# US 2020/0400730 A1 · Frequency spectrum monitoring data structured representation method, and data processing method and compression method

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

- Number: US 2020/0400730 A1
- Document ID: US20200400730A1
- Status: Published application
- Kind: A1
- Application number: 16/979,112
- Filed: 2019-02-21
- Published: 2020-12-24
- Earliest claimed: 2018-03-06
- Assignees: Xi'an Daheng Tiancheng It Co., Ltd.
- Applicants: Xi'an Daheng Tiancheng It Co., Ltd.
- Inventors: Hong YOU; Hongguang MA
- Web page: https://patentgrep.com/patent/US20200400730A1

## CPC

- G01R 23/16: Spectrum analysis; Fourier analysis
- H04N 19/00: Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
- G01R 23/02: Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage
- G06F 16/1744: {using compression, e.g. sparse files}
- G06F 2216/03: Data mining
- H04N 19/61: in combination with predictive coding
- G01R: MEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES (indicating correct tuning of resonant circuits H03J3/12)
- H04N: PICTORIAL COMMUNICATION, e.g. TELEVISION
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Provided is a frequency spectrum monitoring data structured representation method, comprising the following steps: discretizing frequency spectrum monitoring data of a single station in a time dimension and a frequency spectrum dimension to form a two-dimensional frequency spectrum matrix; and arranging frequency spectrum matrices, obtained by all the stations in a certain region, according to a certain rule by taking a certain point as the center, so as to construct a three-dimensional frequency spectrum matrix body. The positive effects of the present invention are that: four dimensions, i.e. time, frequency spectrum, space and energy, of frequency spectrum data can be associated to form a structured organization system meeting frequency spectrum monitoring data; a mathematical operation can be further defined for a frequency spectrum matrix and a frequency spectrum matrix body, thereby conveniently performing information mining on massive monitoring data; and subsequent compression processing and remote transmission of the monitoring data can be facilitated, and requirements of frequency spectrum monitoring station network system construction and big data processing are better met, so that the frequency spectrum monitoring data is utilized more rationally and efficiently.

## Claims

### Claim 1

A structured representation method of spectrum monitoring data, characterized in that, comprising the following steps of:
S1: discretizing spectrum monitoring data of a single station in time dimension and frequency dimension to form a two-dimensional spectrum matrix;
S2: constructing a three-dimensional spectrum matrix by using the two-dimensional spectrum matrix obtained from all stations in a certain area, taking a certain point as a center, and aligning in a position dimension according to certain rules.

### Claim 2 (depends on 1)

The structured representation method of spectrum monitoring data according to claim 1, characterized in that, in the step S1, the two-dimensional spectrum matrix is formed by the following steps of:
S1.1: the spectrum monitoring station, according to a synchronization or calibration method, obtaining a synchronized clock with other stations, then obtaining a uniformly specified sampling time tn, spectrum bandwidth B and frequency sampling points M, where n is 1, 2, 3, . . . , N, N is a positive integer;
S1.2: at the sampling time tn, using the spectrum bandwidth B and the number of frequency sampling points M to discretize the monitoring data in the frequency dimension to obtain a vector Itn, and a sequence Itn is 1×M dimension;
S1.3: for a given monitoring time period, at the sampling time tn, obtaining spectrum monitoring vectors at different sampling time It1, It2, . . . , ItN;
S1.4: aligning the spectrum monitoring vectors at different sampling time in chronological order to form a two-dimensional spectrum matrix W=[It1 It2 . . . ItN]T, the matrix W is N×M dimension.

### Claim 3 (depends on 2)

The structured representation method of spectrum monitoring data according to claim 2, characterized in that, in the step S1, the two-dimensional spectrum matrix defines the following mathematical operations:
assume the two-dimensional frequency spectrum matrix is Wi, Wj, both of which are N×M dimension,
W
i
=
(
x
11
…
x
1

N
⋮
⋱
⋮
x
M


1
…
x
NM
)
,
W
j
=
(
y
11
…
y
1

N
⋮
⋱
⋮
y
M


1
…
y
NM
)
,
then
Spectrum matrix addition:
W
i
+
W
j
=
(
x
11
+
y
11
…
x
1

N
+
y
1

N
⋮
⋱
⋮
x
M


1
+
y
M


1
…
x
NM
+
y
NM
)
Spectrum matrix subtraction:
W
i
-
W
j
=
(
x
11
-
y
11
…
x
1

N
-
y
1

N
⋮
⋱
⋮
x
M


1
-
y
M


1
…
x
NM
-
y
NM
)
Frequency dimension projection of spectrum matrix:
FProj

(
W
)
=
(
∑
i
=
1
N

x
1

i
∑
i
=
1
N

x
2

i
⋮
∑
i
=
1
N

x
M

i
)
=
(
Z
1
Z
2
⋮
Z
M
)
Time dimension projection of spectrum matrix:
TProj

(
W
)
=
(
∑
i
=
1
M

x
i

1
∑
i
=
1
M

x
i

2
…

∑
i
=
1
M

x
i

N
)
=
(
L
1

L
2
…
L
N
)
m is a real number, then multiply the spectrum matrix number:
m

W
=
(
m

x
1

1
…
m

x
1

N
⋮
⋱
⋮
m

x
M

1
…
m

x
N

M
)

