# US 2017/0257248 A1 · Systems and methods for frequency synchronization between transmitters and receivers in a communication system

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2017/0257248 A1
- Document ID: US20170257248A1
- Status: Published application
- Kind: A1
- Application number: 15/063,085
- Filed: 2016-03-07
- Published: 2017-09-07
- Earliest claimed: 2016-03-07
- Assignees: Hong Kong Applied Science and Technology Research Institute Company Limited
- Applicants: Hong Kong Applied Science and Technology Research Institute Co., Ltd.
- Inventors: Shaohua Zhao; Wuxue Ni; Biqiang Luo; Li Li
- Web page: https://patentgrep.com/patent/US20170257248A1

## CPC

- H04L 27/2663: {Coarse synchronisation, e.g. by correlation}
- H04L 27/06: Demodulator circuits; Receiver circuits
- H04L 27/14: Demodulator circuits; Receiver circuits
- H04L 27/2331: {wherein the received signal is demodulated using one or more delayed versions of itself}
- H04L 27/2671: {Time domain}
- H04L 27/2675: {Pilot or known symbols}
- H04L 7/04: Speed or phase control by synchronisation signals {(H04L7/0075 takes precedence)}
- H04L: TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION (arrangements common to telegraphic and telephonic communication H04M)

## Abstract

Systems and methods which provide training sequence or preamble-based synchronization with respect to non-coherent modulated signals and/or differentially coherent modulated signals are described. Embodiments provide for time synchronization using a technique for mitigating the effect of carrier frequency offset (CFO) with respect to the received signal. Embodiments of the present invention provide for frequency synchronization using a technique for estimating CFO using a constant bias induced with respect to the received signal by CFO. Additionally or alternatively, embodiments of the present invention provide for frequency synchronization using a technique for estimating CFO using phase rotation caused by CFO. The time synchronization and frequency synchronization of embodiments may be performed independently, without requiring the results of one synchronization operation for performing the other synchronization operation.

## Claims

### Claim 1

A method for frequency synchronization of packet-based communications, the method comprising:
calculating a first-order differential of a received training sequence signal, wherein the training sequence is non-coherently modulated or differentially coherently modulated;
summing over the first-order differential of the received training sequence signal;
subtracting a summation derived from a known training sequence instance corresponding to the received training sequence from the summation over the first-order differential of the received training sequence signal; and
determining the frequency synchronization based upon the result of the subtracting.

### Claim 2 (depends on 1)

The method of claim 1, wherein a result of the calculating the first-order differential includes a constant bias induced by carrier frequency offset (CFO) in the received training sequence signal and the subtracting the summation derived from the known training sequence instance provides an estimation of the constant bias CFO in the received training sequence signal.

### Claim 3 (depends on 2)

The method of claim 2, wherein the summing over the first-order differential of the received training sequence signal provides
2

π


hT
s

∑
k
=
N
p
N
p
+
NM
-
1

x
~

(
k
)
+
∑
k
=
N
p
N
p
+
NM
-
1

2

π


f
Δ

T
s
+
∑
k
=
N
p
N
p
+
NM
-
1

w
0

(
k
)
and the summation derived from the known training sequence instance provides
S
0
=
2

π


hT
s

∑
k
=
N
p
N
p
+
NM
-
1

x
~

(
k
)
,
and wherein CFO is estimated as
2

π


f
Δ

T
s
=
S
-
S
0
NM
.

### Claim 4 (depends on 1)

The method of claim 1, wherein the calculating the first-order differential is performed as part of calculating a second-order differential of the received training sequence signal for time synchronization of the packet-based communications.

### Claim 5 (depends on 1)

The method of claim 1, further comprising:
providing time synchronization of the packet-based communications.

### Claim 6 (depends on 5)

The method of claim 5, wherein the time synchronization is provided independently with respect to the frequency synchronization, whereby results of one synchronization are not utilized in performing another synchronization operation.

