IP Library Granted Patent US 12683699
Granted Patent B2
US 12683699 · App. 18/864,009 · Granted Jul 14, 2026

Method for measuring Noise RMS of periodic signal and signal-to-noise ratio of sinusoidal signal

Inventors: Erhan Dai (Jiangsu, CN); Zhi Weng (Jiangsu, CN); Linfei Su (Jiangsu, CN)
Assignee: Nanjing University Of Posts And Telecommunications
H04B17/336G01R29/26
View Patent ↗
Loading inventors, assignments & file history…
Monitor This Case
Get email alerts when status or documents change.
Order Certified Copies
Most orders are placed with the USPTO same day — all within 24 business hours.
Order via The Patent Place →
Pre-filled with this patent's details
Quick Facts
Patent No.
US 12683699
App. No.
18/864,009
Granted
Jul 14, 2026
Kind
B2
Abstract

The present application relates to a method for measuring a Noise RMS of a periodic signal and a signal-to-noise ratio of a sinusoidal signal. The technical points of the method comprise: respectively executing signal noise measurement on a target periodic digital signal and a target periodic analog signal; obtaining a series of zero-crossing points of electrical signals, performing analog-to-digital conversion once at a relative time determined by means of the series of zero-crossing points; and further calculating the noise of the periodic signals by means of a series of analog-to-digital conversion results. The method further comprises: respectively executing signal-to-noise ratio measurement on a target sinusoidal digital signal and a target sinusoidal analog signal; obtaining a series of zero-crossing points of electrical signals; according to each zero-crossing point, calculating the frequency or period of the electric signal; and then performing calculation to obtain a signal-to-noise ratio of a sinusoidal signal.

Claims (514)

1 . A method for measuring a signal-to-noise ratio of a sinusoidal signal, applied in an embedded system, used to obtain a signal-to-noise ratio of a target sinusoidal signal, wherein the target sinusoidal signal is a target sine digital signal, and comprises following steps:

Step A′, sampling the target periodic signal by an analog-to-converter, to obtain a sampled signal, and then entering step B′;

Step B′, obtaining, through an analog means or a digital means, a continuous zero-crossing time series T z1 , . . . , T zk , . . . , T zK of at least two periods on the sampled signal after passing a negative peak value or a positive peak value, wherein K denotes a number of zero-crossing time instants in the continuous zero-crossing time series, and T zk denotes a k-th zero-crossing time instant in the continuous zero-crossing time series; and then initializing j=K+1 and entering Step C′;

Step C′, calculating and obtaining, according to T z(j-1) , . . . , T z(j-m) , an average period T, and further calculating and obtaining, according to a preset parameter c that satisfies a value range (0.5,1), a digital initial integration point time instant T sj corresponding to a j-th zero-crossing time instant T zj through T sj =T z(j-1) +T×c, and then entering Step D′, wherein, 1<m<j, and T zj denotes the j-th zero-crossing time instant;

Step D′, selecting, on a basis of each sampling point on the sampling signal, two sampling points that are adjacent front and back to the digital initial integration point time instant T sj , and sequentially forming a first sampling point to be analyzed, and a second sampling point to be analyzed, forming, by combining a time instant corresponding to the sampling point and a sampling value corresponding to the sampling point, a coordinate of the sampling point to be analyzed, and then entering Step E′;

Step E′, obtaining, according to the digital initial integration point time instant T sj , a sampling value x sj corresponding to T sj and at a position between a first sampling point and a second sampling point through a rectangular interpolation means or a trapezoidal interpolation means, to form a coordinate (T sj , x sj ) of a digital integration starting point, and then initializing v=2, sequentially defining each sampling point after the first sampling point and the second sampling point on the sampling signal as each sampling point to be analyzed, and entering Step F′;

Step F′, executing a digital integration operation on the sampling signal from a coordinate position of the digital integration starting point to a coordinate position of a v-th sampling point to be analyzed, to form a digital integration result S v ; and executing a digital integration operation on the sampling signal from a coordinate position of the digital integration starting point to a coordinate position of a v+1-th sampling point to be analyzed, to form a digital integration result S v+1 , and then entering Step G′;

Step G′, determining whether a product of the digital integration result S v , and the digital integration result S v+1 is greater than 0, updating, in a case of yes, a value for v by adding 1 and returning to Step F′; otherwise, entering Step H′;

Step H′, obtaining, according to a condition that a sum of a digital integration from the digital integration starting point to an digital integration ending point is zero to overcome interference from harmonics or noises, a coordinate (T ej , x ej ) of an digital integration ending point located between the v-th sampling point to be analyzed and the v+1-th sampling point to be analyzed through the rectangular interpolation means or the trapezoidal interpolation means, and then entering Step I′;

