WO2019170001A1 - 一种频谱监测数据结构化表示方法、数据处理方法和压缩方法 - Google Patents

一种频谱监测数据结构化表示方法、数据处理方法和压缩方法 Download PDF

Info

Publication number
WO2019170001A1
WO2019170001A1 PCT/CN2019/075666 CN2019075666W WO2019170001A1 WO 2019170001 A1 WO2019170001 A1 WO 2019170001A1 CN 2019075666 W CN2019075666 W CN 2019075666W WO 2019170001 A1 WO2019170001 A1 WO 2019170001A1
Authority
WO
WIPO (PCT)
Prior art keywords
spectrum
matrix
station
monitoring data
dimension
Prior art date
Application number
PCT/CN2019/075666
Other languages
English (en)
French (fr)
Inventor
游鸿
马红光
Original Assignee
西安大衡天成信息科技有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Priority claimed from CN201810181415.1A external-priority patent/CN108616720B/zh
Priority claimed from CN201810181414.7A external-priority patent/CN108509506B/zh
Application filed by 西安大衡天成信息科技有限公司 filed Critical 西安大衡天成信息科技有限公司
Priority to US16/979,112 priority Critical patent/US20200400730A1/en
Publication of WO2019170001A1 publication Critical patent/WO2019170001A1/zh

Links

Images

Classifications

    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/02Arrangements for measuring frequency, e.g. pulse repetition rate; Arrangements for measuring period of current or voltage
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F16/00Information retrieval; Database structures therefor; File system structures therefor
    • G06F16/10File systems; File servers
    • G06F16/17Details of further file system functions
    • G06F16/174Redundancy elimination performed by the file system
    • G06F16/1744Redundancy elimination performed by the file system using compression, e.g. sparse files
    • GPHYSICS
    • G01MEASURING; TESTING
    • G01RMEASURING ELECTRIC VARIABLES; MEASURING MAGNETIC VARIABLES
    • G01R23/00Arrangements for measuring frequencies; Arrangements for analysing frequency spectra
    • G01R23/16Spectrum analysis; Fourier analysis
    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06FELECTRIC DIGITAL DATA PROCESSING
    • G06F2216/00Indexing scheme relating to additional aspects of information retrieval not explicitly covered by G06F16/00 and subgroups
    • G06F2216/03Data mining
    • HELECTRICITY
    • H04ELECTRIC COMMUNICATION TECHNIQUE
    • H04NPICTORIAL COMMUNICATION, e.g. TELEVISION
    • H04N19/00Methods or arrangements for coding, decoding, compressing or decompressing digital video signals
    • H04N19/60Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding
    • H04N19/61Methods or arrangements for coding, decoding, compressing or decompressing digital video signals using transform coding in combination with predictive coding