### Claim 4 (depends on 1)

The structured representation method of spectrum monitoring data according to claim 1, characterized in that, in the step S2, the three-dimensional spectrum matrix is constructed by the following steps:
S2.1: obtaining the monitoring data of all K spectrum monitoring stations in a certain area within a given time period, which are W1, W2, . . . WK;
S2.2: According to the relationship between the positions of the K number of spectrum monitoring stations, arrange W1, W2, . . . WK to form a three-dimensional spectrum matrix body Q=[W1, W2, . . . WK]T, the matrix body Q is N×M×K dimension.

### Claim 5 (depends on 1)

The structured representation method of spectrum monitoring data according to claim 1, characterized in that, in the step S2.2, the spectrum matrix body Q formed by the spectrum matrix W1, W2, . . . WK of the K number of spectrum monitoring stations employs the following steps and rules:
S2.2.1: Calculate a geometric center point V0 on a geographical distribution according to latitude and longitude positions Vn of K number of spectrum monitoring stations, where n takes the value 1, 2, 3, . . . , K;
V
0
=
1
K

∑
i
=
1
K

V
i
S2.2.2: Calculate the di stance Dn between the latitude and longitude position Vn of each spectrum monitoring station and the geometric center point V0, where ∥·∥2 is the second-order norm operation:
Dn=∥Vn−V0∥2
S2.2.3: according to the order from small to large of the distance Dn between the spectrum monitoring station and the geometric center point V0, arrange the corresponding spectrum matrix of the station to construct the spectrum matrix body Q=[W1, W2, . . . WK]T.

### Claim 6 (depends on 4)

The structured representation method of spectrum monitoring data according to claim 4, characterized in that, in the step S2, the three-dimensional frequency spectrum matrix defines the following mathematical operations:
Assuming that the spectrum moments are Qi and Qj, both of which are N×M×K dimensions, Qi=[W1i, W2i, . . . WKi]T, Qj=[W1j, W2j, . . . WKj]T, then
Spectrum matrix addition:
Qi+Qj=[W1i+W1j,W2i+W2j, . . . WKi+WKj]T
Spectrum matrix subtraction:
Qi−Qj;=[W1i−W1j,W2i−W2j, . . . WKi−WKj]T
Frequency dimension projection of spectrum matrix:
FProj(Q)=[FProj(W1),FProj(W2), . . . FProj(WK)]T
Time dimension projection of spectrum matrix:
TProj(Q)=[TProj(W1),TProj(W2), . . . TProj(WK)]T
m is a real number, then multiply the spectrum matrix number:
mQ=[mW1,mW2, . . . mWK]T.

### Claim 7

A compression processing method for spectrum monitoring data of multiple stations, characterized in that, comprising the following steps of:
S1: structured representation of spectrum monitoring data of single station: according to a unified monitoring frequency bandwidth, monitoring a sampling interval of time dimension and a sampling interval of frequency dimension, completing discretization of the frequency dimension and discretization of the time dimension of the monitoring data in a given time period at a given time and a given bandwidth, expressing the spectrum monitoring data of the given monitoring time period and the frequency bandwidth as a spectrum matrix;
S2: structured representation of spectrum monitoring data of multiple stations: arranging all the spectrum matrices obtained by the multiple stations in a given area into a spectrum matrix body according to specific rules of the station position;
S3: Gray-scale processing of spectrum monitoring data: selecting a dimension from the three dimensions of the spectrum matrix body, where the spectrum matrix body can be regarded as an arrangement of a series of matrices in the selected dimension, transforming each of the matrices into a grayscale image by gray-scale processing such that the spectrum matrix body can be regarded as a piece of video data;
S4: Utilization of video compression method to compress monitoring data: for the monitoring data of multiple stations in a given area, there is a large amount of redundant information in the dimensions of time, frequency, and position, compressing the video data corresponding to the spectrum matrix body by using traditional video compression standards.

### Claim 8 (depends on 1)

The compression processing method for spectrum monitoring data of multiple stations according to claim 1, in the step S2, rules for arranging the spectrum matrices into the spectrum matrix body adopt one of the following three methods:
Selecting a monitoring station as a reference station, and arranging the spectrum matrix data corresponding to each station based on a geographic straight-line distance between all stations and the reference point; or
Selecting a geographic position corresponding to the average longitude and latitude of geographic locations of multiple stations in a certain area as a reference point, and arranging the spectrum matrix data corresponding to each station based on a geographic straight-line distance between all stations and the reference point; or
Selecting one particular geographic position as a reference point, and arranging the spectrum matrix data corresponding to each station based on a geographic straight-line distance between all stations and the reference point.