Step I′, obtaining, according to

T

zj

=

T

sj

+

T

ej

2

,

 the j-th zero-crossing time instant T zj , further, obtaining, according to T tj =T ej −T sj , an integration time duration T tj corresponding to the j-th zero-crossing time instant T zj , and forming an integration time duration T tj corresponding to a j-th target zero-crossing time instant T zj , and then entering Step J′;

Step J′, obtaining, according to the target zero-crossing time instant T zj , a measured signal period through T p(j-1) =T zj −T z(j-1) , obtaining, through

F

(

j

-

1

)

=

1

T

p

(

j

-

1

)

=

1

T

zj

-

T

z

(

j

-

1

)

,

a tested signal frequency, and then entering Step K′;

Step K′, determining whether a length of a signal zero-crossing time series {T z(K+1) , . . . , T zj } is less than a preset threshold w+1, entering, in case of yes, Step N′;

otherwise, performing a following calculation: selecting w latest generated integration time durations from the signal integration time duration series {T t(K+1) , . . . , T tj }, calculating and acquiring an average integration time duration T ta , and converting the average integration time duration T ta into a count value n corresponding to a sampling period multiple; selecting w latest generated signal periods from a signal period series {T p(K+1) , . . . , T p(j-1) }, calculating and acquiring an average period value T pa , converting the average period value T pa into a count value N corresponding to the sampling period multiple, and then entering Step L′;

Step L′, calculating and obtaining, according to the w latest generated signal periods selected from the signal period series {T p(K+1) , . . . , T p(j-1) }, a period standard deviation σ T , and obtaining, through

σ

R

=

σ

T

T

pa

,

 a parameter σ R ; or, calculating and obtaining, according to w latest generated signal frequencies selected from a signal frequency series {F (K+1) , . . . , F (j-1) }, a frequency standard deviation σ F , obtaining, through

σ

R

=

σ

F

1

/

T

pa

,

 a parameter σ R , and then entering step M′; and

Step M′, calculating and obtaining, according to count values n and N, as well as the parameter σ R , a signal-to-noise ratio SNR of the target sinusoidal signal through

SNR

=

A

2

2

σ

2

,

 wherein A denotes a peak value for the sinusoidal signal and a denotes a Noise RMS for the signal, and then entering Step N′;

Step N′, updating, by adding 1, a value for j, and returning to Step C′.

2 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 1 , wherein in Step C′, according to T z(j-1) , . . . , T z(j-m) , the average period T is calculated and obtained through

T

=

T

z

(

j

-

1

)

-

T

z

(

j

-

m

)

m

-

1

.

3 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 1 , wherein in Step A′, the target periodic signal is sampled by adopting an equal time interval sampling means or an unequal time interval sampling means to obtain the sampled signal; and a whole period sampling is executed on the target periodic signal to obtain the sampled signal.

4 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 1 , wherein in Step F′, the digital integration operation is either a trapezoidal integration or a rectangular integration.

5 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 1 , wherein in Step M′, the step of calculating and obtaining, according to count values n and N, as well as the parameter σ R , a signal-to-noise ratio SNR of the target sinusoidal signal through

SNR

=

A

2

2

σ

2

follows an arbitrary of following four formulas:

SNR

=

A

2

2

σ

2

=

n

{

2

×

σ

R

×

N

×

sin

[

n

N

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

N

×

2

×

(

1

-

c

)

{

2

×

σ

R

×

N

×

sin

[

n

N

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

n

{

2

×

σ

R

×

N

×

sin

[

2

×

(

1

-

c

)

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

N

×

2

×

(

1

-

c

)

{

2

×

σ

R

×

N

×

sin

[

2

×

(

1

-

c

)

×

π

]

}

2

.

6 . A method for measuring a signal-to-noise ratio of a sinusoidal signal, applied in an embedded system, used to obtain a signal-to-noise ratio of a target sinusoidal signal, wherein the target sinusoidal signal is a target sinusoidal analog signal, and comprises following steps:

Step i′, obtaining, through an analog means or a digital means, a continuous zero-crossing time series T z1 , . . . , T zk , . . . , T zK of at least two periods on the sampled signal after passing a negative peak value or a positive peak value, wherein K denotes a number of zero-crossing time instants in the continuous zero-crossing time series, and T zk denotes a k-th zero-crossing time instant in the continuous zero-crossing time series; and then initializing j=K+1 and entering Step ii′;