Definitions

  • the invention belongs to the technical field of radio spectrum monitoring data processing, and particularly relates to a structured representation method and a data processing method for spectrum monitoring data.
  • electromagnetic spectrum resources are becoming more and more tense, and strengthening electromagnetic spectrum management is becoming more and more important.
  • Monitoring the use of electromagnetic spectrum resources is an important means of electromagnetic spectrum management.
  • electromagnetic spectrum monitoring also requires continuous continuous monitoring, which makes the electromagnetic spectrum monitoring data geographically dispersed and massive in data, and the storage of traditional spectrum monitoring data.
  • processing methods are gradually difficult to adapt to the new requirements for electromagnetic spectrum data transmission and big data processing.
  • the patent CNXXXXXXXXXXXXX.X application date XXX-XX-XX) applied by this unit proposes a data processing system and method for how single spectrum monitoring stations exchange data with other stations, data centers and relay stations. The above system and method describe how data is transmitted between stations and how single station monitoring data is structured. However, how to organize and structure the monitoring data of multiple stations in a geographically dispersed area is not involved.
  • the invention is based on the above patents, further excavating the structural characteristics of the electromagnetic spectrum monitoring data of multiple stations, and proposing new information according to the information correlation of time, frequency, position and energy of the spectrum data of multiple stations.
  • Structured representation, and mathematical operations are defined for single station, multi-station structured representations.
  • the structured representation method provided by the invention will facilitate the deep information mining of the massive monitoring data, facilitate the subsequent compression processing and long-distance transmission of the monitoring data, and is more suitable for the construction requirements of the spectrum monitoring station network system, so that the spectrum Monitoring data is used more reasonably and efficiently.
  • the present invention proposes a data monitoring data structuring method for the electromagnetic spectrum monitoring multi-station data processing requirements.
  • the standardization processing process of the spectrum monitoring data can be realized, a data processing flow conforming to the requirements of the spectrum monitoring network system is formed, and a mathematical operation suitable for a given structured representation is given, which is beneficial to multiple units.
  • Station monitoring data is used more reasonably and efficiently.
  • a structured representation method of spectrum monitoring data comprising the following steps:
  • S1 Discretizing the spectrum monitoring data of a single station in a time dimension and a spectrum dimension to form a two-dimensional spectrum matrix
  • S2 The two-dimensional spectrum matrix obtained by all the stations in a certain area is arranged in a positional dimension with a certain rule as a center, and a three-dimensional spectrum matrix body is constructed.
  • step S1 the construction of the two-dimensional spectrum matrix comprises the following steps:
  • the spectrum monitoring station obtains a synchronous clock with other stations according to the synchronization or calibration method, and obtains information such as a uniformly specified sampling time t n , a spectrum bandwidth B, and a frequency sampling point M, where n is a value of 1, 2 , 3, ..., N, N are positive integers;
  • step S1 the two-dimensional spectral matrix can define the following mathematical operations:
  • the addition of the spectral matrix can be applied to sum the spectral data obtained by multiple stations to obtain an overall average data of the spectrum usage of a certain area.
  • the subtraction of the spectrum matrix can be applied to subtracting the spectrum data obtained by multiple stations to obtain the difference information of the spectrum usage status between different stations; and can also be used to subtract the station monitoring data from the electromagnetic background data. Obtain monitoring data for a specific source of radiation.
  • the spectrum matrix frequency dimension projection can obtain the average spectrum usage of a single station in a certain period of time.
  • the spectral matrix frequency dimension projection can obtain the time occupancy of a single station within a given spectrum segment.
  • step S2 the construction of the three-dimensional spectrum matrix body comprises the following steps:
  • step S2.2 the K spectrum monitoring station spectrum matrices W 1 , W 2 , ... W K are arranged into a spectrum matrix body using the following steps and rules:
  • step S2 the three-dimensional spectrum matrix body can define the following mathematical operations:
  • the addition of the spectral matrix can be applied to sum the spectral data obtained in multiple regions to obtain an overall average of the spectral usage of a larger region.
  • the subtraction of the spectrum matrix body can be applied to subtracting the spectrum data obtained in multiple regions to obtain the difference information of the spectrum usage status between different regions; and can also be used to subtract the regional monitoring data from the regional electromagnetic background data to obtain Monitoring data for radiation sources in specific areas.
  • the frequency matrix projection of the spectrum matrix can obtain the relationship between the spectrum usage and the geographical location in a given period of time in a given area.
  • TProj(Q) [TProj(W 1 ),TProj(W 2 ),...TProj(W K )] T
  • the frequency matrix projection of the spectrum matrix can obtain the relationship between the spectrum usage time and the geographical location in a given period of time in a given area.
  • the invention can correlate the four dimensions of time, spectrum, space and energy of the spectrum data to form a structured organization system in accordance with the spectrum monitoring data.
  • the invention further defines mathematical operations for the spectrum matrix and the spectrum matrix body, thereby conveniently performing information mining on massive monitoring data.
  • the invention is beneficial to monitoring the subsequent compression processing and long-distance transmission of data, and is more suitable for the construction of the spectrum monitoring station network system and the requirements of big data processing, so that the spectrum monitoring data is more rationally and efficiently utilized.
