Journal of

Marine Science
and Engineering

Article
A Carrier-Based Gardner Timing Synchronization Algorithm
for BPSK Signal in Maritime Communication
Fengzhong Qu 1,2 , Zhengchao Li 1 , Minhao Zhang 1 , Xingbin Tu 1,2, * and Yan Wei 1,2

1 The Engineering Research Center of Oceanic Sensing Technology and Equipment, Ministry of Education,
Zhejiang University, Zhoushan 316021, China
2 Donghai Laboratory, Zhoushan 316021, China
* Correspondence: xbtu@zju.edu.cn

Abstract: Wireless communication at sea is an essential way to establish a smart ocean. In the
communication system, however, signals are affected by the carrier frequency offset (CFO), which
results from the Doppler effect and crystal frequency offset. The offset deteriorates the demodulation
performance of the communication system. The conventional Gardner bit-synchronization algorithm
performs timing synchronization on the baseband, but it fails to solve the problem of carrier frequency
offset. In this paper, a carrier-based timing-synchronization Gardner algorithm was proposed. The
algorithm performed error detection in the carrier signal to estimate the synchronization error in real
time and the time-domain expansion in the passband signal. Based on the estimated expansion factor,
the algorithm located the sampling points of the ideal signal and re-interpolated all the sampling
points on the passband to recover the passband signal without Doppler. This algorithm can solve both
the frequency shift and time-domain expansion caused by the Doppler effect without further CFO
estimation and compensation, thus reducing the overall computational complexity of the system. The
simulations showed that the proposed algorithm greatly improves the ability of the communication
system to combat Doppler compared to the conventional Gardner algorithm. The algorithm is
primarily designed for binary phase-shift keying (BPSK)-modulated signals under the additive white
Gaussian noise (AWGN) channel and can be applied to various maritime communication scenarios.

Keywords: Doppler; Gardner algorithm; carrier; maritime wireless communication; medium wave
Citation: Qu, F.; Li, Z.; Zhang, M.; Tu,
X.; Wei, Y. A Carrier-Based Gardner
Timing Synchronization Algorithm
for BPSK Signal in Maritime
1. Introduction
Communication. J. Mar. Sci. Eng.
2023, 11, 829. https://doi.org/
Medium-wave communication is a radio communication technique that uses electro-
10.3390/jmse11040829
magnetic waves with a wavelength of 1000 m~100 m (a frequency range of 300~3000 kHz)
and is an important component of maritime communication. Medium-wave communica-
Academic Editor: Kamal Djidjeli
tion is widely used for maritime emergency rescue, aircraft and ship communication, and
Received: 13 March 2023 radio navigation. However, in digital communication systems, the received carrier signal is
Revised: 8 April 2023 affected by the Doppler effect and the local crystal frequency offset, which creates frequency
Accepted: 11 April 2023 offset in the frequency domain and changes the length of the signal in the time domain. To
Published: 13 April 2023 synchronize the carrier signal with the local clock signal, carrier synchronization and bit
synchronization are required at the receiver. Bit synchronization, also known as symbol
synchronization, extracts the timing pulse sequence for sampling judgments at the receiver
side and is the foundation for accurate sampling judgments.
Copyright: © 2023 by the authors. Bit-synchronization algorithms can be divided into feedforward synchronization algo-
Licensee MDPI, Basel, Switzerland. rithms and feedback synchronization algorithms according to the algorithm structure. It can
This article is an open access article
also be classified as data-assisted and non-data-assisted algorithms [1,2]. The classical syn-
distributed under the terms and
chronization algorithms include the early-late gate (E-LG) algorithm [3], M&M (Mueller &
conditions of the Creative Commons
Muller) algorithm [4], maximum likelihood algorithm [5], wave division multiplex (WDM)
Attribution (CC BY) license (https://
algorithm [6], and Gardner algorithm [7]. Among them, refs. [3,4,6,7] belong to timing
creativecommons.org/licenses/by/
recovery synchronization algorithms. The Gardner timing recovery bit synchronization
4.0/).

J. Mar. Sci. Eng. 2023, 11, 829. https://doi.org/10.3390/jmse11040829 https://www.mdpi.com/journal/jmse

J. Mar. Sci. Eng. 2023, 11, 829 2 of 14

algorithm requires only two resamples in one symbol for error detection. The carrier phase
does not affect the synchronization judgment performance, data assistance and judgment
feedback are not required, and it is easy to implement in high-speed digital communi-
cation systems. Hence, it is widely used in engineering practice. Gardner algorithm is
continuously applied to quadrature amplitude modulation (QAM) signals [8–10], multi-
ple phase shift keying (MPSK) signals [11,12], Gaussian minimum frequency shift keying
(GMSK) signals [13], and other modulation systems [14–16], with an ever-expanding range
of applications.
The Gardner algorithm works by resampling the rough demodulated baseband signal.
The feasibility of the algorithm depends on an assumption that the phase deviation of each
sampled value is constant, which means there is no carrier frequency deviation. When the
assumptions are not satisfied, the performance of the algorithm may degrade. As a result,
improved algorithms for correcting frequency deviation are proposed [10]. In addition,
Gardner algorithm introduces a certain amount of self-source noise when the roll-off factor
(ROF) is small and adjacent symbols are the same, which affects the performance of bit
synchronization.
Improved algorithms using pre-filtering method [8,17] and compensation method [18,19]
have been gradually proposed for Gardner algorithm with large self-jitter in systems with
small ROF. However, the complexity of these improved algorithms is relatively high. For the
method applicable to MPSK modulation proposed in [11], ref. [20] proposed an improved
algorithm applicable to higher-order MPSK modulation. Ref. [21] improved the Gardner
phase detector (GPD) so that the algorithm still performs well in QAM modulated systems
with the ROF close to 0.
When adjacent symbols are the same, the symbol difference is always zero and the
Gardner algorithm cannot capture the timing error, which makes the Gardner algorithm
generate large timing jitter [22] and leads to a high level of self-noise in the system. Some
improved algorithms were proposed to address this issue [23]. Correction algorithms [9]
and preprocessing algorithms [24] were proposed to mitigate self-noise in the Gardner algo-
rithm. To solve the problem of long synchronization timing time, an improved method that
increases the gain control processing was proposed in [25]. Other research directions on bit
synchronization and the Gardner algorithm include frequency-domain bit-synchronization
algorithms [26,27], parallel processing on field programmable gate array (FPGA) [28],
the effect of loop delay [29] on system performance, and performance comparisons with
various algorithms [30,31].
The existing improved methods of the Gardner algorithm all detect timing errors based
on the baseband signal, which is vulnerable to residual-frequency offset. In addition, the
system needs to perform additional frequency-offset estimation and compensation. Based
on the characteristics of the carrier signal, this paper proposed a Carrier-Based Gardner (CB-
Gardner) algorithm for timing synchronization. The algorithm can directly perform error
detection on the carrier signal in the passband, eliminate the carrier frequency offset caused
by Doppler effect and local crystal frequency bias, and perform timing synchronous sam-
pling. The main improvements of the algorithm include: (1) changing the error-detection
point and the error-calculation formula; (2) simplifying the operation of the numerically
controlled oscillator (NCO); and (3) resampling all sampling points on the passband us-
ing an interpolator. This algorithm can be applied to single-carrier signals with binary
phase-shift keying (BPSK) modulation, especially in wireless communication systems.

