝑢 in later consideration)

Go to Step 2

vectors based on the composite channel matrix of selected users