Digital Communication System

Chapter 5 : Synchronization
Lecture 1: Introduction & Receiver
Synchronisation
Introduction
Definition
Cost-Benefit Issues
Approach & Assumptions
Receiver Sync
Frequency & Phase Sync
Lecture 2: Receiver Sync (contd)

Acquisition
Phase Tracking Errors &
Performance
Symbol Sync Discrete Case

Lecture 3: Receiver Sync (contd)
Data-Aided Sync
Non-Data-Aided Sync
Frame Sync
Lecture 4: Network Sync
Open-Loop Transmitter
Sync
Closed-Loop Transmitter
Sync
Conclusion
Synchronization
Synchronization is one of the most critical functions performed at
the receiver of a synchronous communication system. To some
extent, it is the basis of a synchronous communication system.
Carrier synchronization
Symbol/Bit synchronization
Frame synchronization
Synchronization
Carrier synchronization
To recover the signal without distortion, receiver
needs to estimate and compensate for frequency and
phase differences between a received signals carrier
wave and the receivers local oscillator for the purpose
of coherent demodulation.
As to digital communication system, symbol/bit
synchronization and frame synchronization are also
required.
2011.11
Frame synchronization
Receiver can proceed by every group of symbols instead of every
single symbol, such as a frame in TDM system. Similar with symbol/bit
synchronization, the process of extracting such a clock signal is called
frame synchronization.
Symbol/bit synchronization
The output of the receiving filter must be sampled at the
symbol rate and at the precise sampling time instants. Hence, we
require a clock signal. The process of extracting such a clock
signal at the receiver is called symbol/bit synchronization.
Synchronization
Symbol Sync
Frame Sync
Coherent & Non-coherent
Demodulation
Coherent demodulation needs:
1) Phase sync (carrier recovery)
2) Symbol sync
3) Frame sync

Non-coherent demodulation needs:
1) frequency sync (carrier recovery)
2) symbol sync
3) frame sync
Synchronization Defined
Phase synchronization.
In phase lock means that the receivers LO is synchronized in
both frequency and phase with the received signal
Symbol Synchronization
- In symbol lock means that the receiver must, in effect, be able to
produce a square wave that will transition through zero
simultaneously with the incoming signals transitions between
symbols.
Frame Synchronization
- equivalent to being able to generate a square wave at the frame
rate, with the zero crossings coincident with the transitions from one
frame to the next.
- e.g. block code used for forward error control and in TDMA
Synchronization Defined
Frequency Synchronization
- differs from phase sync in that the replica of the carrier that is
generated by the receiver is allowed to have an arbitrary constant
phase offset from the received carrier.
- receiver design leads to degradation of performance versus SNR.

Network Synchronization
- transmitter assumes a more active role in sync by varying the
timing and frequency of its transmissions to correspond to the
expectations of the receiver.
- transmitter sync is also called network sync.

Tradeoff
Between Performance and Versatility (Complexity)

There are added cost to receiver design due to the
implementation of acquisition and tracking circuits.
Time required for synchronization to be achieved.
Energy expended, for instance on pilot signals, for the
purpose of synchronization.
Choice of modulation techniques
Coherent BPSK is 4 dB better compared to non-coherent FSK
for the same BER

Tradeoff
Complexity due to error control coding
Frame, block, message synchronization.
Complexity due to spread spectrum technique.
PN sequence synchronization.

Receiver Synchronization
All digital communication receivers require some
degree of synchronization to the incoming
signal.
The goal of this topic is to grasp the broad
intuitive understanding of sync rather than
attempting to describe a catalog of synchronizer
design methods.
As a result, we will follow traditional analog
development knowing that the sync principles
apply equally well to sampled-data systems (e.g.
phase-locked-loops)
Frequency and Phase
Synchronization
A Phase-Locked-Loop (PLL) is at the heart of nearly all
synchronization circuits.
In modern digital receivers, this loop may be difficult to
recognize, but the functional equivalent is essentially
always present.

Frequency and Phase
Synchronization
PLL are servo-control loops, whose
controlled parameter is the phase of a
locally generated replica of the incoming
carrier signal.
Components of a PLL:
A phase detector
A loop filter
A voltage-controlled oscillator (VCO).
Frequency and Phase
Synchronization
A phase detector determines the difference in phase
between the incoming signal and the reference signal.
The loop filter controls the response of the PLL to the
error signal.
The VCO is an oscillator whose output frequency is a
linear function of its voltage over some range of input
and output.
+ve signal increase frequency beyond the uncontrolled value.
-ve signal reduce frequency below the uncontrolled value.