2. Problem Formulation
2.1. Problem Statement
In this paper, the main problem we address is the effect of the Doppler on the carrier
signal. In a communication system, the carrier frequency of the signal may change due to
the relative motion between the transmitter and the receiver. The carrier frequency with
Doppler effect is
v ± v0
f0 = ( ) fc (1)
v ∓ vs
J. Mar. Sci. Eng. 2023, 11, 829 3 of 14

where f 0 denotes the observed signal frequency, v denotes the wave propagation speed, v0 is
the movement speed of the receiver, and vs denotes the movement speed of the transmitter
source, f c is the original frequency of the signal. The ratio of f 0 and f c is defined as

λ = f 0 / f c = ( f c + f 4 )/ f c = 1 + δ (2)

where f 4 denotes the absolute Doppler shift, and the Doppler factor δ = f 4 / f c denotes
the magnitude of the Doppler shift with respect to the carrier frequency.
In the time domain, the Doppler effect causes the length of the received signal to be
compressed or broadened, resulting a length ratio of 1/λ between the received signal and
the original signal. This length change corresponds to a change in the sampling rate at the
receiver. If the original sampling rate is Fs , then the actual sampling rate is

Fs0 = Fs /λ (3)

Thus, in the time domain, the effect of Doppler is equivalent to signal resampling.
The frequency shift caused by Doppler prevents the receiver from performing accurate
carrier synchronization. The change in sampling rate also affects bit synchronization and
thus degrades the decoding performance. Considering a single-carrier communication
system with BPSK modulation, if the transmitted symbol is xn , the transmitted signal can
be expressed as 
A cos(wc t) , xn = 0
sBPSK (t) = (4)
− A cos(wc t) , xn = 1
The CB-Gardner algorithm proposed in this paper processes the passband signal in
(4). This method can also be extended to higher-order modulations.

2.2. Solutions to Doppler

To eliminate the Doppler effects, two methods are usually used.
Method 1: Resampling the passband signal directly. This method requires estimating
the Doppler factor. One approach is to calculate the Doppler factor by measuring the change
of the length of signal through a synchronization method in the time domain and resample
the passband signal. This method assumes a constant Doppler factor throughout the signal
time, which may not hold true for time-varying Doppler. As a result, the performance of
this method may degrade.
Method 2: Carrier frequency offset (CFO) estimation and compensation, and bit
synchronization. The former eliminates the frequency offset of the signal, allowing for
accurate bit synchronization, which eliminates the spreading effect of the signal in the time
domain. The Gardner algorithm is a canonical example of bit-synchronization algorithms.
Figure 1 shows the closed-loop feedback structure of the Gardner algorithm [32]. The
loop consists of an interpolator, a timing error detector (TED), a loop filter (LF) and an
NCO.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER The input signal is the baseband signal after down-conversion, and the output 4isofthe
REVIEW 16
baseband symbol after decimation.

Figure1.1.The
Figure Thestructure
structureof
ofGardner
Gardneralgorithm.
algorithm.

Figure 2 shows the principle of the Gardner algorithm for TED on the baseband sig-
nal. The figure depicts a baseband signal containing two symbols, where y ( n − 1) ,
y ( n − 1 / 2 ) and y ( n ) are the three detection points required to detect timing error.
J. Mar. Sci. Eng. 2023, 11, 829 Figure 1. The structure of Gardner algorithm. 4 of 14

Figure 2 shows the principle of the Gardner algorithm for TED on the baseband sig-
nal. The figure depicts a baseband signal containing two symbols, where y ( n − 1) ,
Figure 2 shows the principle of the Gardner algorithm for TED on the baseband signal.
They ( nfigure
− 1 / 2 )depicts
and ay (baseband
n ) are the three
signal detectiontwo
containing points required
symbols, whereto ydetect
(n − 1)timing
, y(n − error.
1/2)
andy ( n − 1) and y ( n ) are the optimal sampling instants for the two adjacent symbols
y ( n ) are the three detection points required to detect timing error. y ( n − 1 ) and y ( n ) arein
the optimal sampling instants for the two adjacent symbols in the baseband, and y(n − 1/2)
isthe
thebaseband,
midpoint and y ( ntwo
of these 2 ) is thesymbols.
− 1 /adjacent midpointTherefore,
of these two adjacent algorithm
the Gardner symbols. Therefore,
performs
two resampling points per symbol on average.
the Gardner algorithm performs two resampling points per symbol on average.

Figure2.2.Timing
Figure Timingerror
errordetection
detectionschematic
schematicofofGardner
Gardneralgorithm.
algorithm.

Letx (xn( −
Let n −11))and
andx (nx)( nbe
) the symbols
be the symbols adjudicated
adjudicatedbased y(n y−( 1n)−and
on on
based 1 ) and ( n) .
y(n). yThe
error detection equation of Gardner algorithm is
The error detection equation of Gardner algorithm is
1 1
errnerr n −y(n) *−[ x2(n) −
= yn (= ∗ 1)
[ x (−nx− 1) − x (n)]
(n)] (5)
(5)
2
whereerrerr
where contains
n contains
n
the information
the information that
that the the is
signal signal is extended
extended or compressed
or compressed and
and therefore
Doppler can be detected.
therefore Doppler can be detected.
As
Asseen
seenin inFigure
Figure2,2,the
theshape
shapeofofthe
thebaseband
basebandsignal
signalbears
bearssome
somesimilarity
similarityto
tothe
the
cosine
cosine signal. This similarity serves for timing error detection in CB-Gardner, whichcan
signal. This similarity serves for timing error detection in CB-Gardner, which can
eliminate
eliminatethetheimpact
impactofofDoppler
Dopplerdirectly
directlyon
onthe
thepassband
passbandwhile
whilemaintaining
maintainingthe
thereal-time
real-time
iterative property of the Gardner algorithm.
iterative property of the Gardner algorithm.
3. Materials and Methods
3. Materials and Methods
The proposed CB-Gardner in this paper is a feedback loop consisting mainly of two
The proposed CB-Gardner in this paper is a feedback loop consisting mainly of two
interpolators, a timing error detector, a loop filter, and a numerically controlled oscillator.
J. Mar. Sci. Eng. 2023, 11, x FOR PEER interpolators,
REVIEW a timing error detector, a loop filter, and a numerically controlled oscillator.
5 of 16
This feedback loop is executed once per symbol to counteract the time-varying Doppler.
This feedback loop is executed once per symbol to counteract the time-varying Doppler.
Figure 3 illustrates the structure and model of the proposed CB-Gardner.
Figure 3 illustrates the structure and model of the proposed CB-Gardner.

