Analog Links

Girish M
Asst. Professor
 Content
• Overview of Analog Links
• Carrier-to-Noise Ratio (CNR)
– Carrier Power
– Photo-detector and Pre-amplifier Noises
– Relative Intensity Noise (RIN)
– Reflection Effects on RIN
– Limiting Conditions on CNR
• Multichannel Fiber Transmissions
– Multichannel Amplitude Modulation
– Multichannel Frequency Modulation
– Subcarrier Multiplexing
 Analog Communication Links

(Amplifier
Spontaneous
Emission)

Analog (RF) links are used where converting to digital

signals is undesirable, including
• Analog TV and audio services
• Cable modem services
• Microwave-multiplexed signals (e.g., satellite base
stations)
• Radar signal processing
 Overview of Analog Links
• A bias point on the source is set approximately at the
midpoint of the linear output region. The analog signal can
then be sent with direct intensity modulation technique.
• Other modulation techniques include AM, FM and PM, which
first convert baseband signal onto an electrical subcarrier
prior to intensity modulation.
• Signal impairments in optical source : harmonic distortions,
inter-modulation (IM) products, RIN in the laser, and laser
clipping.
• The fiber should have a flat amplitude and group-delay response
within the passband required to send the signal free of linear
distortion.
• Since modal-distortion-limited bandwidth is difficult to equalize,
it is best to choose a single-mode fiber.
 Carrier-to-Noise Ratio
• The ratio of rms carrier power to rms noise power at
the input of the RF receiver is known as the
carrier- to-noise ratio (CNR).
• For the FSK modulation scheme, BERs of 10-9 and
10-15 translate into CNR values of 36 (15.6 dB) and
64 (18.0 dB), respectively.
• Using AM for studio-quality TV signal requires a
CNR of 56 dB, since the need for bandwidth
efficiency leads to a high SNR. FM only needs CNR
values of 15-18 dB.
 Carrier-to-Noise Ratio
• If CNRi represents the carrier-to-noise ratio related to
a particular signal contaminant, then for N signal-impairment
factors the total CNR is given by
N
1 1
CNR = ∑i=1 CNR
• For single information channel,
i the important signal
impairments include laser intensity noise fluctuations,
laser clipping, photo-detector noise, and optical-amplifier
noise.

• For multiple message channels, the harmonic and inter-

modulation distortions arise.

• The three dominant factors that cause signal impairments in

a fiber link are shot noise, optical-amplifier noise, and laser
clipping.
 Carrier Power
• if the time-varying analog drive signal is s(t), then the envelope of
the output optical power P(t) has the form
P(t) = Pt [1+

power at ms(t) ]
where Pt is the optical output
the bias current level.
•The modulation index m is given
by m = Ppeak / Pt , where Ppeak and
Pt are defined in the right figure.
Typical values of m for analog
applications range from 0.25 to
0.50.
• For a sinusoidal received signal,
C at the output of receiver is

C=
1
(mRMP ) 2
[A
2

] 2
R : Responsivity, M : gain, P :
 Photodetector and Preamplifier Noises
• For the photodiode noise,
iN2 = σN2 ≈ 2q(I p + I D )M 2 F (M )Be ; Ip =
RP current, M is the photodiode gain
• Here, ID is the detector bulk dark
with F(M) being its associated noise figure, and Be is the receiver
bandwidth.
• The CNR for the photodetector only is det
C
= σ2
CNR N

• For the preamplifier 4k BT

i 2 noise,
T= σ2 T Be
Req
= Ft
where Req is the equivalent resistance of the photodetector load and the
preamplifier, and Ft is the noise factor of the preamplifier.
• Then, the CNR for the preamplifier only is
= C
CNR preamp o
2 T
 Relative Intensity Noise (RIN)
• The noise resulting from the random intensity fluctuations (due to
temperature variations or spontaneous emission) is called relative
intensity noise (RIN), which may be defined in terms of the
mean- square intensity variations.
• The resultant mean-square noise current is given by
2
i RIN = σRIN = RIN
2
( RP ) 2
e
where RIN is defined by Bthe noise-to-signal power ratio:
(∆PL )
RIN = 2 [dB/Hz]
PL2
Here, <(PL)2>: mean-square intensity fluctuation, PL : average light
intensity
• Then, the total CNR is given by
CNR =
(mRMP ) / 2