### Claim 9 (depends on 1)

The compression processing method for spectrum monitoring data of multiple stations according to claim 1, in the step S3, for the steps of the gray-scale processing of the spectrum monitoring data, selecting one of the following three methods:
selecting the time dimension as the reference dimension such that the spectrum matrix body can be regarded as data arranged by a series of two-dimensional matrices corresponding to each time point; carrying out grayscale processing on each matrix to convert each matrix into a grayscale image such that in the time dimension, the spectrum matrix body can be regarded as a piece of video data; or
selecting the frequency dimension as the reference dimension such that the spectrum matrix body can be regarded as data arranged by a series of two-dimensional matrices corresponding to each time point; carrying out grayscale processing on each matrix to convert each matrix into a grayscale image such that in the time dimension, the spectrum matrix body can be regarded as a piece of video data; or
selecting the position dimension as the reference dimension such that the spectrum matrix body can be regarded as data arranged by a series of two-dimensional matrices corresponding to each time point; carrying out grayscale processing on each matrix to convert each matrix into a grayscale image such that in the time dimension, the spectrum matrix body can be regarded as a piece of video data.

### Claim 10

A processing method of spectrum monitoring data, characterized in that, comprising:
Arrange the two-dimensional spectrum matrices of multiple stations to form a three-dimensional spectrum matrix body according to inter-relationship between positions of the multiple stations,
wherein the two-dimensional spectrum matrix of each station is formed by the discretization of spectrum monitoring data of the station in time and frequency dimensions.

### Claim 11 (depends on 10)

A processing method of spectrum monitoring data according to claim 10, characterized in that, the two-dimensional spectrum matrix of each station is formed by the following steps:
S1.1: obtaining a synchronized clock of one station with other stations, then obtaining a uniformly specified sampling time tn (1 . . . N), spectrum bandwidth B and frequency sampling points M;
S1.2: at the sampling time tn, using the spectrum bandwidth B and the frequency sampling points M to discretize the monitoring data in the frequency dimension to obtain a vector Itn, and a sequence Itn is 1×M dimension;
S1.3: for a given monitoring time period, at the sampling time tn, obtaining spectrum monitoring vectors at different sampling time It1, It2, . . . , ItN;
S1.4: aligning the spectrum monitoring vectors at different sampling time in chronological order to form a two-dimensional spectrum matrix W=[It1 It2 . . . ItN]T, the matrix W is N×M dimension.

### Claim 12 (depends on 11)

The processing method of spectrum monitoring data according to claim 11, characterized in that,
when the sum of the spectrum monitoring data of a station is required to be obtained, the two-dimensional spectrum matrix of the multiple stations is processed as follows:
W
i
+
W
j
=
(
x
11
+
y
11
…
x
1

N
+
y
1

N
⋮
⋱
⋮
x
M


1
+
y
M


1
…
x
NM
+
y
NM
)
when the difference of the spectrum monitoring data of a station is required to be obtained, the two-dimensional spectrum matrix of the multiple stations is processed as follows:
W
i
+
W
j
=
(
x
11
-
y
11
…
x
1

N
-
y
1

N
⋮
⋱
⋮
x
M


1
-
y
M


1
…
x
NM
-
y
NM
)
wherein
,

W
i
=
(
x
11
…
x
1

N
⋮
⋱
⋮
x
M


1
…
x
NM
)
,
W
j
=
(
y
11
…
y
1

N
⋮
⋱
⋮
y
M


1
…
y
NM
)
,
when the usage condition of the spectrum monitoring data of a station in a certain period of time is required, the two-dimensional spectrum matrix of the station is processed as follows:
TProj

(
W
)
=
(
∑
i
=
1
M

x
i

1
∑
i
=
1
M

x
i

2
…
∑
i
=
1
M

x
i

N
)
=
(
L
1

L
2
…
L
N
)
when the spectrum evaluation usage of the spectrum monitoring data of a station in a certain frequency band is required, the two-dimensional spectrum matrix of the station is processed as follows:
FProj

(
W
)
=
(
∑
i
=
1
N

x
1

i
∑
i
=
1
N

x
2

i
⋮
∑
i
=
1
N

x
M

i
)
=
(
Z
1
Z
2
⋮
Z
M
)
when the multiplication data of the spectrum monitoring data of a station is required, the two-dimensional spectrum matrix of the station is processed as follows:
m


W
=
(
m

x
1

1
…
m

x
1

N
⋮
⋱
⋮
m

x
M

1
…
m

x
N

M
)
wherein, m is a real number.