Figure 3. The
Figure 3. The structure
structure of
of carrier-based
carrier-based Gardner
Gardneralgorithm.
algorithm.

The fundamental difference between CB-Gardner and Gardner algorithm lies in the
The fundamental difference between CB-Gardner and Gardner algorithm lies in the
error detector, which is mainly reflected in the difference of the location of error detection
error detector, which is mainly reflected in the difference of the location of error detection
point and the change of the error detection equation. Error detection in the CB-Gardner can
point and the change of the error detection equation. Error detection in the CB-Gardner
be divided into two steps: finding the error-detection point and calculating the timing error.
can be divided into two steps: finding the error-detection point and calculating the timing
error. An interpolator is also needed before calculating the timing error. Next, we will
describe the detailed process of the algorithm.

3.1. Error Detection Point

J. Mar. Sci. Eng. 2023, 11, 829 5 of 14

An interpolator is also needed before calculating the timing error. Next, we will describe
the detailed process of the algorithm.

3.1. Error Detection Point

For a BPSK signal, (4) shows that the signal can be represented as a cosine waveform.
However, the signal amplitude around the symbol instant decreases gradually due to the
forming filter. Nevertheless, since the symbol period is usually much larger than the carrier
period for medium wave, the amplitude of the filter can be approximated as constant when
considering only the waveform of the first carrier period after the symbol instant. At this
point, the passband waveform can be approximated as a cosine function.
Observing the baseband waveform in Figure 2, we find that its shape is similar to
that of a cosine waveform. Therefore, the three points used for error detection in the CB-
Gardner are chosen with a phase difference of π/2, where the phase of the first detection
point belongs to [0, π2 , π, 3π
2 ], and all detection points are located after the symbol instant.
Specifically, the phase values of the three error-detection points (y1 , y2 , y3 ) are

 ϕy1 ∈ [0, π2 , π, 3π

2 ]
ϕy2 = ϕy1 + π2 (6)
ϕy3 = ϕy1 + π


To minimize the decay of the cosine waveform, we choose the optimal sampling
instant for each symbol as the reference and position of y1 as close to the symbol instant
as possible. Figure 4 shows the positions of the three detection points (y1 , y2 , y3 ) in the
improved algorithm when ϕy1 = 0. S1 is the optimal judgment instant for one symbol
and S2 is the optimal judgment instant for another symbol after S1 . For symbol S1 , the
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW 6 of 16
nearest y1 is located at the instant with a phase of 0. The phases of y2 , y3 are then π/2 and
π, respectively.

Figure4.4.The
Figure Thepositions
positionsof
ofthree
threedetection
detectionpoints
points(y( 1y,1y, y2 ,2 y, y3 3) )ininCB-Gardner.
CB-Gardner.

Suppose
Suppose thethe symbol rate isis FbF,b ,and
symbol rate andthe thesymbol
symbolperiodperiodisisT T==1/F 1 / Fbb. . We
We set
set the
thefirst
first
symbol instant to t1 = 0, its sampling point to P1 = 1. For the nth symbol instant, we have
symbol instant to t = 0 , its sampling point to P = 1 . For the nth symbol instant, we have
tn = (n − 1) T and its1 sampling points are Pn . The1 phase of the nth symbol instant is
tn = ( n − 1) T and its sampling points are Pn . The phase of the nth symbol instant is
( n − 1) w c T , xn = 0
ϕn = w c t n + ϕ x n = c(n −+1)π,
wcxTn = ,1xn = 0 (7)
( n −
ϕn = wc tn + ϕxn =  1 ) w T
(7)
(n − 1)wcT + π, xn = 1
where wc = 2π f c . At this point, the instant of the nearest y1 after tn is
where wc = 2πfc . At this point, the instant of the nearest y 1 after tn is
ty1 = tn +  π2 − [ϕn mod( π2 )] /wc

π π (8)
= tn t+ y1
=π2tn −
+ {[(n−−[ϕ1n)mod(
wc T ]mod π
)]} /( w
2 c) /wc
2 2 (8)
where the time difference ty1 with tn is=4t ty+1 ,n π =
π
− [( n − 1 ) w π
T ] mod ( π
) /w . For the
n
{ − [(n2− 1)wcT ]mod( )} / wc 2 c c
current nth symbol, we can use the estimated 2 λn−1 from the previous 2 symbol as a priori
π π
where the time difference ty1 with tn is ty1 ,n = { − [(n − 1)wcT ]mod( )} / wc . For the
2 2
current nth symbol, we can use the estimated λ n−1 from the previous symbol as a priori
information. Considering the variation of the signal length due to Doppler, the number of
J. Mar. Sci. Eng. 2023, 11, 829 6 of 14

information. Considering the variation of the signal length due to Doppler, the number of
sampling points corresponding to 4ty1 ,n is

4p = 4
 ty1 ,n Fs /λn−1
= π2 − [(n − 1)wc T ]mod( π2 ) /wc ∗ Fs /λn−1
(9)
= 1/λn−1 ∗ π2 − [(n − 1)wc T ]mod( π2 ) /(2π) ∗ Ff cs


= 1/λn−1 ∗ spc ∗ π2 − [(n − 1)wc T ]mod( π2 ) /(2π)



where spc = Fs / f c denotes the average number of samples per carrier cycle. Therefore, the
positions of the error-detection points (y1 , y2 , y3 ) are
 π π
 py1 ,n = Pn + 1/λn−1 ∗ spc ∗ 2 − [(n − 1)wc T ]mod( 2 ) /(2π)
p = py1 ,n + 1/λn−1 ∗ spc/4 (10)
 y2 ,n
J. Mar. Sci. Eng. 2023, 11, x FOR PEER REVIEW py3 ,n = py1 ,n + 1/λn−1 ∗ spc/2 7 of 16

Once we get the positions of all three error detection points, we can obtain their
corresponding values using the interpolation method, which will be described later.
3.2. Timing Error Detection
3.2. Timing
UnlikeError Detection
(5), the CB-Gardner performs error detection by utilizing three sampled val-
ues Unlike
on the (5),
passband instead of
the CB-Gardner determined
performs symbols. by
error detection When ϕy1 three
utilizing ∈ [0, πsampled
] , for example
values
on the passband instead of determined symbols. When
ϕy1 = 0 , as shown in Figure 5, and assuming that the position ϕ y 1
∈ [ 0, π ] , for example ϕ
of y 1 is relatively accurate,y 1
= 0,
as shown in Figure 5, and assuming that the position of y1 is relatively accurate, the signal
the signal
may broaden mayduebroaden due toofthe
to the effect effect of(i.e.,
Doppler Doppler
λ < 1), andλthe
(i.e., < 1detection
), and thepointsdetection points
y2 and y3
y 2 and
may shift. y 3 may shift.