Step ii′, calculating and obtaining, according to T z(j-1) , . . . , T z(j-m) , an average period T, and further calculating and obtaining, according to a preset parameter c that satisfies a value range (0.5,1), a digital initial integration point time instant T sj corresponding to a j-th zero-crossing time instant T zk through T sj =T z(j-1) +T×c, and then entering Step iii′, wherein 1<m<j, T zk denotes the j-th zero-crossing time instant;

Step iii′, initializing, before the time instant T sj , an integrator, to return an integrator output voltage to zero, and starting an analog integration at a digital integration starting point time instant T sj , terminating, in a case where the integrator output voltage returns to zero again, the integration, and recording an integration ending point time instant T ej , to overcome interference from harmonics or noises, and then entering Step iv′;

Step iv′, obtaining, according to

T

zj

=

T

sj

+

T

ej

2

,

 the j-th zero-crossing time instant T zj , and further obtaining, according to T tj =T ej −T sj , an integration time duration T tj corresponding to the j-th zero-crossing time instant T zj , and forming an integration time duration T tj corresponding to a j-th target zero-crossing time instant T zj , and then entering Step v′;

Step v′, obtaining, according to the target zero-crossing time instant T zj , a measured signal period through T p(j-1) =T zj −T z(j-1) , obtaining, through

F

(

j

-

1

)

=

1

T

p

(

j

-

1

)

=

1

T

zj

-

T

z

(

j

-

1

)

,

 a tested signal frequency, and then entering Step vi′;

Step vi′, determining whether a length of a signal zero-crossing time series {T z(K+1) , . . . , T zj } is less than a preset threshold w+1, entering, in case of yes, Step ix′; otherwise, performing a following calculation: selecting w latest generated integration time durations from the signal integration time duration series {T t(K+1) , . . . , T tj }, calculating and acquiring an average integration time duration T ta , and converting the average integration time duration T ta into a count value n corresponding to a counter period multiple; selecting w latest generated signal periods from a signal period series {T p(K+1) , . . . , T p(j-1) }, calculating and acquiring an average period value T pa , converting the average period value T pa into a count value N corresponding to the counter period multiple, and then entering Step vii′;

Step vii′, calculating and obtaining, according to the w latest generated signal periods selected from the signal period series {T p(K+1) , . . . , T p(j-1) }, a period standard deviation σ T , and obtaining, through

σ

R

=

σ

T

T

pa

,

 a parameter σ R ; or, calculating and obtaining, according to w latest generated signal frequencies selected from a signal frequency series {F (K+1) , . . . , F (j-1) }, a frequency standard deviation, obtaining, through

σ

R

=

σ

F

1

/

T

pa

,

 a parameter σ R , and then entering Step viii′; and

Step viii′, calculating and obtaining, according to count values n and N, as well as the parameter σ R , a signal-to-noise ratio SNR of the target sinusoidal signal through

SNR

=

A

2

2

σ

2

,

 wherein A denotes a peak value for the sinusoidal signal and σ denotes a Noise RMS for the signal, and then entering Step ix′; and

Step ix′, updating, by adding 1, a value for j, and returning to Step ii′.

7 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 6 , wherein, according to T z(j-1) , . . . , T z(j-m) , the average period T is calculated and obtained through

T

=

T

z

(

j

-

1

)

-

T

z

(

j

-

m

)

m

-

1

.

8 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 6 , wherein after respectively obtaining the digital integration starting point time instant T sj and the digital integration ending point time instant T ej , T sj and T ej are respectively calibrated and updated according to circuit delay parameters.

9 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 6 , wherein in Step viii′, the step of calculating and obtaining, according to count values n and N, as well as the parameter σ R , a signal-to-noise ratio SNR of the target sinusoidal signal through

SNR

=

A

2

2

σ

2

follows an arbitrary of following four formulas:

SNR

=

A

2

2

σ

2

=

n

{

2

×

σ

R

×

N

×

sin

[

n

N

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

N

×

2

×

(

1

-

c

)

{

2

×

σ

R

×

N

×

sin

[

n

N

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

n

{

2

×

σ

R

×

N

×

sin

[

2

×

(

1

-

c

)

×

π

]

}

2

SNR

=

A

2

2

σ

2

=

N

×

2

×

(

1

-

c

)

{

2

×

σ

R

×

N

×

sin

[

2

×

(

1

-

c

)

×

π

]

}

2

.

10 . The method for measuring the signal-to-noise ratio of the sinusoidal signal according to claim 6 , wherein in Step iii′, the analog integration is implemented by utilizing an operational amplifier or other devices with integration functions.