### Claim 13 (depends on 12)

The processing method of spectrum monitoring data according to claim 12, characterized in that, a geometric center point V0 on a geographical distribution of K number of spectrum monitoring stations in an area is obtained by the following step:
V
0
=
1
K

∑
i
=
1
K

V
i
where n takes the value 1, 2, 3, . . . , K; Vi refers to the latitude and longitude positions of the i-th number station.

### Claim 14 (depends on 13)

The processing method of spectrum monitoring data according to claim 13, characterized in that, a distance Dn between the latitude and longitude position Vn of each spectrum monitoring station and the geometric center point V0 is obtained by the following step:
Dn=∥Vn−V0∥2
where ∥·∥2 is the second-order norm operation.

### Claim 15 (depends on 14)

The processing method of spectrum monitoring data according to claim 14, characterized in that, the three-dimensional frequency spectrum matrix body Q is formed by the following steps:
according to the order from small to large of the distance Dn between the spectrum monitoring station and the geometric center point V0, arrange the corresponding spectrum matrix of the station to construct the three-dimensional spectrum matrix body Q=[W1, W2, . . . WK]T, the spectrum matrix body Q is in N×M×K dimension, W1, W2, . . . WK refers to the two-dimensional spectrum matrix of K stations in an area within a given time period.

### Claim 16 (depends on 15)

The processing method of spectrum monitoring data according to claim 15, characterized in that,
when the sum of the spectrum monitoring data of the stations in a multiple area is required to be obtained, the three-dimensional spectrum matrix of the stations in each area is processed as follows:
Qi+Qj=[W1i+W1j,W2i+W2j, . . . WKi+WKj]T
when the difference of the spectrum monitoring data of stations in two area is required to be obtained, the three-dimensional spectrum matrix of the stations in two area is processed as follows:
Qi−Qj=[W1i−W1j,W2i−W2j, . . . WKi−WKj]T
when the spectrum evaluation usage of the spectrum monitoring data of a station in a certain frequency band is required, the two-dimensional spectrum matrix of the stations in the area is processed as follows:
TProj(Q)=[TProj(W1),TProj(W2), . . . TProj(WK)]T
when the spectrum evaluation usage of the spectrum monitoring data of the stations of the area in a certain frequency band is required, the three-dimensional spectrum matrix of the stations in the area is processed as follows:
FProj(Q)=[FProj(W1),FProj(W2), . . . FProj(WK)]T
when the multiplication data of the spectrum monitoring data of the stations of the area in a certain frequency band is required, the three-dimensional spectrum matrix of the stations in the area is processed as follows:
mQ=[mW1,mW2, . . . mWK]T
where Qi and Qj are the three-dimensional spectrum matrix of stations in an area, both are N×M×K dimensions, Qi=[W1i+W2i, . . . WKi]T, Qj=[W1j, Q2j, . . . WKj]T.

### Claim 17

A compression processing method for spectrum monitoring data of multiple stations, characterized in that, comprising:
S1: arrange the two-dimensional spectrum matrices of multiple stations to form a three-dimensional spectrum matrix body according to the mutual relationship between the positions of the stations, wherein the two-dimensional spectrum matrix of each station is formed by the discretization of the spectrum monitoring data of the station in the time and frequency dimensions;
S2: Select one dimension from the three dimensions of the three-dimensional frequency spectrum matrix body, carrying out grayscale processing of each matrix to convert to grayscale image;
S3: compress the spectrum monitoring data in the selected dimension by using video compression method.

### Claim 18 (depends on 17)

The compression processing method for spectrum monitoring data of multiple stations according to claim 17, in the step S1, the two-dimensional matrix is arranged as follows:
Selecting a station as a reference station, and arranging the spectrum matrix data corresponding to each station based on a standard of geographic straight-line distance between all stations and the reference point; or
Selecting a geographic position corresponding to the average longitude and latitude of geographic locations of multiple stations in a certain area as a reference point, and arranging the spectrum matrix data corresponding to each station based on a standard of geographic straight-line distance between all stations and the reference point; or
Selecting one particular geographic position as a reference point, and arranging the spectrum matrix data corresponding to each station based on a standard of a geographic straight-line distance between all stations and the reference point.

### Claim 19 (depends on 17)

The compression processing method for spectrum monitoring data of multiple stations according to claim 17, in the step S2, the steps of gray-scale processing of the spectrum monitoring data is selected from one of the following three method:
selecting the time dimension as the reference dimension and carrying out grayscale processing on each matrix to convert each matrix into a grayscale image; or
selecting the frequency dimension as the reference dimension and carrying out grayscale processing on each matrix to convert each matrix into a grayscale image; or
selecting the position dimension as the reference dimension and carrying out grayscale processing on each matrix to convert each matrix into a grayscale image.