Figure5.5.The
Figure Thepositions
positionsof
ofthree
threeactual
actualdetection
detectionpoints
points(y( 1y,1y, y2 ,2 y, y3 3),),when λ <1,1ϕ, y1ϕ=
whenλ < =0.
y1 0.

Referring
Referringto tothe
theerror
errordetection
detectionEquation
Equation(5) (5)of
ofthe
theGardner
Gardneralgorithm,
algorithm,we wedefine
definethethe
detection error
detection error asas
errn = y2 ∗ (y1 − y3 ) (11)
errn = y 2 * ( y1 − y 3 ) (11)
It can be inferred that errn > 0, which leads to a decrease in the control word in the
NCO.ItConsequently,
can be inferred that
the NCO > 0 , which
errnestimates leads
that λn <to1.a Conversely,
decrease in the control
when word
λ > 1, in the
the NCO
estimates that λ
NCO. Consequently,n > 1. the NCOn estimates that λ n < 1 . Conversely, when λ > 1 , the NCO
Thus, err accurately indicates the directional information of Doppler
and can be used as timing synchronization error. But when ϕy1 ∈ [ π2 , 3π ], for example,
estimates that λ > 1 . Thus, errn accurately indicates the directional2 information of
ϕy1 = π2 , if λ < 1, nas shown in Figure 6, according to (11), we can see that errn < 0 at this
point. Thisand
willcan
give n > 1.But when ϕ ∈ [ ,
π 3π
Doppler bethe NCO
used as incorrect information of λerror.
timing synchronization y1
] , for
2 2
π
example, ϕy1 = , if λ < 1 , as shown in Figure 6, according to (11), we can see that
2
errn < 0 at this point. This will give the NCO incorrect information of λ n > 1 .
 estimates that λ n > 1 . Thus, errn accurately indicates the directional information of
π 3π
Doppler and can be used as timing synchronization error. But when ϕy1 ∈ [ , ] , for
2 2
π
example, ϕy1 = , if λ < 1 , as shown in Figure 6, according to (11), we can see that
J. Mar. Sci. Eng. 2023, 11, 829 2 7 of 14
errn < 0 at this point. This will give the NCO incorrect information of λ n > 1 .

Figure 6.
Figure 6. The
The positions
positions of
of three
three actual
actual detection
detectionpoints
points(y
( y1 ,1 ,yy22, ,yy33),),when λ <1,1ϕ
whenλ < , yϕ
1 y=
1
=π/2.
π/2.

Therefore, we need to correct (11). To differentiate between the two cases of ϕy1 ∈ [0, π]
and ϕy1 ∈ [ π2 , 3π
2 ], a simple method is to compare the absolute values of y1 and y2 when
the signal-to-noise ratio (SNR) is not very low. Thus, the final equation for CB-Gardner’s
error detection is 
y2 ∗ ( y1 − y3 ) , | y1 | > | y2 |
errn = (12)
y2 ∗ ( y3 − y1 ) , | y1 | < | y2 |
In this case, the sign of errn will correctly indicate the directional information of
Doppler. In addition, errn also contains information about the magnitude of the Doppler,
enabling the NCO to accurately detect both the magnitude and direction of the Doppler.
Since the error detection is carrier-based and independent of the specific symbol, the
CB-Gardner does not fail even when there are a large number of consecutive identical
symbols.
For BPSK signals, (11) is used for error detection, while for quadrature phase shift
keying (QPSK) and higher-order modulation methods, (12) is required.

3.3. Loop Filter

For the nth symbol, once the synchronization error errn is detected by the error detector,
it is fed into the loop filter. The loop filter here is identical to the one used in the Gardner
algorithm, which is a second-order proportional integral filter. The output of the filter can
be obtained by
n
ELF,n = c1 errn + c2 ∑ errn (13)
i =1

where c1 and c2 are scaling and integration coefficients, respectively. The filter output ELP,n
is then fed to the NCO.

3.4. Numerically Controlled Oscillator

The NCO is used to determine the position of the optimal sampling instant for the
next symbol, i.e., Pn+1 . Figure 7 shows the flowchart of NCO in the CB-Gardner algorithm.
In the CB-Gardner, we modified the expression of the control word in the NCO.
Since the NCO only needs to obtain the position of the next symbol and does not require
knowledge of the midpoint position between two symbols, the NCO only needs to overflow
once per symbol period, i.e., the NCO overflows with a period of Ti = T. Therefore, the
initial value of the control word is
1 1 1 Fs
W0 = Ts /Ti = /T = / == 1/( ) = 1/sps (14)
Fs Fs Fb Fb
 filter can be obtained by
n
ELF ,n = c1 errn + c 2  errn (13)
i =1

J. Mar. Sci. Eng. 2023, 11, 829 where c1 and c2 are scaling and integration coefficients, respectively. The filter output 8 of 14
ELP ,n is then fed to the NCO.

3.4. Numerically
where Ts isControlled period, and sps = FFs denotes the average number of samples per
Oscillator
the sampling
b
symbol.
The NCOFor the nth
is used symbol, thethe
to determine control word
position of within the NCO
the optimal is
sampling instant for the
next symbol, i.e., Pn + 1 . Figure 7 shows the flowchart of NCO in the CB-Gardner algorithm.
Wn = W0 − ELP,n = 1/sps − ELF,n (15)

FigureFigure 7. The flowchart

7. The flowchart of NCOofinNCO in CB-Gardner.
CB-Gardner.

Below, we provide a brief derivation of the register values in the NCO after each
In the CB-Gardner, we modified the expression of the control word in the NCO. Since
overflow. We define the location of each NCO overflow as the best sampling instant for
the NCO only needs to obtain the position of the next symbol and does not require
the symbols. For the nth symbol, we assume that the NCO overflows at the position of
knowledge of the midpoint position between two symbols, the NCO only needs to
Pn = mn + un , where mn is the integer part of Pn and un is the fractional part. The register
value after the overflow is ηn , which is

ηn = (mk + uk )Wn (16)

where mk is an integer and uk is a fractional number. From Pn , ηn and Wn , we can directly

obtain the position and register value of the next NCO overflow, as given by
ηn
Pn+1 = mn + 1 + mk + uk = int( Pn ) + 1 + (17)
Wn

ηn
ηn+1 = ηn − (mk + 1)Wn + 1 = ηn − [int( ) + 1]Wn + 1 (18)
Wn
Then, from the nth symbol to the (n + 1)th symbol, the NCO estimates the signal
expansion factor as
ηn
λn = sps/( Pn+1 − Pn ) = sps/[int( Pn ) + 1 + − Pn ] (19)
Wn

In this way, we estimate the magnitude of the signal expansion factor in real time.
Subsequently, we again search for the nearest three error detection points after the (n + 1)th
symbol to form a feedback loop.

