Transcript Physical layer network coding for next generation wireless

Physical layer network
coding for next generation
wireless broadband
Alister Burr, University of York
[email protected]
Agisilaos Papadogiannis, Chalmers University
11th March 2010
1
Outline

The challenge of next generation wireless
networks

Next generation network architectures

MIMO and MIMO cellular

Multi-user and Network MIMO

Physical layer network coding

Conclusions
11th March 2010
2
Wireless networks – some history



Began with Marconi in 1890’s
…
1st generation


2nd generation


digital, some data ~1992
3rd generation


analogue, telephony ~1980 (Japan)
CDMA, flexible services, up to 384 kbit/s ~2002
4th generation

OFDM(A), full Internet access, up to 1 Gbit/s ???
11th March 2010
3
4G

The “next generation” has been discussed ever
since 3G standards were finalised a decade ago


However starting from 2002 ITU-R has defined
the requirements for IMT-Advanced



however it was not initially clear what form the
“fourth generation” might take
which has since then been generally accepted as
the definition of 4G
Key requirement is 100 Mbit/s for high mobility
and 1 Gbit/s for low mobility
Standards currently under development:


Mobile WiMAX (IEEE 802.16e/m)
3GPP LTE-Advanced
11th March 2010
4
The dream

To provide full Internet connectivity to everyone,
anywhere


which means wirelessly
Next generation wireless research has usually
focussed on a ‘headline’ maximum data rate

but of course this will not be the rate most users
experience,

and probably is not the most important figure

In densely-populated cities a network for everyone
must provide extremely high capacity densities
11th March 2010
5
Required capacity density

Average population density in European
cities ranges from 3400 - 5400/km2

however in commercial district in working
hours it will be much higher

say 8000/km2

Suppose 10% subscribe, and 20% of those
require access at busy hour

Expected data rate 5 Mbit/s
8000/km2  10%  20%  5 Mbit/s = 800 Mbit/s/km2
11th March 2010
6
Capacity density of 1G

e.g. AMPS, U.S.A:

~400 channels in each direction

~15 km radius cells  700 km2

re-use ~1:10  0.06 channels/km2,
 equivalent

to approx 1.8 kbit/s/km2
in approx 50 MHz
11th March 2010
7
Current and 4G systems

Currently one base station serves about
1km2
4G bandwidths proposed are ~ 40 MHz
 Best available bandwidth efficiency
averages about 2 bits/s/Hz across cell
 hence capacity density is 80 Mbit/s/km2
 assumes 100% frequency re-use



We need an order of magnitude more!
10 more bandwidth unlikely to be available
11th March 2010
8
Increased density

We can also
increase
number of cells

BUT

need many
more cell sites

interference
11th March 2010
9
Cell edge problem

We may be able to increase
bandwidth efficiency
(bits/s/Hz)



use (e.g) advanced MIMO
techniques
BUT mobile close to cell
edge suffers interference
from adjacent cells
Conventionally we reduce
frequency re-use

but this reduces available
bandwidth by factor 3 or
more
11th March 2010
10
BuNGee

University of York is part of a European project
tackling these problems


Proposes:




Beyond Next Generation Mobile Broadband
(BuNGee)
new hierarchical network architecture based on
wireless backhaul
Advanced MIMO techniques for high bandwidth
efficiency
Self-organising network for
optimal spectrum use
Goal: 1 Gbit/s/km2
11th March 2010
11
Outline

The challenge of next generation wireless
networks

Next generation network architectures

MIMO and MIMO cellular

Multi-user and Network MIMO

Physical layer network coding

Conclusions
11th March 2010
12
Achieving capacity density

Will probably need a combination of the
approaches mentioned:
More spectrum
 Improved bandwidth efficiency

 especially
increased use of MIMO
Increased frequency re-use (100%)
 Reduced cell size

 requires
low base station installation cost,
 and a cost-effective backhaul network
11th March 2010
13
Wireless backhaul

Simple comparison with 4G proposals
suggests we may need ~10 BSs per km2!