Frequency and Phase
Synchronization
For a normalized
input signal of the
form:


Consider a
normalized VCO
output of the form:

)] ( cos[ ) (
0
t t w t r u + =
)] (

sin[ 2 ) (
0
t t w t x u + =
Frequency and Phase
Synchronization
Output error signal at
the phase detector
output:



If the filter output is
low-pass, we will
have
)] (

) ( 2 sin[ )] (

) ( sin[
)] ( cos[ )] (

sin[ 2 ) ( ) ( ) (
0
0 0
t t t w t t
t t w t t w t r t x t e
u u u u
u u
+ + + =
+ + = =
)] (

) ( sin[ ) ( t t t y u u =
Frequency and Phase
Synchronization
The low pass filter produces an output that is solely the
function of difference in phase between the two signals.
The VCO output will be a linear function of y(t).
Deviation in frequency is given as (also known as
linearized loop equation)
) ( * )] (

) ( [ ) ( * ) (
) (
)] (

[
) (
t f t t K t f t e K
t y K
dt
t d
t
o o
o
u u
u
e
~ =
= = A
Gain of the VCO
Loop filter impulse response
Frequency and Phase
Synchronization
The Fourier transform of the difference equation


leads to


Reorganizing, we have
) ( * )] (

) ( [ ) ( * ) (
) (
)] (

[
) (
t f t t K t f t e K
t y K
dt
t d
t
o o
o
u u
u
e
~ =
= = A
) ( )] (

) ( [ ) ( w F K j
o
e e e e O O = O
) (
) (
) (
) (
) (

e
e e
e
H
w F K j
w F K
o
o
=
+
=
O
O
Closed-loop transfer function
Frequency and Phase
Synchronization
The order of the PLL is the order of the
highest term in jw in the denominator of
H(w).
Previous equation indicates that this is
always one more than the order of the
loop filter F(w).
Order of PLL is critical for determining the
loops steady-state response to a steady-
state input.
Steady state tracking Characteristic
of the PLL
The expression for the Fourier transform of the
phase error can be given as:


The steady-state error is the residual error after all
transients have died away.

) (
) (
) ( )) ( 1 ( ) (

) ( )} ( { ) (
w F K j
j
H t e F E
o
+
O
= O = O O = =
e
e e
e e e e e
) (
) ( ) (
)} ( { ) (
2
0
lim lim
w F K j
j
t e F j t e
o
j t
+
O
= =

e
e e
e
e
Steady state tracking Characteristic
Refer to Example 10.2, 10.3 and 10.4.
Consider a loops steady-state response to
a phase step at the loop input
Next, consider a loops steady-state
response to a frequency step at the input.
Lastly, consider a loops steady-state
response when the input frequency is
changing linearly with time (a frequency
ramp function).
Steady state tracking Characteristic
In practice, vast majority of PLL designs are second
order due to 2 reasons:
1) a second-order loop can be made to be unconditionally stable.
Stable loops will always try to track the input. No set of input
conditions regardless of how extreme will cause the loop to
respond in the inappropriate direction to changes in the input.

2) Second-order loops will track out the effect of a frequency step
(Doppler shift) and they are relatively easy to analyse since the
closed-form results obtained for first-order loops are good
approximations for second-order loop performance.


Performance in noise
The input might be noisy, as is the case in many
communication systems.


n(t) can be expanded into quadrature
components.
) ( ] cos[ ) (
0
t n t w t r + + = u
t w t n t w t n t n
s c 0 0
sin ) ( cos ) ( ) ( + =
Performance in noise
The output of the phase detector can be written as

The loop filter eliminates the high-frequency components. We
are then left with
Let us denote the variance of n
/
(t) by
2
n
.
It can be shown that the variance of the output phase is:

For the special case of white noise, the phase variance is:


This phase variance is also related to the

frequency) carrier the at twice terms ( ]

cos[ ) ( ]

sin[ ) ( ]

sin[ ) ( ) ( ) ( + + + = = u u u u t n t n t r t x t e
c s
]

cos[ ) ( ]

sin[ ) ( ]

sin[ ) ( u u u u t n t n t n
c s
+ + = '
e e e
t
o
u
d H G
2 2

| ) ( | ) (
2
1
}

=
e e
t
o
u
d H
N
o
2 2

| ) ( |
2
}

=
L o
B N 2
2

=
u
o
Performance in Noise
The phase variance is a measure of the amount of jitter
or wobble in the VCO output due to noise at the input.
Clearly, one would wish phase variance to be small,
which for a given noise level implies small loop
bandwidth, B
L
which implies a narrow H(w).
This infers the narrower the effective bandwidth of H(w),
the poorer will be the loops ability to track incoming
signal phase changes, (w).
Thus, a loop design must balance noise response with
desired input phase response.
Acquisition
Acquisition is the process of getting the
PLL to lock with the incoming signal.
Can be accomplished by:
Aided acquisition
With the aid of external circuits or signals
Self-acquisition
Without the aid of external circuits or signals
(unaided PLL)
Symbol synchronization
Symbol synchronization is needed in order to achieve
optimum demodulation.
2 classes of symbol synchronizers:
Non-Data Aided (NDA)
class of synchronizers that knows nothing about the actual data
sequence e.g. cyclic prefix in intrinsic structure of OFDM symbols.