  • the present invention provides a multi-station spectrum monitoring data compression processing method, based on a structured representation method of spectrum monitoring data, Using the similarity characteristics between spectrum monitoring data of multiple stations in a given area, and referring to the principle of traditional video compression to process massive spectrum monitoring data of multiple stations, in the construction of radio spectrum monitoring network and application of data transmission and big data processing The broad prospects are conducive to the more reasonable and efficient use of spectrum monitoring data of radio stations.
  • the technical solution adopted by the present invention is to provide a spectrum monitoring data compression processing method, which includes the following steps:
  • the structured monitoring data of a single station is structured, and the frequency of the monitoring data at a given time and a given bandwidth is discretized and the time dimension of the given time period is obtained according to the unified monitoring frequency bandwidth, the monitoring time dimension sampling interval and the frequency dimension sampling interval.
  • Discretization representing spectrum monitoring data for a given monitoring period and frequency bandwidth as a spectrum matrix
  • the multi-station spectrum monitoring data is structured to represent that all spectrum matrices obtained by multiple stations in a given area are arranged into a spectrum matrix according to specific rules at the station position;
  • the grayscale processing of the spectrum monitoring data selects one dimension from the three dimensions of the spectrum matrix body, and the spectrum matrix body can be regarded as the arrangement of a series of matrices in the dimension, and each matrix is grayed out into a grayscale image.
  • the spectrum matrix body can be regarded as a piece of video data;
  • the video compression method is used to compress the monitoring data. For the monitoring data of multiple stations in a given area, there is a large amount of redundant information in the time, frequency and position dimensions, which can use the traditional video compression standard. The video data corresponding to the spectrum matrix body is compressed.
  • the following two methods can be adopted for the spectrum matrix to be arranged as a spectrum matrix body rule:
  • the geographic location corresponding to the average value of the latitude and longitude of the plurality of stations in a certain area is selected as a reference point, and the spectral matrix data corresponding to each station is arranged by using the geographical distance between all the stations and the reference point as a standard.
  • the spectrum matrix body can be grayed out by the following three methods:
  • the spectrum matrix body can be regarded as data of a series of two-dimensional matrices corresponding to each time point. Each matrix can be grayed out into a grayscale image, and in the time dimension, the spectrum matrix can be regarded as a piece of video data. Obviously, the monitoring data is very correlated in the adjacent monitoring time;
  • the spectrum matrix body can be regarded as data of a series of two-dimensional matrices corresponding to each time point.
  • Each matrix can be grayscaled into a grayscale image, and in the frequency dimension, the spectrum matrix can be regarded as a piece of video data. Monitoring data is also relevant in adjacent frequency bands.
  • the spectrum matrix body can be regarded as data of a series of two-dimensional matrices corresponding to each time point. Each matrix can be converted into a grayscale image by grayscale processing, and in the position dimension, the spectrum matrix body can be regarded as a piece of video data. Monitoring data obtained at different times between adjacent stations also has a strong correlation.
  • the present invention provides a method for processing spectrum monitoring data, including:
  • the two-dimensional spectrum matrix of the plurality of stations is arranged to form a three-dimensional spectrum matrix body according to the mutual relationship of the stations, and the two-dimensional spectrum matrix of the plurality of stations is arranged.
  • the two-dimensional spectrum matrix of each station is formed by the discretization of the time dimension and the spectrum dimension by the spectrum monitoring data of the station.
  • the two-dimensional spectrum matrix of each station is formed by the following steps:
  • a station obtains a synchronous clock with other stations, and obtains a uniformly specified sampling time t n (1...N), a spectrum bandwidth B, and a frequency sampling point number M;
  • the two-dimensional spectrum matrix of the plurality of stations is processed as follows:
  • the two-dimensional spectrum matrix of the plurality of stations is processed as follows:
  • the two-dimensional spectrum matrix of the station is processed as follows:
  • the two-dimensional spectrum matrix of the station is processed as follows:
  • the two-dimensional spectrum matrix of the station is processed as follows:
  • the distance D n between the latitude and longitude position V n of each station and the geometric center point V 0 is obtained by the following steps:
  • ⁇ 2 is a second-order norm operation.
  • the three-dimensional spectrum matrix body Q is formed by the following steps:
  • the three-dimensional spectrum matrix of the stations of each area is processed as follows:
  • the three-dimensional spectrum matrix of the stations in the two regions is processed as follows:
  • the two-dimensional spectrum matrix of the station in the area is processed as follows:
  • TProj(Q) [TProj(W 1 ),TProj(W 2 ),...TProj(W K )] T
  • the three-dimensional spectrum matrix processing of the station in the region is as follows:
  • the three-dimensional spectrum matrix processing of the station of the area is as follows:
  • Q i and Q j are the three-dimensional spectrum matrices of the stations in a region, both of which are N ⁇ M ⁇ K dimensions.
  • the present invention provides a multi-station spectrum monitoring data compression processing method, including:
  • S1 arranging a two-dimensional spectrum matrix of the plurality of stations according to the mutual relationship of the stations, and arranging the two-dimensional spectrum matrix of the plurality of stations to form a three-dimensional spectrum matrix body, wherein each station The two-dimensional spectrum matrix is formed by the discretization of the time dimension and the spectrum dimension by the spectrum monitoring data of the station;
  • S2 selecting one dimension from three dimensions of the three-dimensional spectrum matrix body, and converting each matrix into a grayscale image by performing grayscale processing;
  • step S1 the two-dimensional spectrum matrix is arranged as follows:
  • step S2 the spectrum monitoring data grayscale processing step selects one of the following three methods:
  • the position dimension is selected as the reference dimension, and each matrix is grayed out into a grayscale image.
  • the invention fully exploits the data correlation and redundancy of the spectrum monitoring data of multiple stations in time, frequency and position dimension, and realizes the compression processing of structured spectrum monitoring data by using video compression method. .
  • FIG. 1 is a schematic diagram of an application scenario of a structured representation method for spectrum monitoring data according to the present invention.
  • FIG. 2 is a schematic diagram of a spectrum matrix body of a structured representation method for spectrum monitoring data according to the present invention.
  • FIG. 3 is an example of a gray scale diagram of a spectrum matrix of the spectrum monitoring data.
  • FIG. 4 is a diagram showing the steps of a method for compressing data monitoring of a plurality of stations according to the present invention.
  • FIG. 5 is a schematic diagram of an application scenario of a multi-station spectrum monitoring data compression processing method according to the present invention.
  • FIG. 6 is a schematic diagram of the grayscale processing of the spectrum matrix body with the time dimension as a reference dimension according to the present invention.
  • FIG. 7 is a schematic diagram of the grayscale processing of the spectrum matrix body with the frequency dimension as a reference dimension according to the present invention.
  • FIG. 8 is a schematic diagram of the grayscale processing of the spectrum matrix body with the position dimension as a reference dimension according to the present invention.
  • the invention mainly applies the compression and transmission of spectrum monitoring data of multiple stations, which is beneficial to the more reasonable and efficient utilization of monitoring data of multiple stations.
  • FIG. 1 is a schematic diagram of an application scenario of a structured display method for spectrum monitoring data according to an embodiment of the present invention.
  • the invention provides a structured representation method of spectrum monitoring data, which comprises the following steps:
  • the spectrum matrix obtained by all the stations in a certain area is arranged in a positional dimension according to a certain rule around a certain point, and a three-dimensional spectrum matrix body is constructed.
  • the construction of the two-dimensional spectrum matrix includes the following steps:
  • Step 1 The spectrum monitoring station obtains a synchronous clock with other stations according to a certain synchronization or calibration method, and obtains information such as a uniformly specified sampling time t n (1...N), a spectrum bandwidth B, and a frequency sampling point M;
  • Step 2 At the sampling time point t n , use the spectrum bandwidth B, the frequency sampling point number M and other information to discretize the monitoring data in the frequency dimension to obtain a vector.
  • Step 3 For a given monitoring time period, obtain the spectrum monitoring vector at different sampling times at the sampling time t n (1...N)
  • Step 4 Arranging the spectrum monitoring vectors at different sampling times in chronological order to form a two-dimensional spectrum matrix
  • the matrix W is N x M dimensions.
  • the structured monitoring method of the spectrum monitoring data, the construction of the three-dimensional spectrum matrix body comprises the following steps:
  • Step 1 Obtain monitoring data of all K spectrum monitoring stations in a certain area within a given time period, namely W 1 , W 2 , ... W K ;
  • the structured monitoring method of spectrum monitoring data can define the following mathematical operations:
  • the addition of the spectral matrix can be applied to sum the spectral data obtained by multiple stations to obtain an overall average of the spectrum usage of a region.
  • the subtraction of the spectrum matrix can be applied to subtracting the spectrum data obtained by multiple stations to obtain the difference information of the spectrum usage status between different stations; and can also be used to subtract the station monitoring data from the electromagnetic background data. Obtain monitoring data for a specific source of radiation.
  • the spectrum matrix frequency dimension projection can obtain the average spectrum usage of a single station in a certain period of time.
  • the spectral matrix frequency dimension projection can obtain the time occupancy of a single station within a given spectrum segment.
  • the structured monitoring method of the spectrum monitoring data, the three-dimensional spectrum matrix body can define the following mathematical operations:
  • the addition of the spectral matrix can be applied to sum the spectral data obtained in multiple regions to obtain an overall average of the spectral usage of a larger region.
  • the subtraction of the spectrum matrix body can be applied to subtracting the spectrum data obtained in multiple regions to obtain the difference information of the spectrum usage status between different regions; and can also be used to subtract the regional monitoring data from the regional electromagnetic background data to obtain Monitoring data for radiation sources in specific areas.
  • the frequency matrix projection of the spectrum matrix can obtain the relationship between the spectrum usage and the geographical location in a given period of time in a given area.
  • TProj(Q) [TProj(W 1 ),TProj(W 2 ),...TProj(W K )] T
  • the frequency matrix projection of the spectrum matrix can obtain the relationship between the spectrum usage time and the geographical location in a given period of time in a given area.
  • the K spectrum monitoring station spectrum matrices W 1 , W 2 , ... W K are arranged as a spectrum matrix body, and the following steps and rules can be adopted:
  • Step 1 Calculate the geometric center point V 0 of the geographical distribution according to the latitude and longitude position V n (1...K) of the K spectrum monitoring stations:
  • Step 2 Calculate the distance D n between each spectrum monitoring station V n and the geometric center point V 0 , where ⁇ 2 is a 2nd order norm operation:
  • the present invention represents it in the form of a spectrum matrix.
  • the advantages are as follows: First, it is advantageous to use a conventional spectral matrix such a structured representation and calculation method for subsequent data processing; secondly, the spectrum data is in time— The information in the two dimensions of the frequency is correlated with the bottom layer, similar to the waterfall graph and the time-frequency graph, which helps the further information mining. Finally, the spectrum matrix representation of the spectrum monitoring data can map the spectral data into grayscale images. And can be compressed and processed using an image compression method. For example, using a spectral matrix representation can assist in the following tasks:
  • the present invention proposes a concept and a structured representation of the spectrum matrix body.
  • the advantages are as follows: First, the method extends the traditional matrix form to a new dimension for identifying the location dimension of the spectrum monitoring data. Thereby, the four dimensions of time, spectrum, space and energy of the spectrum data are correlated to form a structured organization system conforming to the spectrum monitoring data.
  • the present invention further defines a mathematical operation for the spectrum matrix body, thereby conveniently Monitoring data for information mining; (3)
  • the spectrum matrix body can convert multi-station spectrum data of a certain area into video data for compression and transmission. For example, the use of spectral matrix representations and operations can assist in the following tasks:
  • the structured monitoring method of spectrum monitoring data proposed by the present invention is more suitable for the construction of the spectrum monitoring station network system and the requirements of big data processing, so that the spectrum monitoring data is more rationally and efficiently utilized.
  • Figure 3 shows the spectrum monitoring data obtained by using the spectrum analyzer at 55 points in Linyi. It is represented as a spectrum matrix.
  • the time domain takes 50 sampling moments, the sampling bandwidth is 500MHz, and the sampling points in the frequency domain are 501.
  • the matrix body Q is 50 ⁇ 501 ⁇ 55 dimensions. Shown in the figure is a spectral matrix grayscale map obtained by mapping 55 spectral matrices into 8th-order gray scales and composing a spectrum matrix body.
  • FIG. 4 is a schematic diagram of a multi-station spectrum monitoring data compression processing method according to the present invention
  • FIG. 5 is a schematic diagram of an application scenario of a multi-station spectrum monitoring data compression processing method according to an embodiment of the present invention.
  • a multi-station spectrum monitoring data compression processing method includes the following steps:
  • Structured data monitoring of single station spectrum monitoring According to the unified monitoring frequency bandwidth, monitoring time dimension sampling interval and frequency dimension sampling interval, the frequency of the monitoring data at a given time and a given bandwidth is discretized and the time dimension of a given time period is maintained. Discretization, representing spectrum monitoring data for a given monitoring period and frequency bandwidth as a two-dimensional spectrum matrix;
  • Multi-station spectrum monitoring data structured representation All spectrum matrices obtained by multiple stations in a given area are arranged into a three-dimensional spectrum matrix according to specific rules at the station position.
  • the construction of the two-dimensional spectrum matrix can take the following steps:
  • Step 1 The spectrum monitoring station obtains a synchronous clock with other stations according to a certain synchronization or calibration method, and obtains information such as a uniformly specified sampling time t n (1...N), a spectrum bandwidth B, and a frequency sampling point M;
  • Step 2 At the sampling time point t n , use the spectrum bandwidth B, the frequency sampling point number M and other information to discretize the monitoring data in the frequency dimension to obtain a vector.
  • Step 3 For a given monitoring time period, obtain the spectrum monitoring vector at different sampling times at the sampling time t n (1...N)
  • Step 4 Arranging the spectrum monitoring vectors at different sampling times in chronological order to form a two-dimensional spectrum matrix
  • the matrix W is N x M dimensions.
  • the construction of the 3D spectrum matrix can take the following steps:
  • Step 1 Obtain monitoring data of all K spectrum monitoring stations in a certain area within a given time period, namely W 1 , W 2 , ... W K ;
  • Grayscale processing of spectrum monitoring data Select one dimension from three dimensions of the spectrum matrix body.
  • the spectrum matrix body can be regarded as the arrangement of a series of matrices in the dimension, and each matrix is grayed out into a grayscale image.
  • the spectrum matrix body can be regarded as a piece of video data;
  • the video compression method is used to compress the monitoring data: for multiple station monitoring data in a given area, there is a large amount of redundant information in all dimensions of time, frequency and position, which can utilize the traditional video compression standard.
  • the video data corresponding to the spectrum matrix body is compressed.
  • Common video compression standards are H.264, H.265, MPEG4, MJPEG, and the like.
  • H.264 has a high compression ratio and also guarantees high quality and smooth images.
  • H.265 is more suitable for HD/Ultra HD transmission. Specific compression methods and standards can be selected according to actual needs.
  • the following two methods can be adopted for the spectrum matrix to be arranged as a spectrum matrix body rule:
  • a monitoring station as the reference station, and arrange the spectrum matrix data corresponding to each station by using the geographical distance between all stations and the reference station as the standard.
  • the location dimension is the geography between the station and the reference station.
  • the straight line distance is used as the scale;
  • the spectrum matrix body can be grayed out by the following three methods:
  • the spectrum matrix body when the time dimension is selected as the reference dimension, can be regarded as data in which a series of two-dimensional matrices corresponding to each time point are arranged. Each matrix can be grayed out into a grayscale image, and in the time dimension, the spectrum matrix can be regarded as a piece of video data.
  • the spectrum matrix body can be regarded as data in which a series of two-dimensional matrices corresponding to each time point are arranged.
  • Each matrix can be grayscaled into a grayscale image, and in the frequency dimension, the spectrum matrix can be regarded as a piece of video data.
  • the spectrum matrix body when the position dimension is selected as the reference dimension, can be regarded as data in which a series of two-dimensional matrices corresponding to each time point are arranged. Each matrix can be converted into a grayscale image by grayscale processing, and in the position dimension, the spectrum matrix body can be regarded as a piece of video data.