### Claim 7 (depends on 5)

The method of claim 5, wherein the providing time synchronization comprises:
calculating a second-order differential of the received training sequence signal;
calculating a cross-correlation between a second-order differential vector provided by the second-order differential calculation and a vector derived from a known training sequence instance corresponding to the received training sequence; and
determining the time synchronization based upon a peak provided by the cross-correlation.

### Claim 8 (depends on 1)

The method of claim 1, wherein the calculating the cross-correlation comprises:
calculating the cross-correlation along a moving widow, wherein the peak provided by the cross-correlation comprises a peak identified within the moving window.

### Claim 9 (depends on 8)

The method of claim 8, wherein the second-order differential vector is represented by D(k)=[{circumflex over (d)}(k−NM−1), {circumflex over (d)}(k−NM−2), . . . {circumflex over (d)}(k)], the vector derived from the known training sequence is represented by C=[c(0), c(1), . . . c(NM−1)], c(i)ε{+1,−1}, and the moving window in which the cross-correlation is performed is represented by [−M−1, M−1], and wherein the cross-correlation is performed as
R

(
k
)
=
D

(
k
)

C
=
∑
r
=
k
-
NM
-
1
k

d
^

(
r
)

c

(
r
)
.

### Claim 10 (depends on 8)

The method of claim 8, wherein the peak provided by the cross-correlation comprises a greatest cross-correlation in the moving window determined as
k
ST
=
argmax
k
∈
[
-
M
-
1
,
M
-
1
]

(
D

(
k
)
)
.

### Claim 11 (depends on 7)

The method of claim 7, wherein the calculating the cross-correlation comprises:
establishing an index for a moving window cross-correlation;
calculating the cross-correlation of the second-order differential vector and the vector derived from the known training sequence instance for a current index point of the moving window;
comparing the cross-correlation for the current moving window index point with a stored moving window peak correlation value;
determining whether the stored moving window peak is smaller than the cross-correlation for the current index point in the moving window and, if the stored moving window peak is smaller, replacing the stored moving window peak with the cross-correlation for the current index point;
incrementing the moving window index; and
repeating the calculating the cross-correlation, the comparing the cross-correlation, and the determining whether the stored moving window peak is smaller until an end of the moving widow is reached.

### Claim 12 (depends on 11)

The method of claim 11, wherein the establishing an index comprises setting N0=Pcur−(N−1)M, l=−(M−1), P=0, the calculating the cross-correlation comprises calculating NM point cross-correlation as
R

(
l
)
=
∑
r
=
N
0
-
l
-
NM
-
1
N
0
-
l

d
^

(
r
)

c

(
r
)
,
the comparing the cross-correlation comprises comparing P with R(l)), the replacing the stored moving window peak comprises replacing P by R(l), and the incrementing the moving window index comprises setting l=l+1.

### Claim 13

A system for frequency synchronization of packet-based communications, the system comprising:
a first-order differentiator circuit operable to calculate a first-order differential of a received training sequence signal, wherein the training sequence is non-coherently modulated or differentially coherently modulated;
a summation circuit operable to provide a summation over the first-order differential of the received training sequence signal, subtract a summation derived from a known training sequence instance corresponding to the received training sequence from the summation over the first-order differential of the received training sequence signal, and provide a result for use in determining the frequency synchronization.

### Claim 14 (depends on 13)

The system of claim 13, wherein a result of the first-order differential calculation includes a constant bias induced by carrier frequency offset (CFO) in the received training sequence signal and a result of subtraction of the summation derived from the known training sequence instance provides an estimation of the constant bias CFO in the received training sequence signal.

### Claim 15 (depends on 14)

The system of claim 14, wherein the summation over the first-order differential of the received training sequence signal provides
2

π


hT
s

∑
k
=
N
p
N
p
+
NM
-
1

x
~

(
k
)
+
∑
k
=
N
p
N
p
+
NM
-
1

2

π


f
Δ

T
s
+
∑
k
=
N
p
N
p
+
NM
-
1

w
0

(
k
)
and the summation derived from the known training sequence instance provides
S
0
=
2

π


hT
s

∑
k
=
N
p
N
p
+
NM
-
1

x
~

(
k
)
,
and wherein CFO is estimated as
2

π


f
Δ

T
s
=
S
-
S
0
NM
.