We believe that the only cost-effective way
to provide this is by wireless backhaul

However must allow for spectrum used by
backhaul links

Hence must minimise backhaul load
11th March 2010
14
BuNGee architecture
11th March 2010
15
Some figures

Assume HBS serves 1 km2

Assume total 40 MHz available

20 MHz for MS-ABS (access links);

20 MHz ABS-HBS (backhaul)

Assume average 2 bits/s/Hz across cell

Then capacity per ABS = 20  2 = 40 Mbit/s

No. ABS per HBS = 1 Gbit/s / 40 Mbit/s = 25

Area served by ABS = 1 km2/25 = 40 000 m2, or
200m square
11th March 2010
16
Wireless mesh network

Since cells are very small,




mobile (MS) may be
served by more than one
ABS
MSs now served by
optimum combination of
available ABSs
Practically abolishes
HBS
concept of cells!
Overall network looks more
like a wireless mesh
network
11th March 2010
ABS
MS
ABS
MS
Relay
ABS
17
Wireless relaying

A cell can be extended
by adding fixed, or
infrastructure relays

very similar
architecture to wireless
backhaul

with relays replacing
ABSs

may allow direct
connection of MSs to
hub
11th March 2010
Relay
Relay
18
Hierarchical wireless network


A generalised framework
for network architectures
involving wireless backhaul
and/or relaying
we might allow:




more than one layer of
relays
direct connections between
nodes on the same level
MSs to connect to different
relay levels
Again, similar to mesh
network in structure
11th March 2010
19
Outline

The challenge of next generation wireless
networks

Next generation network architectures

MIMO and MIMO cellular

Multi-user and Network MIMO

Physical layer network coding

Conclusions
11th March 2010
20
MIMO link model
ENCODER



nT  




H21
HnR1
s
MIMO = Multiple Input, Multiple Output


H11
DECODER


 n
R


r
i.e. multiple antennas at each end of a link
Input and output signal can be modelled as (1  nT) and
(1  nR) vectors, s and r; noise (1  nR): n
channel modelled as a matrix H: element Hij gives propagation
between transmit antenna j and receive antenna i
r=Hs+n
11th March 2010
21
Eigendecomposition
n
1
s'
nT


H
r
n
n
r'
VH
nR
Multiply by matrices U and V at input and output of channel, where


U
s
columns of U,V are transmit and receive eigenvectors of the channel
Then U H VH = 

a diagonal matrix with the square roots of the eigenvalues of HHH on the
diagonal, and r’i = i s’i + n’i

i.e. we create a set of uncoupled channels, whose power gains are the
eigenvalues
Each eigenvector can be treated as a steering vector for antenna array
 transmit/receive beam patterns
11th March 2010
22
Beamforming model
n
s'

U
VH
r'
Another way of viewing MIMO:



each input stream corresponds to a beam from the
antenna array towards a multipath signal
can create as many such beams as there are
antennas
hence can transmit up to n = min(nT, nR) beams

provided there are enough multipaths
11th March 2010
23
MIMO capacity

Capacity and
bandwidth efficiency
approximately
multiplied by no. of
streams, n
Slope of curves
proportional to n


called multiplexing
gain
Dramatic capacity
gain!
35
25
20
15
10
5
0
11th March 2010
nT=4 nR=4
nT=2 nR=4
nT=1 nR=4
nT=4 nR=2
nT=2 nR=2
nT=4 nR=1
nT=1 nR=1
30
Capacity (bits/s/Hz)

Rayeigh capacity
0
5
10
15
SNR(dB)
20
25
30
24
MIMO in interference
Uint
d int
sint
Beamformer
H int
r
d