Landscapes

  • Physics & Mathematics (AREA)
  • Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Mathematical Physics (AREA)
  • Data Mining & Analysis (AREA)
  • Databases & Information Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Multimedia (AREA)
  • Signal Processing (AREA)
  • Monitoring And Testing Of Transmission In General (AREA)

Abstract

本发明提出了一种频谱监测数据结构化表示方法,包括如下步骤:将单台站频谱监测数据在时间维、频谱维离散化形成二维频谱矩阵;将一定区域内所有台站获得的频谱矩阵以某点为中心按照一定规则在位置维排列,构建三维频谱矩阵体。本发明的积极效果是:能够将频谱数据的时间、频谱、空间、能量四个维度关联起来,形成符合频谱监测数据的结构化组织体系;针对频谱矩阵和频谱矩阵体能够进一步定义数学运算,从而方便的对海量监测数据进行信息挖掘;能够有利于监测数据后续的压缩处理和远距离传输,更为适应频谱监测台站网系的建设和大数据处理的要求,使得频谱监测数据被更加合理和高效的利用。

Description

一种频谱监测数据结构化表示方法、数据处理方法和压缩方法 技术领域
本发明属于无线电频谱监测数据处理技术领域,特别涉及一种频谱监测数据结构化表示方法和数据处理方法。
背景技术
由于信息技术的发展,电磁频谱资源越发显得紧张,加强电磁频谱管理越来越重要。对电磁频谱资源使用情况进行监测,是电磁频谱管理的重要手段。然而,由于电磁频谱监测固定台站、移动监测车辆等数量众多,电磁频谱监测也需要不间断的连续监测,使得电磁频谱监测数据具有地理位置上分散、数据海量的特点,传统频谱监测数据的存储和处理方法渐渐难以适应对电磁频谱数据传输和大数据处理新要求。本单位申请的专利CNXXXXXXXXXXX.X(申请日XXXX-XX-XX)中针对单频谱监测台站如何与其他台站、数据中心和中继台站进行交换数据提出了一种数据处理系统和方法。上述系统和方法对台站间如何传输数据以及单台站监测数据如何结构化进行了描述。但对于地域上分散的一定区域的多台站监测数据如何组织和结构化处理没有涉及。
本发明是在上述专利的基础上,进一步挖掘了多台站电磁频谱监测数据的结构化特征,根据多台站频谱数据的在时间、频率、位置、能量等维度的信息相关性,提出新的结构化表示方法,并对单台站、多台站结构化表示定义了数学运算。本发明给出的结构化表示方法将有利于对海量监测数据进行深度的信息挖掘,有利于监测数据后续的压缩处理和远距离传输,更为适应频谱监测台站网系的建设要求,使得频谱监测数据被更加合理和高效的利用。
另外,随着频谱资源重要性和重视程度的提高,电磁频谱台站数量的急剧扩大,对电磁频谱监测数据处理以获取频谱使用详细状况是合理和高效进行频谱资源利用和频谱管理的基础。尤其是随着频谱网系建设的发展,频谱监测台站及不同频谱管理单位间的数据传输和处理需求不断提高,由于频谱监测数据具有天然的海量、分散的特点,使得传统频谱监测数据的存储和处理方法渐渐难以适应对电磁频谱数据传输和大数据处理新要求。本单位在专利(CNxxxxxxxxx.x,申请日:2018-XX-XX)中将单台站频谱监测数据表示为频 谱矩阵形式,可转为灰度图像进行压缩和传输,在专利(CNxxxxxxxxx.x,申请日:2018-XX-XX)中提出了多台站频谱监测数据结构化表示方法,将其表示为频谱矩阵体。这种结构不仅形式和结构上和视频数据上类似,从数据特点上来说,给定区域的多台站频谱监测数据在时间、频率、位置三个维度上的数据之间相关性非常强。比如,相邻监测时间,频谱监测数据相关性非常强;相邻台站,不同时间获得监测数据也具有强相关性;相邻频段,监测数据也有相关性。因此,以频谱矩阵体形式表示的监测数据具有相当程度的信息冗余。
发明内容
第一方面,在借鉴和发展现有方法和理论的基础上,本发明针对电磁频谱监测多台站数据处理需求,提出一种频谱监测数据结构化方法。利用本发明提出的系统和方法,能够实现频谱监测数据的规范化处理过程,形成符合频谱监测网系要求的数据处理流程,并且给出了适用于给定结构化表示的数学运算,有利于多台站监测数据被更加合理和高效的利用。
为了实现上述目的,本发明采用的技术方案是:
一种频谱监测数据结构化表示方法,包括如下步骤:
S1:将单台站频谱监测数据在时间维、频谱维离散化形成二维频谱矩阵;
S2:将一定区域内所有台站获得的二维频谱矩阵以某点为中心按照一定规则在位置维排列,构建三维频谱矩阵体。
在步骤S1中,二维频谱矩阵的构建包括以下步骤:
S1.1:频谱监测台站根据同步或校准方法,获得和其它台站同步时钟,获取统一规定的采样时间t n、频谱带宽B以及频率采样点数M等信息,其中n取值为1、2、3、……、N,N为正整数;
S1.2:在采样时刻点t n,利用频谱带宽B、频率采样点数M对监测数据进行频率维上的离散化,得到向量
Figure PCTCN2019075666-appb-000001
数列
Figure PCTCN2019075666-appb-000002
为1×M维;
S1.3:给定的监测时间段,在采样时间t n,获取不同采样时刻频谱监测向量
Figure PCTCN2019075666-appb-000003
S1.4:按照时间先后顺序将不同采样时刻频谱监测向量进行排列,组成二维频谱矩阵
Figure PCTCN2019075666-appb-000004
矩阵W为N×M维。
在步骤S1中,二维频谱矩阵可定义以下数学运算:
(1)假设二维频谱矩阵为W i、W j,均为N×M维,则
频谱矩阵加法:
Figure PCTCN2019075666-appb-000005
频谱矩阵的加法可应用于将多台站获得的频谱数据进行求和,从而获得某区域的频谱使用状况的总体平均数据。
频谱矩阵减法:
Figure PCTCN2019075666-appb-000006