### Claim 16 (depends on 13)

The system of claim 13, wherein the first-order differential is calculated as part of a calculation of a second-order differential of the received training sequence signal for time synchronization of the packet-based communications.

### Claim 17 (depends on 13)

The system of claim 13, further comprising:
time synchronization circuitry operable to provide time synchronization of the packet-based communications.

### Claim 18 (depends on 17)

The system of claim 17, wherein the time synchronization is provided independently with respect to the frequency synchronization, whereby results of one synchronization are not utilized in performing another synchronization operation.

### Claim 19 (depends on 17)

The system of claim 17, wherein the time synchronization circuitry comprises:
a differentiator circuit operable to calculate a second-order differential of a received training sequence signal, wherein the training sequence is non-coherently modulated or differentially coherently modulated;
a cross-correlator circuit coupled to an output of the differentiator circuit and operable to calculate a cross-correlation between a second-order differential vector provided by the second-order differential calculation circuit and a vector derived from a known training sequence instance corresponding to the received training sequence and to identify a peak provided by the cross-correlation for use in determining the time synchronization.

### Claim 20 (depends on 13)

The system of claim 13, wherein the cross-correlator circuit is operable to calculate the cross-correlation along a moving widow, wherein the peak provided by the cross-correlation comprises a peak identified within the moving window.

### Claim 21 (depends on 20)

The system of claim 20, wherein the second-order differential vector is represented by D(k)=[{circumflex over (d)}(k−NM−1), {circumflex over (d)}(k−NM−2), . . . {circumflex over (d)}(k)], the vector derived from the known training sequence is represented by C=[c(0), c(1), . . . c(NM−1)], c(i)ε{+1,−1}, and the moving window in which the cross-correlation is performed is represented by [−M−1, M−1], and wherein the cross-correlation performed by the cross-correlator circuit is performed in accordance with
R

(
k
)
=
D

(
k
)

C
=
∑
r
=
k
-
NM
-
1
k

d
^

(
r
)

c

(
r
)
.

### Claim 22 (depends on 21)

The system of claim 21, wherein the peak provided by the cross-correlator circuit comprises a greatest cross-correlation in the moving window determined in accordance with
k
ST
=
argmax
k
∈
[
-
M
-
1
,
M
-
1
]

(
D

(
k
)
)
..

### Claim 23 (depends on 13)

The system of claim 13, wherein the cross-correlator circuit is operable to establish an index for a moving window cross-correlation, calculate the cross-correlation of the second-order differential vector and the vector derived from the known training sequence instance for a current index point of the moving window, compare the cross-correlation for the current moving window index point with a stored moving window peak correlation value, determine whether the stored moving window peak is smaller than the cross-correlation for the current index point in the moving window and, if the stored moving window peak is smaller, replace the stored moving window peak with the cross-correlation for the current index point, increment the moving window index, and repeat the calculating the cross-correlation, the comparing the cross-correlation, and the determining whether the stored moving window peak is smaller until an end of the moving widow is reached.

### Claim 24 (depends on 23)

The system of claim 23, wherein the cross-correlator circuit is operable to establish index by setting N0=Pcur−(N−1)M, l=−(M−1), P=0, calculate the cross-correlation by calculating NM point cross-correlation as
R

(
l
)
=
∑
r
=
N
0
-
l
-
NM
-
1
N
0
-
l

d
^

(
r
)

c

(
r
)
,
compare the cross-correlation by comparing P with R(l)), replace the stored moving window peak by replacing P by R(l), and increment the moving window index by setting l=l+1.