3.5. Interpolator
After performing the aforementioned operation, we can obtain the position of optimal
sampling instant Pn for each symbol. However, as we are dealing with a passband carrier
signal, we need to determine the position of each sampling point and resample them in
order to recover the passband signal without Doppler effects. This necessitates the use of
interpolation algorithms.

3.5.1. Location of Sampling Points

There should be sps sampling points within each symbol. Assuming that Doppler is
approximately constant within a symbol period, these sampling points should be equidis-
J. Mar. Sci. Eng. 2023, 11, 829 9 of 14

tantly distributed. Therefore, the position interval of each sampling point between the nth
symbol and the (n + 1)th symbol is

P4n = ( Pn+1 − Pn )/sps= 1/λn (20)

The location of each sampling point is:

Pn,i = Pn + i ∗ P4n = Pn + i/λn , i = 1, 2, . . . , sps (21)

After that, we can get the values of all sampling points again according to Pn,i . For
n = 1, the position of the first symbol is P1 = 1.

3.5.2. Interpolation Algorithm

The interpolation method of CB-Gardner uses the four-point Lagrange cubic interpo-
lation method with Farrow structure [33]. Once the positions of the error detection points
(py,n ) and each sampling point (Pn,i ) are obtained, an interpolation algorithm is applied
to resample them and obtain their true values without the influence of Doppler. For any
resampling position P = m + u, where m is the integer part of P and u is the fractional part
of P, let s be the received original passband signal, the interpolated value after resampling
according to P is
yin = XT BT U (22)
where X consists of four sampling points in s:
 
s ( m + 2)
 s ( m + 1) 
X=
 s(m)

 (23)
s ( m − 1)

B is the Farrow cubic interpolation structure matrix:

0 0 1 0
 
− 1 1 − 21 − 31 
B= 6  (24)
1 1 
 0
2 −1 2
1
6 − 12 1
2 − 16

and U is the matrix determined by u:


1
 u 
U= 
 u2  (25)
u3

The resampling value for the sampling point location Pn,i is yin_n,i , and the combination
of yin_n,i forms the final passband signal sCB after Doppler elimination, i.e.,

sCB [(n − 1) ∗ sps + i ] = yin_n,i , n = 1, 2, . . . ; i = 1, 2, . . . , sps (26)

and sCB (1) = s(1). The algorithm terminates when the number of symbols (n) in a packet
reaches its maximum value (N). In the subsequent demodulation process, we only need to
perform down-conversion, matched filtering, and equalization operations on sCB to decode
the transmitted symbols x.
Algorithm 1 summarizes the proposed CB-Gardner algorithm.
J. Mar. Sci. Eng. 2023, 11, 829 10 of 14

Algorithm 1: CB-Gardner algorithm.

1: s, sps, N
2: Initialization of λ0 = 1, P1 = 1, η1 = 1, sCB (1) = s(1);
3: for n = 1 : N
4: Step 1: Finding detection points
5: getting py1 ,n , py2 ,n , py3 ,n
6: interpolating to get y1 , y2 , y3
7: Step 2: Timing Error detection
8: errn = y2 ∗ (y1 − y3 ) or errn = y2 ∗ (y3 − y1 )
9: Step 3: Loop filter
n
10: ELF,n = c1 errn + c2 ∑ errn
i =1
11: Step 4: Numerically controlled oscillator
12: Wn = 1/sps − ELF,n
13: Pn+1 = int( Pn ) + 1 + ηn /Wn
14: ηn+1 = ηn − [int(ηn /Wn ) + 1]Wn + 1
15: λn = sps/( Pn+1 − Pn )
16: Step 5: Calculating sCB
17: for i = 1 : sps
18: Pn,i = Pn + i/λn
19: interpolating to get yin_n,i
20: sCB [(n − 1) ∗ sps + i ] = yin_n,i
21: end for
22: end for

4. Simulation
To evaluate the effectiveness of the CB-Gardner algorithm, a simulation was performed
using Matlab R2017b. The transmitted signal was a BPSK single-carrier signal that passed
through an additive white Gaussian noise (AWGN) channel and had artificially Doppler
added to simulate the signal distortion caused by the Doppler effect and local crystal
frequency offset. In our method, the received signal was processed by the CB-Gardner
algorithm, after which the output signal underwent down-conversion and matched filtering.
The method differed from the traditional Gardner method where these two procedures were
inverted. Finally, equalization was performed, and the bit error rate (BER) was calculated.
The transmitted signal consisted of both pilot and data. The pilot was an m-sequence of
length 511, and the data consisted of N symbols, with both the pilot and data modulated
with BPSK.
For the Gardner algorithm, the receiver needs to perform CFO estimation and com-
pensation before applying the Gardner algorithm. The CFO estimation method used in
this paper is as follows: assume that the maximum CFO is ± f 4max , all frequency offsets
between f c − f 4max and f c + f 4max are traversed with an interval of f d , and then frequency
compensation is performed. After that, the m-sequence part of the received signal is
correlated with the local m-sequence signal, and the CFO is estimated as the frequency
offset resulting in the maximum correlation peak. When the CB-Gardner is used, the
passband signal is directly fed into it, and the receiver-side process is much simpler because
CB-Gardner eliminates both the frequency offset and the sampling bias.
Although the carrier signal is not affected by multipaths in the AWGN channel,
residual frequency offset still exists when using the Gardner algorithm due to the inaccuracy
of the estimated CFO, which can lead to an increasing phase shift of the symbols. The
phase-locked loop of the equalizer is utilized to eliminate the residual phase distortion of
the symbols. Therefore, it is necessary to use an equalization method when implementing
the Gardner algorithm. For the sake of fairness, the CB-Gardener also employs equalization,
which uses a decision-feedback equalizer based on root recursive least squares [34].
Since the CB-Gardner requires multiple sampling points in each carrier cycle of the
carrier signal for error detection, this method is suitable for medium-wave communication
scenarios with relatively low frequencies. To simulate the medium-wave communication
 the symbols. Therefore, it is necessary to use an equalization method when implementing
the Gardner algorithm. For the sake of fairness, the CB-Gardener also employs equaliza-
tion, which uses a decision-feedback equalizer based on root recursive least squares [34].
Since the CB-Gardner requires multiple sampling points in each carrier cycle of the
carrier signal for error detection, this method is suitable for medium-wave communication
J. Mar. Sci. Eng. 2023, 11, 829 11 of 14
scenarios with relatively low frequencies. To simulate the medium-wave communication
system at sea, the simulation parameters were set as follows: the number of symbols N
is 10,000, the sampling rate Fs is 2.4 MHz, the symbol rate Fb is 100 k, the carrier fre-
system atf sea,
quency c
theMHz,
is 0.4 simulation parameters
the roll-off factor were
of theset as follows:
forming filter the
α isnumber
0.2, theof lengthNisis
symbols
filter
10,000, the sampling rate Fs is 2.4 MHz, the symbol rate Fb is 100 k, the carrier frequency
40, and the feedforward and feedback filter lengths of the equalizer are 5 and 4. The num-
f c is 0.4 MHz, the roll-off factor of the forming filter α is 0.2, the filter length is 40, and
ber
theof simulationsand
feedforward is set to 4000. filter
feedback The added
lengthsfrequency offset isare
of the equalizer measured PPM (parts
5 and 4.inThe number of
per million), i.e.,
simulations is set to 4000. The added frequency offset is measured in PPM (parts per
million), i.e., PPM
PPM= δ = f / fc (27)
106 = δ = f 4 / f c (27)
106
First, we set the SNR of the passband signal to 25 dB and varied the PPM from −32,000
First, we set the SNR of the passband signal to 25 dB and varied the PPM from −32,000
to 32,000 with an interval of 4000. Figure 8 shows the simulated BER variation curves for
to 32,000 with an interval of 4000. Figure 8 shows the simulated BER variation curves
different PPM when using the Gardner algorithm and the CB-Gardner algorithm. As
for different PPM when using the Gardner algorithm and the CB-Gardner algorithm. As
shown in the figure, the BER is almost 0.5 when PPM ≥ 8000 or PPM ≤ −12,000 when using
shown in the figure, the BER is almost 0.5 when PPM ≥ 8000 or PPM ≤ −12,000 when
the Gardner algorithm, indicating that the communication system cannot work at all. For
using the Gardner algorithm, indicating that the communication system cannot work at
the CB-Gardner, however, the communication system fails when PPM ≥ 28,000 or PPM ≤
all. For the CB-Gardner, however, the communication system fails when PPM ≥ 28,000 or
−32000. Therefore, the CB-Gardner can resist a maximum PPM of about ±28000 , which is
PPM ≤ −32000. Therefore, the CB-Gardner can resist a maximum PPM of about ±28, 000,
much larger than the corresponding ±8000 value for the Gardner. For the Gardner algo-
which is much larger than the corresponding ±8000 value for the Gardner. For the Gardner
rithm, its BER is much less than 10−4 when −8000 ≤ PPM ≤ 4000. Correspondingly, the BER
algorithm, its BER is much less than 10−4 when −8000 ≤ PPM ≤ 4000. Correspondingly,
for the CB-Gardner is much less than 10−4 when −28,000 ≤ PPM ≤ 20,000. The use of the
the BER for the CB-Gardner is much less than 10−4 when −28,000 ≤ PPM ≤ 20,000. The
CB-Gardner significantly enhanced the ability of the communication system against Dop-
use of the CB-Gardner significantly enhanced the ability of the communication system
pler.
against Doppler.