Data Aided (DA)
class of synchronizers that use known information about the data
stream.
This knowledge may be obtained by feeding back decisions on
received data or because a known sequence has been injected
into the data stream such as training symbols,
preambles/midambles or pilot frequencies.
Symbol synchronization
It is clear that DA techniques result in better synchronisation
performance, but some part of bandwidth or data
transmission capacity is lost for the pilot signals or training
sequences.
3 Types of Open-loop Symbol Synchronization (Bit Synchronizers)
Closed-loop symbol synchronization
(Early/late-gate data synchronizer)
Closed-loop symbol
synchronization (Symbol Timing)
Correct receiver timing
Early receiver timing
Symbol Synchronization Errors and Symbol
Error Performance
Fig. 10.15 shows the effect of symbol-sync error on bit-
error probability for a BPSK signal in additive white
Gaussian noise
From the figure, the degradation is less than about 1 dB
in SNR for a fractional timing jitter of less than 5%.
Figure 10.15
Example 10.7 Effect of Timing Jitter
Through the use of Fig. 10.15, determine the effect of a
10% symbol-fractional timing jitter on a system required
to maintain a 10
-3
bit error probability.

Solution

It can be seen from Fig. 10.15 that a 10
-3
bit error
probability will require a SNR of about 6.7 dB in the
absence of all timing jitter. The same figure indicates
that for a fractional timing jitter of 10% (
e
/T = 0.1), a
SNR of about 12.9 dB is required.

Example 10.7 (contd)
Thus, the ability to accommodate this large timing jitter
would require a 6.2 dB higher SNR than that needed to
maintain a 10
-3
bit-error probability without jitter. This
illustrates a use to which Fig. 10.15 can be put; however,
this example is clearly extreme. No communication
system would be designed with over four times the
nominally required power level in order to accommodate
a large-symbol sync error.
Frame synchronization
Applies when the transmitted data have
some sort of block or frame structure.
Application dependent
Examples:
Frame marker (e.g. used in T1 carrier)
Codeword transmission (e.g. system with
inconsistent or bursty transmissions)
Frame synchronization
Frame Marker

The simplest frame sync aid (advantage).
A single bit or a short pattern of bits that
transmitter injects periodically into the data
stream.
The receiver, having achieved data sync,
correlates the known pattern with the incoming
data stream at the known injection interval.
Frame Synchronization
Frame synchronization
Sync Codeword

Approach for systems with inconsistent or bursty
transmissions or with rapid acquisition requirements.
A sync codeword would typically be sent as part of a
message header.
The receiver must know the codeword and be constantly
searching for it in the data stream, possibly with a
matched filter correlator.
Detection of codeword indicate a known position
(typically the beginning) in the data frame.
Network synchronization
Receiver sync
Coherent modulation
Broadcast (one direction communication)
Single-link communication (e.g. microwave link, land-
line links or fiber-optic links)
Terminal sync-transmission parameters
are changed to obtain sync
Noncoherent modulation
Multi user systems
Satellite communication
TDMA systems
Network Sync
For communication systems using non-coherent modulation
techniques or that involve many users accessing a central
communication node such as many satellite communication
systems, it often makes sense for sync to be mostly or entirely a
terminal function.
This means that the terminal transmitter parameters are modified to
achieve sync rather than modifying the central nodes receiver
parameters.
This must be the approach if the system uses TDMA. In TDMA,
each user is allotted a segment of time in which to transmit its
information.
The terminal transmitter must be synchronized with the system in
order for its transmitted burst of data to arrive at the central node at
the time when the node is prepared to receive the data.

Network Sync
Transmitter sync procedures may be classified as being either open
loop or closed loop.
Open loop techniques dont rely on any measurement of the arriving
signal parameters at the central node.
The terminal pre-corrects its transmission based on stored
knowledge of link parameters that have been provided by some
external authority but may possibly be modified by observations of a
return signal from the central node.
The main advantages of the open loop methods are that acquisition
is fast the procedure can work without a return link and the amount
of real-time computation that is required is small.
The disadvantages are that they require the existence of the
external authority that provides knowledge of the required link
parameters and that they are relatively inflexible.

Network Sync
Close-loop techniques, however, require little in the way of a priori
knowledge of link parameters.
Closed-loop methods involve measurements of the sync accuracy of
the incoming transmissions from the terminal upon their arrival at the
central node and the return of the results of these measurements to
the terminal via a return path.
The advantages are that no external source of knowledge is
required for the system to work and the responses on the return link
allow the system to adapt easily and quickly to changing geometries
and link conditions.
The disadvantages are that they require a relatively large amount of
real-time processing, requiring two-way links to every terminal and
that acquisition can take a relatively long time.