### Claim 25

A method for frequency synchronization of packet-based communications, the method comprising:
calculating an autocorrelation of a received training sequence signal and a delayed version of the received training sequence, wherein the training sequence is non-coherently modulated or differentially coherently modulated, and wherein the training sequence is a repetitive pattern;
estimating carrier frequency offset (CFO) from results of the autocorrelation; and
determining the frequency synchronization based upon the CFO estimate.

### Claim 26 (depends on 25)

The method of claim 25, the calculating the autocorrelation comprises:
calculating an autocorrelation of a vector of sampling points of the received training sequence signal and a delayed version of the vector, wherein the delayed version of the vector is delayed with interval Δt.

### Claim 27 (depends on 26)

The method of claim 26, wherein the calculating the autocorrelation further comprises:
calculating a product (R0) between a currently input sampling point and a previous sampling point;
calculating a product (R1) between corresponding sampling points of a delayed version of the received training sequence vector; and
calculating an autocorrelation of the foregoing products (S=S+R0−R1).

### Claim 28 (depends on 25)

The method of claim 25, further comprising:
providing time synchronization of the packet-based communications.

### Claim 29 (depends on 28)

The method of claim 28, wherein the time synchronization is provided independently with respect to the frequency synchronization, whereby results of one synchronization are not utilized in performing another synchronization operation.

### Claim 30 (depends on 28)

The method of claim 28, wherein the providing time synchronization comprises:
calculating a second-order differential of the received training sequence signal;
calculating a cross-correlation between a second-order differential vector provided by the second-order differential calculation and a vector derived from a known training sequence instance corresponding to the received training sequence; and
determining the time synchronization based upon a peak provided by the cross-correlation.

### Claim 31 (depends on 25)

The method of claim 25, wherein the calculating the cross-correlation comprises:
calculating the cross-correlation along a moving widow, wherein the peak provided by the cross-correlation comprises a peak identified within the moving window.

### Claim 32 (depends on 31)

The method of claim 31, wherein the second-order differential vector is represented by D(k)=[{circumflex over (d)}(k−NM−1), {circumflex over (d)}(k−NM−2), . . . {circumflex over (d)}(k)], the vector derived from the known training sequence is represented by C=[c(0), c(1), . . . c(NM−1)], c(i)ε{+1,−1}, and the moving window in which the cross-correlation is performed is represented by [−M−1, M−1], and wherein the cross-correlation is performed as
R

(
k
)
=
D

(
k
)

C
=
∑
r
=
k
-
NM
-
1
k

d
^

(
r
)

c

(
r
)
.

### Claim 33 (depends on 31)

The method of claim 31, wherein the peak provided by the cross-correlation comprises a greatest cross-correlation in the moving window determined as
k
ST
=
argmax
k
∈
[
-
M
-
1
,
M
-
1
]

(
D

(
k
)
)
..

### Claim 34 (depends on 30)

The method of claim 30, wherein the calculating the cross-correlation comprises:
establishing an index for a moving window cross-correlation;
calculating the cross-correlation of the second-order differential vector and the vector derived from the known training sequence instance for a current index point of the moving window;
comparing the cross-correlation for the current moving window index point with a stored moving window peak correlation value;
determining whether the stored moving window peak is smaller than the cross-correlation for the current index point in the moving window and, if the stored moving window peak is smaller, replacing the stored moving window peak with the cross-correlation for the current index point;
incrementing the moving window index; and
repeating the calculating the cross-correlation, the comparing the cross-correlation, and the determining whether the stored moving window peak is smaller until an end of the moving widow is reached.

### Claim 35 (depends on 34)

The method of claim 34, wherein the establishing an index comprises setting N0=Pcur−(N−1)M, l=−(M−1), P=0, the calculating the cross-correlation comprises calculating NM point cross-correlation as
R

(
l
)
=
∑
r
=
N
0
-
l
-
NM
-
1
N
0
-
l

d
^

(
r
)

c

(
r
)
,
the comparing the cross-correlation comprises comparing P with R(l)), the replacing the stored moving window peak comprises replacing P by R(l), and the incrementing the moving window index comprises setting l=l+1.