Figure
Figure8.8.BER
BERfor differentPPM
fordifferent PPMwith
withGardner
Gardnerand
andCB-Gardner
CB-Gardneralgorithm,
algorithm,SNR
SNR==25
25dB.
dB.

Basedon
Based onFigure
Figure8,8,it itcan
canbebe seen
seen that
that thethe
BER BER of the
of the system
system experiences
experiences a sharp
a sharp in-
when the PPM
crease when the PPM exceeds a certain value, irrespective of whether the Gardner ororCB-
increase exceeds a certain value, irrespective of whether the Gardner CB-
Gardner algorithm is employed. The result indicates that the stability
Gardner algorithm is employed. The result indicates that the stability of the communica-of the communication
system
tion undergoes
system undergoes an an
abrupt
abrupt change
change instead
insteadofofaasmooth
smooth change
change asas the PPMincreases.
thePPM increases.
Additionally, the
Additionally, the results
results showed
showedthat thatthe
thesystem
systemcan canresist a greater
resist a greaterDoppler
Doppler when PPM
when PPM <0
in comparison with PPM > 0 for both Gardner and CB-Gardner
< 0 in comparison with PPM > 0 for both Gardner and CB-Gardner algorithm. This couldalgorithm. This could be
attributed to the smaller jitter of the Doppler factor estimated by the NCO
be attributed to the smaller jitter of the Doppler factor estimated by the NCO of the algo- of the algorithm
whenwhen
rithm PPM < PPM0, with
< 0, the
withcurrent loop filter
the current loop parameters used in
filter parameters thein
used simulation.
the simulation.
We conducted further simulations to verify the performance of CB-Gardner on SNR.
The PPM was set as −2000, and the SNR of the passband signal ranged from −10 dB to
50 dB with an interval of 5 dB. Figure 9 shows the BER variation curves for different SNRs
when using the Gardner and CB-Gardner algorithms for BPSK- and QPSK-modulated
signals. Figure 9 demonstrates that the BER using the Gardner algorithm for BPSK signals
is less than 10−6 when SNR > 10 dB. In contrast, when using CB-Gardner, the BER for BPSK
signals is less than 10−6 when SNR ≥ 20 dB. However, for QPSK signals, the BER remains
greater than 10−2 when SNR ≥ 50 dB. The simulation results indicate that CB-Gardner is
 when using the Gardner and CB-Gardner algorithms for BPSK- and QPSK-modulated
signals. Figure 9 demonstrates that the BER using the Gardner algorithm for BPSK signals
is less than 10−6 when SNR > 10 dB. In contrast, when using CB-Gardner, the BER for BPSK
signals is less than 10−6 when SNR ≥ 20 dB. However, for QPSK signals, the BER remains
J. Mar. Sci. Eng. 2023, 11, 829 12 of 14
greater than 10−2 when SNR ≥ 50 dB. The simulation results indicate that CB-Gardner is
currently only suitable for BPSK signals, and the required SNR is less than 10 dB larger
than that of the Gardner. This is because the error-detection points of the Gardner algo-
currently
rithm only suitable
are located in thefor BPSK signals,
symbol envelope,andwhereas
the required SNR is less thanpoint
the error-detection 10 dBoflarger than
the CB-
that of the Gardner. This is because the error-detection points of the Gardner
Gardner algorithm are located within one carrier cycle. As a result, it can be inferred that algorithm
areerror
the located in the in
detection symbol envelope, is
the CB-Gardner whereas the error-detection
more susceptible point
to noise than of of
that thethe
CB-Gardner
Gardner.
algorithm are located within one carrier cycle. As a result, it can be inferred
Although the error-detection equation (12) of the CB-Gardner can theoretically be that theused
error
detection in the CB-Gardner is more susceptible to noise than that of the Gardner.
for QPSK, the BER is relatively high. Thus, the CB-Gardner is currently less applicable to Although
the error-detection
QPSK signals. equation (12) of the CB-Gardner can theoretically be used for QPSK, the
BER is relatively high. Thus, the CB-Gardner is currently less applicable to QPSK signals.

Figure
Figure9.9.BER
BERfor
fordifferent
differentSNRs
SNRswith
withGardner
Gardnerand
andCB-Gardner
CB-Gardneralgorithms.
algorithms.