U
s
H
n
W
x
V
dˆ
matched
filter
MIMO cellular system also subject to interference
Beamformer applies linear weights to maximise
SNIR at output

filters signals from different directions to maximise
signal to noise-plus-interference ratio
11th March 2010
25
Capacity of MIMO cellular





4 antenna
elements
MIMO system
capacity around
3  SISO
and more than
1.5  SIMO
(smart antenna)
Beamformer
(“prewhitening”)
very important
Interference
limited
11th March 2010
26
Implications for 4G networks

MIMO can dramatically increase link
capacity

and significantly increase cellular capacity

Note that capacity is mainly affected by
n = min(nT, nR)

Still severely limited by inter-cell
interference

Can we reduce the effect of interference?
11th March 2010
27
Outline

The challenge of next generation wireless
networks

Next generation network architectures

MIMO and MIMO cellular

Multi-user and Network MIMO

Physical layer network coding

Conclusions
11th March 2010
28
Multi-user MIMO systems

It has been known since the 1960s that the
optimum means of sharing a channel
between several users may be by
simultaneous, mutually interfering
transmission


as opposed to time-division multiplexing, or
other orthogonal multiplexing
Information-theoretic approach:
Multiple Access Channel - MAC (uplink)
 Broadcast Channel - BC (downlink)

11th March 2010
29
2-user SISO MAC




R1  C1  I  X1 ; Y X 2   log 2 1  P1 N 
Rate region: set of
achievable rates of the two
R2  C2  I  X 2 ; Y X1   log 2 1  P2 N 
sources
R1  R2  Csum  I  X1 , X 2 ; Y 
C1 and C2 are capacities of
 log 2 1   P1  P2  N 
two channels without the
R2
other
“Corner points” achieved
by successive interference
cancellation
“Time sharing” sum rate
limited to dashed line

C2
time-sharing
rate
Csum
in general achievable sum
rate Csum exceeds this
11th March 2010
C1
R301
Multi-user MIMO MAC

In a multi-user MIMO multiple access channel
(uplink),


sum rate capacity limit is capacity of MIMO channel
formed by combining all Tx antennas of all users
For nU users with nT Tx antennas, average Tx
power Si and Tx time Ti (each)
CTS


where:
nU
S
i 1 i

 S ; i U1Ti  T ;
n
1 nU
  Ti CnT ,nR
T i 1
Csum  CnU nT ,nR
 T
 Si   CnT ,nR  S 
 Ti 
   C
nU
S
i 1 i
nU nT ,nR
S 
CnT ,nR  S  denotes capacity of nTnR MIMO channel
with SNR S
11th March 2010
31
Multi-user MIMO

Conventional TDMA/FDMA is equivalent to timesharing




divides “headline” rate by no. of channels
MU-MIMO allows several users to share same
time slot/channel
Users/BS can act as a single nT nU  nR MIMO
system
nT
Usually more BS
than terminal antennas

multiplexing gain no
longer limited by
no. terminal antennas
nU
BS
nR
11th March 2010
32
Sum rate (black) and time-share capacity (red), 8 users, nT=4, nR=4
40
Sum rate (black) and time-share capacity (red), 8 users, nT=2, nR=8
80
35
70
30
60
Capacity (bits/s/Hz)
Capacity (bits/s/Hz)
Symmetric and asymmetric
25
20
15
50
40
30
10
20
5
10
0


0
5
10
15
SNR (dB)
20
25
30
0
0
5
10
15
SNR (dB)
20
25
30
Modest advantage when nT = nR (symmetric links)
Large advantage when nT  nR (asymmetric links)

max. multiplexing gain becomes min(nR, nTnU)
11th March 2010
33
Broadcast channel

Broadcast channel simply means one transmitter
to many receivers




obvious example is radio/TV broadcasting, where
same message is intended for all
also applies to cellular downlink, where a different
message is intended for each receiver
R2
In latter case can define a
C2
capacity region like that for
multi-user MIMO
time-sharing
Csum
rate
Again time sharing
(TDMA/FDMA) is sub-optimal
11th March 2010
C1
R
341
Dirty paper coding (DPC)