频谱矩阵的减法可应用于将多个台站获得的频谱数据进行相减,从而获得不同台站间的频谱使用状况的差异信息;也可以用于将台站监测数据与电磁背景数据相减,获得特定辐射源的监测数据。
(2)频谱矩阵频率维投影:
Figure PCTCN2019075666-appb-000007
频谱矩阵频率维投影可以获得单台站在一定时间段内的频谱平均使用情况。
(3)频谱矩阵时间维投影:
Figure PCTCN2019075666-appb-000008
频谱矩阵频率维投影可以获得单台站在给定频谱段内的时间占用度情况。
(4)m为实数,则频谱矩阵数乘:
Figure PCTCN2019075666-appb-000009
在步骤S2中,三维频谱矩阵体的构建包括以下步骤:
S2.1:获取一定区域内的所有K个频谱监测台站在给定时间段内的监测数据,即W 1,W 2,…W K
S2.2:按照K个频谱监测台站在位置上的相互关系,对W 1,W 2,…W K进行排列,组成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维。
在步骤S2.2中,K个频谱监测台站频谱矩阵W 1,W 2,…W K排列为频谱矩阵体采用如下步骤和规则:
S2.2.1:根据K个频谱监测台站的经纬度位置V n计算出地理分布上的几何中心点V 0,其中n取值为1、2、3、……、K;
Figure PCTCN2019075666-appb-000010
S2.2.2:计算每个频谱监测台站的经纬度位置V n与几何中心点V 0的距离D n,其中‖·‖ 2为2阶范数运算:
D n=‖V n-V 02
S2.2.3:按照频谱监测台站与几何中心点V 0的距离D n从小到大的顺序,排列台站对应的频谱矩阵,构建频谱矩阵体Q=[W 1,W 2,…W K] T
在步骤S2中,三维频谱矩阵体可定义以下数学运算:
(1)假设频谱矩阵为Q i、Q j,均为N×M×K维,则
频谱矩阵体加法:
Figure PCTCN2019075666-appb-000011
频谱矩阵体的加法可应用于将多个区域获得的频谱数据进行求和,从而获得更大区域范围的频谱使用状况的总体平均数据。
频谱矩阵体减法:
Figure PCTCN2019075666-appb-000012
频谱矩阵体的减法可应用于将多个区域获得的频谱数据进行相减,从而获得不同区域间的频谱使用状况的差异信息;也可以用于将区域监测数据与区域电磁背景数据相减,获得特定区域辐射源的监测数据。
(2)频谱矩阵体频率维投影:
FProj(Q)=[FProj(W 1),FProj(W 2),…FProj(W K)] T
频谱矩阵体频率维投影可以获得给定区域一定时间段内的频谱使用情况与地理位置间的关系。
(3)频谱矩阵时间维投影:
TProj(Q)=[TProj(W 1),TProj(W 2),…TProj(W K)] T
频谱矩阵体频率维投影可以获得给定区域一定时间段内的频谱使用时间与地理位置间的关系。
(4)m为实数,则频谱矩阵体数乘运算:
mQ=[mW 1,mW 2,…mW K] T
本发明能够将频谱数据的时间、频谱、空间、能量四个维度关联起来,形成符合频谱监测数据的结构化组织体系。
本发明针对频谱矩阵和频谱矩阵体进一步定义了数学运算,从而方便地对海量监测数据进行信息挖掘。
本发明有利于监测数据后续的压缩处理和远距离传输,更为适应频谱监测台站网系的建设和大数据处理的要求,使得频谱监测数据被更加合理和高效的利用。
第二方面,针对频谱监测数据单台站频谱矩阵和多台站频谱矩阵体结构化表示形式,本发明提出一种多台站频谱监测数据压缩处理方法,基于频谱监测数据的结构化表示方法,利用给定区域内多台站频谱监测数据间的相似性特点,借鉴传统视频压缩的原理来处理多台站海量频谱监测数据,在无线电频谱监测网系的建设和数据传输、大数据处理方面应用前景广阔,有利于无线电台站频谱监测数据被更加合理和高效的利用。
为了实现上述目的,本发明采用的技术方案是,提出一种频谱监测数据压缩处理方法,包括以下步骤:
单台站频谱监测数据结构化表示,根据统一的监测频率带宽、监测时间维采样间隔和频率维采样间隔,完成给定时刻、给定带宽的监测数据频率维离散化和给定时段时间维上的离散化,将给定监测时间段和频率带宽的频谱监测数据表示为频谱矩阵;
多台站频谱监测数据结构化表示,将给定区域内的多台站获得的所有频谱矩阵,按照台站位置上的特定规则排列成频谱矩阵体;
频谱监测数据灰度化处理,从频谱矩阵体的三个维度中选择一个维度,频谱矩阵体可看做该维度上一系列矩阵的排列,将每个矩阵进行灰度化处理转化为灰度图像,则频谱矩阵体可视为一段视频数据;
采用视频压缩方法对监测数据进行压缩,对给定区域内的多台站监测数据来说,在时间、频率、位置各维度上,均有大量的冗余信息,可利用传统视频压缩标准,对频谱矩阵体对应的视频数据进行压缩。
在提出的多台站频谱监测数据压缩处理方法的多台站频谱监测数据结构化表示步骤中,频谱矩阵排列为频谱矩阵体规则可采用以下两种方法:
选择某一监测台站为基准台站,以所有台站与基准台站间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;
选择某区域多台站地理位置经纬度的平均值对应的地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据。
在提出多台站频谱监测数据压缩处理方法的频谱监测数据灰度化处理步骤中,频谱矩阵体可采用以下三种方法来进行灰度化:
选择时间维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在时间维上,频谱矩阵体可视为一段视频数据。显而易见,相邻监测时间,监测数据相关性非常强;
选择频率维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在频率维上,频谱矩阵体可视为一段视频数据。相邻频段,监测数据也有相关性。
选择位置维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在位置维上,频谱矩阵体可视为一段视频数据。相邻台站间,不同时间获得监测数据也具有强相关性。
另一方面,本发明提出一种频谱监测数据的处理方法,包括:
将多个台站的二维频谱矩阵,依据台站在位置上的相互关系,对所述多个台站的二维频谱矩阵进行排列以形成三维频谱矩阵体,