5.5.Discussion
Discussion
TheCB-Gardner
The CB-Gardneralgorithm
algorithm proposed
proposed in in this
this paper
paperopens
opensupupseveral
severalavenues
avenuesforfor
future
fu-
research and discussion, as described below.
ture research and discussion, as described below.
(1) In this paper, the first sampling point was set as the exact first symbol instant.
(1) In this paper, the first sampling point was set as the exact first symbol instant.
However, in a practical communication system, it is challenging to locate the first symbol
However, in a practical communication system, it is challenging to locate the first symbol
instant with such precision. One potential solution is to use pilot signals to detect the
instant with such precision. One potential solution is to use pilot signals to detect the ar-
arrival time of the signal;
rival time of the signal;
(2) CB-Gardner requires knowledge of the phase of each symbol instant to locate the
(2) CB-Gardner requires knowledge of the phase of each symbol instant to locate the
error-detection point. However, in a multipath environment, the phase of the symbol instant
error-detection point. However, in a multipath environment, the phase of the symbol in-
may be random, making it impossible for the algorithm to locate the error-detection point.
stant may be random, making it impossible for the algorithm to locate the error-detection
Although we have explored alternative methods to estimate phase, they also eliminate
point. Although we have explored alternative methods to estimate phase, they also elim-
the phase variation caused by Doppler, leading to error-detection failure. Therefore, CB-
inate the phase variation caused by Doppler, leading to error-detection failure. Therefore,
Gardner is currently unsuitable for multipath environments, especially complex underwater
CB-Gardner is currently unsuitable for multipath environments, especially complex un-
acoustic environments, but it may be useful in wireless communication with fewer channel
derwater
taps; acoustic environments, but it may be useful in wireless communication with
fewer (3)
channel
In ourtaps;
simulations, CB-Gardner had a high BER for QPSK modulated signals.
Several factors may have CB-Gardner
(3) In our simulations, contributed had a high
to this BERFirst,
result. for QPSK modulated
the phase signals.
variations Sev-
in QPSK
eral factors may have contributed to this result. First, the phase variations in
are more diverse, resulting in a waveform at the symbol instant that is less like a cosineQPSK are
more diverse, resulting in a waveform at the symbol instant that is less like
waveform. Second, the error detection Equation (12) may fail, causing the localized symbol a cosine
instant to drift. Third, the constellations of the QPSK signal are relatively dense and require
a higher SNR. These factors can increase the self-jitter of error detection, resulting in a
higher BER.

6. Conclusions
The Gardner algorithm and various improved algorithms are based on baseband
signals for error detection. In this paper, we proposed an improved Gardner algorithm
applied to the carrier signals based on the similarity between the carrier signal and the
baseband symbols in the waveform. Compared to the Gardner algorithm, the CB-Gardner
algorithm redefined the position of the error detection point, improved the error-detection
J. Mar. Sci. Eng. 2023, 11, 829 13 of 14

equation, simplified the working process of the NCO, and resampled all passband sampling
points using an interpolator. The improved algorithm performed error detection directly
on the passband while eliminating the frequency offset and sampling-point bias caused by
Doppler; thus, the estimation and compensation of CFO at the receiver side are no longer
required, simplifying the processing flow. The error detection remained valid when adjacent
symbols were the same, reducing the self-noise to a certain extent. CB-Gardner retains the
real-time iterative property of the Gardner algorithm and can resist time-varying Doppler.
The CB-Gardner algorithm was primarily designed for single-carrier BPSK signals
under AWGN channels, particularly for wireless communication systems located in the
medium-wave band, making it suitable for maritime wireless communication. Simulations
showed that the performance of CB-Gardner against Doppler is about three times better
than that of the Gardner algorithm. However, CB-Gardner requires a higher SNR for the
system. The main improvement direction for CB-Gardner may be to continue reducing
self-noise, making it suitable for QPSK and other modulation systems.

Author Contributions: Conceptualization, F.Q.; methodology, F.Q., Z.L. and X.T.; software, Z.L.
and M.Z.; validation, F.Q., Z.L. and M.Z.; formal analysis, F.Q., Z.L. and M.Z.; investigation, Z.L.;
resources, F.Q., X.T. and Y.W.; data curation, Z.L.; writing—original draft preparation, F.Q. and Z.L.;
writing—review and editing, X.T. and M.Z.; visualization, Z.L. and X.T.; supervision, F.Q. and Y.W.;
project administration, Y.W.; funding acquisition, F.Q., Y.W. and X.T. All authors have read and agreed
to the published version of the manuscript.
Funding: This work was supported in part by the National Science Fund for Distinguished Young
Scholars under Grant 62225114 and in part by the National Natural Science Foundation of China
under Grant 62171405 and Grant 62101489.
Institutional Review Board Statement: Not applicable.
Informed Consent Statement: Not applicable.
Data Availability Statement: Code for data analysis is available on request from the authors.
Conflicts of Interest: The authors declare no conflict of interest.

References
1. Mengali, U. Synchronization Techniques for Digital Receivers; Springer Science & Business Media: Berlin, Germany, 1997.
2. Meyr, H.; Moeneclaey, M.; Fechtel, S.A. Digital Communication Receivers: Synchronization, Channel Estimation, and Signal Processing;
Wiley Online Library: New York, NY, USA, 1998; Volume 444.
3. Louveaux, J.; Vandendorpe, L.; Cuvelier, L.; Pollet, T. An early-late timing recovery scheme for filter-bank-based multicarrier
transmission. IEEE Trans. Commun. 2000, 48, 1746–1754. [CrossRef]
4. Mueller, K.; Muller, M. Timing recovery in digital synchronous data receivers. IEEE Trans. Commun. 1976, 24, 516–531. [CrossRef]
5. Classen, F.; Meyr, H.; Sehier, P. Maximum likelihood open loop carrier synchronizer for digital radio. In Proceedings of the
Proceedings of ICC’93-IEEE International Conference on Communications, Geneva, Switzerland, 23–26 May 1993; pp. 493–497.
6. Suzuki, T.; Takatori, H.; Ogawa, M.; Tomooka, K. Line equalizer for a digital subscriber loop employing switched capacitor
technology. IEEE Trans. Commun. 1982, 30, 2074–2082. [CrossRef]
7. Gardner, F. A BPSK/QPSK timing-error detector for sampled receivers. IEEE Trans. Commun. 1986, 34, 423–429. [CrossRef]
8. D’Andrea, A.N.; Luise, M. Optimization of symbol timing recovery for QAM data demodulators. IEEE Trans. Commun. 1996, 44,
399–406. [CrossRef]
9. Leng, W.; Zhang, Y.; Yang, Z. A modified gardner detector for multilevel PAM/QAM system. In Proceedings of the 2008
International Conference on Communications, Circuits and Systems, Fujian, China, 25–27 May 2008; pp. 891–895.
10. Chen, Q.; Li, M. Modified gardner algorithm for bit synchronization in high-order QAM system. In Proceedings of the 2013
International Conference on Computational Problem-Solving (ICCP), Jiuzhai, China, 26–28 October 2013; pp. 427–432.
11. Lim, D. A modified Gardner detector for symbol timing recovery of M-PSK signals. IEEE Trans. Commun. 2004, 52, 1643–1647.
[CrossRef]
12. Zhao, H.; Gong, Y. Demodulation of 8PSK Signal with Gardner Bit Synchronization Using FPGA. In Proceedings of the
Communications, Signal Processing, and Systems: Proceedings of the 2017 International Conference on Communications, Signal
Processing, and Systems, Harbin, China, 14–17 July 2017; pp. 661–668.
13. Azam, M.; Akram, E.; Ahmad, J.; Shah, I. An improved, non-data aided symbol timing recovery for GMSK modulated signals. In
Proceedings of the 7th International Multi Topic Conference—INMIC 2003, Islamabad, Pakistan, 8–9 December 2003; pp. 107–111.
J. Mar. Sci. Eng. 2023, 11, 829 14 of 14