This is an information-theoretic result which applies to a
channel subject to interference where the interference is
known at the transmitter



Achievable capacity of this channel is
log2(1 + PS/Pn)



r=s+i+n
where s is the information signal, i is the interference, and n is
noise
i.e. the same capacity as if the interference were not present
note this is not achieved by “pre-cancelling” the interference
On a broadcast downlink, the signal to one user is
interference to another, and is known at the transmitter
11th March 2010
35
Precoding in practice



Dirty paper precoding in principle
operates by selecting a codebook
(set of transmitted codewords)
depending on the interference
A more practical scheme following
the same principle is TomlinsonHarashima precoding (THP)
Use modulo function f(x), and
transmit:


where d is data, i is interference and
k is some integer
At the receiver apply the modulo
operation again:


interference is removed
some degradation due to “folding” of
noise
11th March 2010
f(x)
x
A
A  A

f  x   mod  x  , A  
2  2

s  f  d  i   d  i  kA
d   f r   f  s  i  n
 f  d  n  kA
 d  n  k A
36
Linear beamforming

Or simpler still, we can simply form beams to
each user


ensuring also that we null interference to other
users
this is a purely linear operation
11th March 2010
37
Network MIMO


So far we have assumed that signals from
other cells must be treated as interference
However it is possible for several (in
principle, all) base stations to cooperate to
transmit to a given mobile


Then there is in principle no CCI!


or to receive from that mobile
since all received signals are exploited as
signals
The entire system then operates as a multiuser MIMO system with (on the uplink)
nTnU nC transmit and nRnC receive
antennas


where nC is the number of cooperating
cells
in principle multiplexing gain approaches
min(nRnC, nTnU nC)
11th March 2010
38
Practical limitations

There are of course practical limitations to this
concept:

Where should the processing be performed?
In a hierarchical network like BuNGee, at the hub
 Distributed methods also possible, with processing
at cooperating BSs
 Computational complexity


Synchronisation


Is it feasible to keep cooperating base stations
phase synchronous? (especially for downlink)
Backhaul capacity
11th March 2010
39
Backhaul capacity requirements

On the downlink, if two BSs cooperate to
communicate with an MS, that MS’s data
should be sent to both


On the uplink, neither may be able to
decode the MS without the signal from
the other



could double backhaul requirements
hence analogue signal may need to be
transmitted over the backhaul in high
precision
may increase backhaul requirements by
several times
b1s1
b2s2
r1
r2
s1
s2
Need to ensure backhaul links are
efficiently used
11th March 2010
a1 r1 + a2 r2
40
Limitations of in-band backhauling


If we use wireless
backhaul, we must
account for bandwidth
occupied
Can we re-use the same
spectrum in backhaul and
access segments?


in-band backhauling
Duplexing restrictions of
ABSs usually prevent
same resources being
used in the two segments
in the same place
11th March 2010
41
Outline

The challenge of next generation wireless
networks

Next generation network architectures

MIMO and MIMO cellular

Multi-user and Network MIMO

Physical layer network coding

Conclusions
11th March 2010
42
Network coding

A network node applies a joint
coding function to two (or more)
incoming data streams


instead of simply switching between
them
In this simple example (the “butterfly
network”) the central node applies
the XOR function (modulo-2
addition)


then both streams can be recovered
at both output nodes
without network coding the central
link would have to have twice the
capacity to achieve this
11th March 2010
43
Two-way relay channel (2WRC)
SA
DB
b
a
ab
R
ab
SB
DA

Allows a relay to support transmissions in two
directions at once

Relay broadcasts XOR combination of two
incoming streams

Each destination can then reconstruct data
intended for it by XOR combination with the data
it transmitted
11th March 2010
44
Physical layer network coding
(PLNC)