RIN(RP ) Be + 2q(I p +2I D )M 2 F (M )Be + (4kBT / R )Be F

2

eq
 Relative Intensity Noise (RIN) (2)
Example 9-1:
• The right figure shows RIN for
two buried-heterostructure lasers .
The noise level was measured at
100 MHz.
• For injection currents
sufficiently above threshold (i.e.,
for IB/Ith > 1.2), the RIN of these
index-guided lasers lies between
-140 and -150 dB/Hz.
 Relative Intensity Noise (RIN) (3)
Example 9-2: The figure below shows the RIN of an InGaAsP buried-
heterostructure laser as a function of modulation frequency at several different
bias levels. The RIN is essentially independent of frequency below several
hundred MHz, and it peaks at the resonant frequency. At a bias level of 60-mA,
which gives a 5-mW output, the RIN is typically < -135 dB/Hz for modulation
frequencies up to 8-GHz. For received optical signal levels of -13 dBm (50-W)
or less, the RIN of buried-heterostructure InGaAsP lasers lies sufficiently below
the noise level of a 50- amplifier with a 3-dB noise figure.
 Reflection Effects on RIN
Back-reflected signals can increase the RIN by 10-20 dB as shown in the
figure below. These curves show the increase in RIN for bias points
ranging from 1.24 to 1.62 times the threshold-current level. The
feedback power ratio here is the amount of optical power reflected back
into the laser relative to the light output from the source. The dashed
line shows that at 1.33Ith the
feedback ratio
must be less than -
60 dB in order to
maintain an RIN of
less than -140
dB/Hz.
 Limiting Conditions on CNR
The dominant terms of the noise differ depending on the
power level at the receiver as follows:
• Low level : the preamplifier circuit noise
• Intermediate level : the quantum noise
• High level : the RIN
Example 9-3: Consider a link with a laser transmitter and a
PIN receiver having the characteristics given below.
 Limiting Conditions on CNR (2)
The figure below show a plot of C/N as a function of the optical power
level at the receiver.
Observations:
•At high received powers the source noise dominates to give a
constant
C/N.
•At intermediate levels, the
quantum noise is the main
contributor, with a 1-dB drop in
C/N for every 1-dB decrease in
received optical power.
•For low light levels, the
thermal noise of the receiver
is the limiting noise term,
yielding a 2-dB rolloff in C/N
for each 1-dB drop in received
optical power.
 Multichannel Fiber Transmissions
• In broadband analog applications, such as CATV super-
trunks, one can employ multiplexing technique in which a
number of baseband signals are superimposed on a set of N
subcarriers that have different frequencies f1, f2, ... , fN.
• The modulated subcarriers are combined through FDM to
form a composite signal that directly modulates a single
optical source.
• Methods for achieving this include VSB-AM, FM, and
SCM.
• AM is simple and cost-effective in that it is compatible with
the equipment interfaces of a large number of CATV
customers.
 Multichannel Fiber Transmissions (2)
• Although FM requires a larger bandwidth than AM, it provides a
higher SNR and is less sensitive to source nonlinearities.
• Microwave SCM operates at higher frequencies than AM or FM
and is an interesting approach for broadband distribution of
both analog and digital signals.
• The fiber-optic-based TV networks operate in a frequency
range from 50 to 88 MHz and from 120 to 550 MHz. The band
from 88 to 120 MHz is reserved for FM radio broadcast.
• The CATV networks can deliver over 80 AM-VSB video
channels, each having a noise bandwidth of 4 MHz
within a channel bandwidth of 6 MHz, with SNRs
exceeding 47 dB.
 Multichannel Amplitude Modulation
• Used in CATV networks, which can deliver up to 80 AM-VSB video
channels.
• The figure below depicts the technique for combining N independent
messages. An i-th channel information-bearing signal amplitude-
modulates a carrier wave that has a frequency fi, i = 1, 2, ..., N.
• An RF power combiner sums these N amplitude-modulated carriers to
yield a composite FDM signal which intensity-modulates a laser diode.
• Following the optical receiver, a bank of parallel bandpass filters
separates the combined carriers back into individual channels.

• The individual message signals are recovered from the carriers

by standard RF techniques.
 Multichannel Amplitude Modulation (2)
For N channels the optical modulation index m is related to the
per- channel modulation index mi by
1/
 N
2 2 
m= ∑
 mi
i=1
 modulation
If each channel   index mi has the same value mc, then
m = mcN 1/ 2
As a result, when N signals are frequency-multiplexed and used to
modulate a single optical source, the CNR of a single channel is
degraded by 10logN.
Multichannel Amplitude Modulation (3)
• When multiple carrier frequencies pass through a nonlinear
device such as a laser diode, inter-modulation (IM) products
other than the original frequencies can be produced.
• Among the IM products, only the 2nd-order and 3rd-order
terms are considered, since higher-order products tend to be
significantly smaller.
• The 3rd-order IM distortion products at frequencies fi + fj - fk
(which are known as triple-beat IM products) and 2fi - fj
(which are known as two-tone IM products) are the most
dominant.
 Multichannel Amplitude Modulation (4)
• A 50-channel CATV network operating over 55.25-373.25 MHz has
39 second-order IM products at 54.0 MHz and 786 third-order IM
tones at 229.25 MHz.