14. Hu, H.; Zhang, C.; Li, Y.; Jia, W. Demodulation of continuous phase 2FSK signal with Gardner synchronization. In Proceedings of
the 2016 6th International Conference on Electronics Information and Emergency Communication (ICEIEC), Beijing, China, 17–19
June 2016; pp. 228–231.
15. Hua, J.; Zhou, L.; Chen, C.; Jiang, L. Synchronization for QDPSK—Costas loop and Gardner algorithm using FPGAs. In
Proceedings of the 2014 IEEE/ACIS 13th International Conference on Computer and Information Science (ICIS), Taiyuan, China,
4–6 June 2014; pp. 27–31.
16. Bazdresch, M.; Al-Hamiri, M. Symbol synchronization of the Alamouti space-time block code with the Gardner algorithm. In
Proceedings of the 2017 8th IEEE Annual Information Technology, Electronics and Mobile Communication Conference (IEMCON),
Vancouver, BC, Canada, 3–5 October 2017; pp. 635–639.
17. D’Andrea, N.A.; Luise, M. Design and analysis of a jitter-free clock recovery scheme for QAM systems. IEEE Trans. Commun.
1993, 41, 1296–1299. [CrossRef]
18. Ogmundson, P.G.; Driessen, P.F. Zero-crossing DPLL bit synchronizer with pattern jitter compensation. IEEE Trans. Commun.
1991, 39, 603–612. [CrossRef]
19. Lankl, B.; Sebald, G. Jitter-reduced digital timing recovery for multilevel PAM and QAM systems. In Proceedings of the
Proceedings of ICC’93-IEEE International Conference on Communications, Geneva, Switzerland, 23–26 May 1993; pp. 804–810.
20. Gappmair, W.; Cioni, S.; Corazza, G.E.; Koudelka, O. Extended Gardner Detector for Improved Symbol-Timing Recovery of
M-PSK Signals. IEEE Trans. Commun. 2006, 54, 1923–1927. [CrossRef]
21. Stojanović, N.; Xie, C.; Zhao, Y.; Mao, B.; Gonzalez, N.G.; Qi, J. Modified Gardner phase detector for Nyquist coherent optical
transmission systems. In Proceedings of the 2013 Optical Fiber Communication Conference and Exposition and the National
Fiber Optic Engineers Conference (OFC/NFOEC), Anaheim, CA, USA, 17–21 March 2013; pp. 1–3.
22. Gappmair, W. Self-noise performance of zero-crossing and Gardner synchronisers applied to one/two-dimensional modulation
schemes. Electron. Lett. 2004, 40, 1–2. [CrossRef]
23. Yu, Q.; Huang, Z.; Ba, J. An Improved Gardner Feedback Timing Synchronization Loop. In Proceedings of the 2022 10th
International Conference on Intelligent Computing and Wireless Optical Communications (ICWOC), Chongqing, China, 10–12
June 2022; pp. 63–67.
24. Yang, D.; Yan, C.; Wang, H.; Kuang, J.; Wu, N. Extension to Gardner timing error detector for QPSK signals. In Proceedings of the
2010 International Conference on Wireless Communications & Signal Processing (WCSP), Suzhou, China, 21–23 October 2010;
pp. 1–5.
25. Zhu, J.; Jiang, Y.; Zhao, K.; Ma, P. An Improved Timing Recovery Method Based on Gardner Algorithm. In Proceedings of the 8th
International Symposium on Test Automation & Instrumentation (ISTAI 2020), Online, 28–29 November 2020; pp. 1–5.
26. Barton, S.; Al-Jalili, Y. A symbol timing recovery scheme based on spectral redundancy. In Proceedings of the IEE Colloquium on
Advanced Modulation and Coding Techniques for Satellite Communications, London, UK, 27 January 1992; pp. 3/1–3/6.
27. Luan, M.; Zhu, H.; Li, M. High-Speed Frequency Domain Bit Synchronization Algorithm for Parallel Implementation. In
Proceedings of the 2018 10th International Conference on Communication Software and Networks (ICCSN), Chengdu, China, 6–9
July 2018; pp. 479–483.
28. Hu, J.; Zhu, L.; Wang, J. The Implementation of High Speed Parallel Timing Synchronization Algorithm Based on FPGA. In
Proceedings of the 2018 10th International Conference on Communication Software and Networks (ICCSN), Chengdu, China, 6–9
July 2018; pp. 484–487.
29. Barletta, L.; Magarini, M.; Scardoni, F.; Spalvieri, A. Impact of loop delay on the performance of Gardner timing recovery. IEEE
Photonics Technol. Lett. 2013, 25, 1797–1800. [CrossRef]
30. Gappmair, W. Open-loop analysis of a nonlinearly modified Gardner synchronizer. IEEE Commun. Lett. 2014, 18, 1963–1966.
[CrossRef]
31. Dorize, C.; Ghazisaeidi, A.; Renaudier, J.; Charlet, G. Performance analysis of nonlinear and gardner timing error detectors with
frequency selective pulse shaping. In Proceedings of the 2015 European Conference on Optical Communication (ECOC), Valencia,
Spain, 27 September–1 October 2015; pp. 1–3.
32. Gardner, F.M. Interpolation in digital modems. I. Fundamentals. IEEE Trans. Commun. 1993, 41, 501–507. [CrossRef]
33. Erup, L.; Gardner, F.M.; Harris, R.A. Interpolation in digital modems. II. Implementation and performance. IEEE Trans. Commun.
1993, 41, 998–1008. [CrossRef]
34. Hsu, F. Square root Kalman filtering for high-speed data received over fading dispersive HF channels. IEEE Trans. Inf. Theory
1982, 28, 753–763. [CrossRef]

Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual
author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to
people or property resulting from any ideas, methods, instructions or products referred to in the content.