In a wireless network, we do not have
discrete, non-interfering paths


except by using TDMA or FDMA
Signals:
are broadcast to all nodes within range
 combine additively in signal space


However it is still possible to extract a joint
information stream equivalent to XOR
combination
11th March 2010
45
PLNC for 2WRC

a
DB
ab
b
a+b
R
System operates in two
phases



SA
Phase 1: sources transmit
simultaneously
Phase 2: relay transmits
Assume both sources
transmit BPSK

{+1, -1}  {1, 0}
11th March 2010
ab
SB
DA
Phase 1
Phase 2
SA, SB
SA, SB
time
a
0
0
1
1
b
0
1
0
1
a+b
-2
0
0
+2
ab
0
1
1
0
46
For comparison
SA
DB



a
ab
Without network
coding
Network layer
network coding
Physical layer
network coding
R
b
SB
ab
DA
SA
SB
R(A)
R(B)
time
SA
SB
R(AB)
time
SA, SB
R(AB)
time
11th March 2010
47
Relay data compression
R
S




Cooperative diversity: relay provides extra diversity if S –
D link fades
Phase 1: Source transmits to relay and destination
Phase 2: Relay transmits to destination
Note signal at relay is correlated with Phase 1 signal at
destination


D
since both arise from the same data
This allows distributed compression using Slepian-Wolf
coding
 reduces
11th March 2010
the data relay must transmit
48
Slepian-Wolf coding


If two data sources are
correlated, their joint
information content is less
than the sum of their
separate content
Can exploit this to
compress the data


S1
S2
C1
C2
R1
R2
S’1
D
S’2
R2
even though encoders
are separate
White area on graph gives
rate region

range of possible
compressed rates to
allow reconstruction
11th March 2010
R1
49
PLNC in network MIMO





T1
Example:
2 terminals connected
to hub via 3 BS
B1
T2
B2
B3
B2 can use PLNC, to share
its link with the hub between
Hub
two terminals
It can use distributed compression, since
the data is correlated with that via B1 and
B3
Reduces backhaul load
11th March 2010
50
Two-way relaying in hierarchical
wireless network

Duplexing constraints can be
alleviated using 2-way relaying




allows access and backhaul
resource to be shared at each ABS
however neighbouring ABSs and
corresponding MSs may interfere
We have analysed this
considering a simple scenario
Assume 2 ABSs share resources
over 2 slots


i.e. both ABSs and MSs transmit
simultaneously on same channel
consider amplify-and-forward and
network coded decode-and-forward
relaying
11th March 2010
51
Capacity results

Plot resulting capacity
against strength of
interfering links




for small interference
DF is better
for large interference
AF capacity increases
again
both MSs can exploit
both ABSs
 exploits a form of
network MIMO
Generally 2 slots are
better than three or four
11th March 2010
52
Conclusions (1)


Next generation wireless networks will be
required to handle much larger capacity densities
MIMO techniques can greatly increase link
capacity


Multi-user MIMO can greatly increase capacities


but still seriously affected by interference in cellular
networks
especially for asymmetric systems
Network MIMO can further increase capacity


and also largely eliminate interference
however increases backhaul load
11th March 2010
53
Conclusions (2)

Physical Layer Network Coding can better
exploit backhaul capacity




by allowing backhaul links to be shared
by exploiting correlation of signals travelling by
different routes
by overcoming duplexing constraints on spectrum
sharing
All these techniques will be essential elements of
next generation networks!
11th March 2010
54
Acknowledgements

With thanks to our collaborators :




Prof. Jan Sykora, Czech Technical University,
Prague
Prof. Tad Matsumoto, JAIST, Japan
Prof. Meixia Tao, Shanghai Jiao Tong University
and funders:




European Commission, FP7 project “BuNGee”
RCUK “Sciences Bridges China” programme
COST 2100 programme
Czech Technical University
11th March 2010
55