Interferometric Noise Penalty in SMF
Links - Experimental Results and
Comparison with Theory
Petar Pepeljugoski
George Sefler
IBM
Outline
Measurement Setup
Comparison of measured and calculated
penalties for 3dB and 3.7 dB extinction
ratio
Conclusion
Proposed solution
IBM
Experimental Setup
setup closely follows analytic model
worst case IN controlled with polatization controller
50/50
splitter
50/50
splitter
Polarization
Controller
Attenuator 2
TX
RX
Attenuator 1
Pattern
Generator
BERT
IBM
Calculation of IN
Improved calculation follows formalism adopted by P. Legg
(JLT. vol. 14, no. 9, pp. 1943-1954)
expressions for conditional error probabilities
Takes into account ER, ISI
IBM
Interferometric Noise Penalty - 3dB ER
IN penalty 3.2 dB with maximum polarization alignment - leaves
ample of positive margin for 24 dB RL and shortest links
IN penalty 0.6 dB for longest links with no connector loss - leaves
2dB margin
0
∞
34 dB
32 dB
30dB
28 dB
26 dB
24 dB
-2
-4
BER
-6
-8
-10
-12
-14
-9
-8
-7
-6
-5
Attenuation
-4
-3
-2
IBM
Comparison of Measured and Calculated
Interferometric Noise Penalty - 3dB ER
0 or 2 dB link loss due to connectors
shortest links have only 3.2 dB (0 dB loss) ~1.5 dB (2dB loss) worst case penalty
(polarization aligned)
excellent agreement between measurements and analytical model
Link with 2 dB Connector Loss
8
5
4
(10
-2) Ava
dB ilab
- (0 le L
in
.5
dB k Bu
/km dg
)L- et
Lin
kP
en
.
3
2
Measurement
Calculation
9
Margin
~0.24 dB
1
8
RLRx = RLTx = -12 dB
Worst-case Polarization
ER = 3 dB
7
6
(1 0
-0) Availa
dB
b
- (0 le Lin
.5 d k B
B/k udg
m)L et
- Li
nk
Margin
6
Margin
Worst-case IN Penalty (dB)
RLRx = RLTx = -12 dB
Worst-case Polarization
ER = 3 dB
Worst-case IN Penalty (dB)
Measurement
Calculation
7
0
Link with 0 dB Connector Loss
10
5
4
3
Pe
n.
2
1
0
2
4
6
Link Length (km)
8
10
0
0
2
4
6
8
10
Link Length (km)
IBM
Interferometric Noise Penalty - 3.7 dB ER
IN penalty 2.5 dB with maximum polarization alignment - leaves
even more positive margin
No IN penalty with minimum polarization alignment
-2
0
10
10
worst case
-2
10
34 dB
32 dB
30dB
28 dB
26 dB
24 dB
22 dB
-4
10
-6
-4
10
14 dB
16 dB
18 dB
20 dB
22 dB
24 dB
26 dB
28 dB
30 dB
32 dB
34 dB
-6
10
BER
10
BER
best case
∞
-8
10
-8
10
∞
-10
10
-10
10
-12
10
-12
10
-14
-14
10
-9
-8
-7
-6
-5
Attenuation
-4
-3
-2
10
-9
-8.5
-8
-7.5
-7
-6.5
-6
-5.5
-5
Attenuation [dB]
4 dB ER will be even more robust!
IBM
Comparison of Measured and Calculated
Interferometric Noise Penalty 3.7 dB ER
0 or 2 dB link loss due to connectors
shortest links have 2.5 (0 dB loss) or 1.1 dB (2dB loss) worst case penalty
(polarization aligned)
very good agreement with analytical model
Link with 2 dB Connector Loss
8
3
2
1
Margin
~0.46 dB
en
.
8
RLRx = RLTx = -12 dB
Worst-case Polarization
ER = 3.7 dB
7
(10
-0) Availa
dB
b
- (0 le Lin
.5 d k B
B/k udg
m)L et
-L
ink
6
Margin
4
(10
-2) Ava
dB ilab
- (0 le L
in
.5
dB k Bu
/km dg
)L- et
Lin
kP
Measurement
Calculation
9
5
Margin
Worst-case IN Penalty (dB)
6
RLRx = RLTx = -12 dB
Worst-case Polarization
ER = 3.7 dB
Worst-case IN Penalty (dB)
Measurement
Calculation
7
0
Link with 0 dB Connector Loss
10
5
4
3
Pe
n.
2
1
0
2
4
6
Link Length (km)
8
10
0
0
2
4
6
8
10
Link Length (km)
IBM
Dependence of BER on Time
BER drifts from maximum polarization
initial link settings for BER = 0.5e-9
-9
10
-10
10
-11
BER
10
-12
10
-13
10
-14
10
0
2
4
6
8
10
Time [Hours]
12
14
16
IBM
Conclusions
Interferometric noise penalty is real and
can be large
Measurements and calculations in excellent
agreement
1300 nm serial links still have positive
margin after allocating IN penalty
we have 0.4 dB margin left
IBM
Proposed Solution
Include interferometric noise penalty in the link
budget
Set minimum Extinction Ratio to 4 dB
Proposed solution does not require any component
redesign, does not cause pain to anybody
IBM