• The amplitudes of the triple-beat products are 3 dB higher than

the two-tone third-order IM products.

• Since there are N(N-1)(N-2)/2 triple-beat terms compared with N(N-1)

two-tone terms, the triple-beat products tend to be the major source
of IM noise.

• If a signal passband contains a large number of equally spaced carriers,

several IM terms will exist at or near the same frequency. This so-
called beat stacking is additive on a power basis.
Multichannel Amplitude Modulation (5)
• For N equally spaced equal-amplitude carriers, the
number of third-order IM products that fall right on the
r-th carrier is given by
D1,2 = 1  
N − 2 −1 1− (−1)N (−1 r 
2 2 ) 
[
for two-tone terms ]
 of the type 2fi – fj ,
• and by
D1,1, = r (N − r +1) +1 (N − 32) − 5 −1 1− (−1)N (−1 N+
r


2 4 2 ) 
[ triple-beat termsof the type fi + ]fj – fk .
1

for
Multichannel Amplitude Modulation (6)

D1,2

D1,1,1
Multichannel Amplitude Modulation (7)
• The results of beat stacking are commonly referred to as
composite second order (CSO) and composite triple beat
(CTB), and are used to describe the performance of
multichannel AM links:
peak carrier power
CSO =
peak power in composite 2nd-order IM tone
peak carrier power
CTB =
peak power in composite 3rd-order IM tone
Multichannel Amplitude Modulation (8)

Example 9-4:
• Figures 9-8 and 9-9 show the predicted
relative 2nd-order and 3rd-order IM
performance, respectively, for 60 CATV
channels in the frequency range 50-450
MHz.
• The effect of CSO is most significant at the
passband edges, whereas CTB contributions
are most critical at the center of the band.
 Multichannel Amplitude Modulation (9)
Predicted relative CSO and CTB performance for 60 amplitude-
modulated CATV channels in 50-450 MHz

CTB

CSO
 Multichannel Frequency Modulation
• The use of AM-VSB signals has a C/N or S/N requirement of > 40-
dB for each AM channel, which places very stringent requirement on
laser and receiver linearity.
• FM scheme requires a wider bandwidth (30-MHz versus 4-MHz for
AM), but yields a SNR improvement over the CNR.
• The S/N at the output of an FM detector is much larger than the C/N
at the input of the detector. The improvement is given by
3 B  ∆f pp 
2

  =  
S C 
e
N
 out +10 log   fν   +
  
 N  2 
 w
where B is the required bandwidth,
ni Dff pp is the peak-to-peak frequency
v
ν
deviation of the modulator, f is the highest videofrequency,
 and w is a
weighting factor used to account for the nonuniform response of the
eye pattern to white noise in the video bandwidth.
• The total S/N improvement is generally in the range of 36 ~ 44 dB.
 Multichannel Frequency Modulation (2)
Example 9-5: RIN vs OMI.
Assume that
•RIN noise dominates
•S/N = C/N + 40 dB for the
FM system
• AM bandwidth per
channel
= 4 MHz
• FM bandwidth per
channel
= 30 MHz
Observations:

•If the per-channel OMI is 5 %, then a RIN < -120 dB/Hz is needed for
each FM TV program, requiring S/N > 56 dB.
•For an AM system a laser with an RIN value of -140 dB/Hz can barely
Multichannel Frequency Modulation (3)
Example 9-6: Power budget vs
OMI. Assume that
•Laser power coupled into SMF =
0 dBm
• RIN = -140 dB/Hz
•pin receiver with a 50- front
end
• Preamplifier noise figure = 2
dB
•AM bandwidth per channel = 4
MHz
•FM bandwidth per channel = 30
MHz
• FM bandwidth per channel = 30
MHz
•Assuming a per-channel OMI of 5 %, the AM system has a power
margin of about 10-dB for a 40-dB SNR, whereas the FM system has a
 Subcarrier Multiplexing
• The figure below shows the basic concept of an SCM system. The
input to the transmitter consists of a mixture of N independent
analog and digital baseband signals.
• Each incoming signal si(t) is mixed with a local oscillator (LO) having
a frequency fi. The LO frequencies employed are in the 2 to 8 GHz
range and are known as the subcarriers.
• Combining the modulated subcarriers gives a composite FDM signal
which is used to drive a laser diode.
 Subcarrier Multiplexing (2)
• At the receiving end, the optical signal is directly detected
with a wideband InGaAs PIN photodiode and reconverted to a
microwave signal.

• For long-distance links, one can also employ a wideband

InGaAs APD with a 50 to 80-GHz gain-bandwidth product or
use an optical preamplifier.

• For amplifying the received microwave signal, one can use a

wideband LNA or a pin-FET receiver.