Professional Documents
Culture Documents
al
ainen
Introduction
Taneli Riihonen
Motivation
Focus:
Fixed infrastructure-based relay node
. Amplify and forward (AF) protocol
. Frequency-domain processing
.
Relay node
Base station
Taneli Riihonen
User
System Model
Taneli Riihonen
xt ( )
x
t ( )
DFT
PA
Yr [n]
CP removal
htr ( )
t
CP insertion
inverse DFT
Xt [n]
1
2}
E{|V
[n]|
t
Pt
Taneli Riihonen
x
R ( )
DFT
CP removal
PA
hRD ( )
xR ( )
YD [n]
R
CP insertion
inverse DFT
x
S ( )
XR [n]
amplification
xS ( )
CP removal
PA
hSR ( )
DFT
YR [n]
S
CP insertion
inverse DFT
XS [n]
PR
2
2
2
(KS +S )PS |HSR [n]|2 +R
Amplification by [n] =
KS2 SR [n]
S2 SR [n]+1
KS2 SR [n]
S2 SR [n]+1
2
KR
RD [n]
2
R RD [n]+1
2
KR
RD [n]
2
R
RD [n]+1
+1
Taneli Riihonen
where
SR [n] =
RD [n] =
PS |HSR [n]|2
2
R
PR |HRD [n]|2
2
D
Multipath Channels
The source and the relay are fixed nodes, and the destination is mobile
Relay node
Base station
User
BEP Analysis
Taneli Riihonen
BEP Derivation
[n]RD [n]
in which
Reformulate the end-to-end SNR as [n] = 2 [n]
RD [n]+1
[n] =
[n] =
2
SR [n]
KS2 KR
1
2 (KS2 + S2 ) SR [n] + 1
2
2
2 2
2 2
2 2
+ R
KS R + S KR + S R SR [n] + KR
(KS2 + S2 ) SR [n] + 1
X
(1)k k + 32
1
1
1
2 2 [n]
RD [n]
k!(k
+
2)
k=0
k+ 21
[n]
1
3
U k + , 2,
[n]
2
[n]
RD [n]
Taneli Riihonen
BEP Bounds
Linear PAs by substituting KS = KR = 1 and S2 = R2 = 0
Asymptotic lower bounds
e [n] P SR [n] with
. First hop limited by distortion: P
SR
SR
2
1 KS2 KR
,
[n] =
lim [n] =
2
2
2 KS + S
SR [n]
2
2
2
+ S2 R
+ S2 KR
KS2 R
[n] =
lim [n] =
KS2 + S2
SR [n]
r
[n]
[n]
1
2
1
erfc
2
1
2
RD [n]
RD [n]+2
SR [n]
2
Discussion
Taneli Riihonen
At
|xt ( )|,
t
At Pt ,
|xt ( )| t Pt
|xt ( )| > t Pt
t2
At
t
1 exp t2 +
erfc (t )
t
2
A2t
2
2
1
exp
K
t
t
t2
Taneli Riihonen
1.5
5
1.52
5
1.5
5
5
1.57
) = arg min
(S , R
(S ,R ) Pe [n]
subject to S > 0, R > 0
1.
1.6
1.625
1.65
25
1.6
75
62
2
R
[dB]
1.5
67
1.6
1.
1.69
1.6
67
Taneli Riihonen
1.69
1.
25
1.65
2
S2 [dB]
1.6
1.5
1.5
1.
62
75
1.
S =S
1.675
65
1.
1.65
KS2 SR [n]
= arg min 2
, S > 0
S
[n]
+
1
S SR
R
= arg min Pe [n]
, R > 0
1.
4 1.6
Can be decomposed as
1.
5
52
1. 5
55
1.575
1.
1.55
Fig. 2. Contour plot for log10 Pe [n] in terms of the input back-off
factors when
SR [n] = 15dB,
RD [n] = 20dB, and AS = AR = 1. The
minimal bit error probability Pe [n] = 2.0 102 is reached with S2 =
2 = 2.18dB (marked with ).
2.85dB and R
RD [n] [dB]
5
3
2
15
20
SR [n] [dB]
(a) S2 [dB]
25
30
3
2
15
1
10
1
1
2
5 3
4
5
6
0
0
10
2
3
4
5
6
15
RD [n] [dB]
1
2
2
1
10
5
4
3
3
5
4
6
5
5
RD [n] [dB]
4
0
20
20
1.75
1.5
15
1.25
1.25
10
75
2.25
0.
2.5
15
10
20
25
2.75
2.2
1.75
1.5
25
1.25
5
0.75
25
30
30
20
SR [n] [dB]
2 [dB]
(b) R
25
3
4
5
6
30
0.75
0.5
10
15
20
25
30
SR [n] [dB]
`
Fig. 3. Contour plots for the optimized input back-off factors and the minimal bit error probability when AS = AR = 1. The SNR pair (
SR [n],
RD [n])
considered in Fig. 2 is marked with .
Taneli Riihonen
101
Pe [n]
102
2
S2 = 6dB, R
= 0.5dB
2
S2 = 2dB, R
= 3dB
10
2
2
S2 = S2 , R
= R
2
2
S linear, R
= R
S2 = S2 , R linear
Both PAs linear
104
10
15
20
25
30
[dB]
Fig. 4. Average bit error probability when
SR [n] =
,
RD [n] =
+ 5dB
and AS = AR = 1.
2
S2 = 6dB, R
= 0.5dB
2
S2 = 2dB, R
= 3dB
2
2
S2 = S2 , R
= R
2
2
S linear, R
= R
101
101
S2 = S2 , R linear
Pe [n]
Pe [n]
102
102
10
15
SR [n] [dB]
20
25
30
Fig. 5. Average bit error probability in terms of the first hop SNR when the
second hop SNR
RD [n] = 20dB and AS = AR = 1.
10
15
RD [n] [dB]
20
25
30
Fig. 6. Average bit error probability in terms of the second hop SNR when the
first hop SNR
SR [n] = 15dB (hence S2 = 2.85dB) and AS = AR = 1.
Taneli Riihonen
Conclusion
Taneli Riihonen
Conclusion
Taneli Riihonen
Thank you!
Taneli Riihonen