其中,每个台站的二维频谱矩阵通过该台站的频谱监测数据在时间维和频谱维的离散化形成。
进一步,所述每个台站的二维频谱矩阵通过如下步骤形成:
S1.1:一台站获得和其它台站的同步时钟,获取统一规定的采样时间t n(1…N)、频谱带宽B以及频率采样点数M;
S1.2:在采样时刻点t n,利用频谱带宽B和频率采样点数M对频谱监测数据进行频率维上的离散化,得到向量
Figure PCTCN2019075666-appb-000013
数列
Figure PCTCN2019075666-appb-000014
为1×M维;
S1.3:在给定的监测时间段,在采样时间t n,获取不同采样时刻的频谱监测向量
Figure PCTCN2019075666-appb-000015
S1.4:按照时间先后顺序将不同采样时刻的频谱监测向量进行排列,组成二维频谱矩阵
Figure PCTCN2019075666-appb-000016
矩阵W为N×M维。
进一步,当要获得台站的总体频谱监测数据时,按照如下方式对所述多个台站的二维频谱矩阵进行处理:
Figure PCTCN2019075666-appb-000017
当要获得台站的频谱监测数据的差异时,按照如下方式对所述多个台站的二维频谱矩阵进行处理:
Figure PCTCN2019075666-appb-000018
其中,
Figure PCTCN2019075666-appb-000019
当要获得台站的频谱监测数据在一定时间段的使用情况时,按照如下方式对该台站的二维频谱矩阵处理:
Figure PCTCN2019075666-appb-000020
当要获得台站的频谱监测数据在一定频段的频谱评价使用情况时,按照如下方式对该台站的二维频谱矩阵处理:
Figure PCTCN2019075666-appb-000021
当要获得台站的频谱监测数据的加倍数据时,按照如下方式对该台站的二维频谱矩阵处理:
Figure PCTCN2019075666-appb-000022
其中,m为实数。
进一步,通过如下步骤获得一个区域内的K个台站的地理分布上的几何中心点V 0
Figure PCTCN2019075666-appb-000023
n取值为1、2、3、……、K;V i表示第i个台站的经纬度位置。
进一步,通过如下步骤获得每个台站的经纬度位置V n与几何中心点V 0的距离D n
D n=‖V n-V 02
其中‖·‖ 2为2阶范数运算。
进一步,通过如下步骤形成所述三维频谱矩阵体Q:
按照所述区域内的台站与几何中心点V 0的距离D n从小到大的顺序,排列台站对应的二维频谱矩阵,形成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维,W 1,W 2,…W K表示一个区域内的K个台站在给定时间段内的二维频谱矩阵。
进一步,当要获得多个区域的台站的总体频谱监测数据时,按照如下方式对每个区域的台站的三维频谱矩阵进行如下处理:
Figure PCTCN2019075666-appb-000024
当要获得两个区域的台站的频谱监测数据的差异时,按照如下方式对两个 区域的台站的三维频谱矩阵进行如下处理:
Figure PCTCN2019075666-appb-000025
当要获得台站的频谱监测数据在一定频段的频谱评价使用情况时,按照如下方式对该区域的台站的二维频谱矩阵处理:
TProj(Q)=[TProj(W 1),TProj(W 2),…TProj(W K)] T
当要获得一个区域的台站的频谱监测数据在一定频段的使用情况时,按照如下方式对该区域的台站的三维频谱矩阵处理:
FProj(Q)=[FProj(W 1),FProj(W 2),…FProj(W K)] T
当要获得一个区域的台站的频谱监测数据的加倍数据时,按照如下方式对该区域的台站的三维频谱矩阵处理:
mQ=[mW 1,mW 2,…mW K] T
其中,Q i和Q j为一个区域内的台站的三维频谱矩阵,均为N×M×K维,
Figure PCTCN2019075666-appb-000026
Figure PCTCN2019075666-appb-000027
另一方面,本发明提出一种多台站频谱监测数据压缩处理方法,包括:
S1:将多个台站的二维频谱矩阵,依据台站在位置上的相互关系,对所述多个台站的二维频谱矩阵进行排列以形成三维频谱矩阵体,其中,每个台站的二维频谱矩阵通过该台站的频谱监测数据在时间维和频谱维的离散化形成;
S2:从所述三维频谱矩阵体的三个维度中选择一个维度,将每个矩阵进行灰度化处理转化为灰度图像;
S3:采用视频压缩方法对该维度上的频谱监测数据进行压缩。
进一步,步骤S1中,所述二维频谱矩阵的排列如下:
选择某一台站为基准台站,以所有台站与该基准台站之间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;或
选择某区域多台站地理位置经纬度的平均值对应的地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;或
选择任意地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据。
进一步,步骤S2中,频谱监测数据灰度化处理步骤选择以下三种方法之一:
选择时间维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像;或
选择频率维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像;或
选择位置维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像。
与现有技术相比,本发明充分挖掘了多台站频谱监测数据在时间、频率、位置维上的数据相关性和冗余性,利用视频压缩的方法实现了结构化频谱监测数据的压缩处理。
附图说明
图1为本发明所述频谱监测数据结构化表示方法应用场景示意图。
图2为本发明所述频谱监测数据结构化表示方法频谱矩阵体示意图。
图3所述频谱监测数据频谱矩阵体灰度图示例。
图4本发明所述多台站频谱监测数据压缩处理方法步骤图。
图5本发明所述多台站频谱监测数据压缩处理方法应用场景示意图。
图6本发明所述以时间维为基准维进行频谱矩阵体灰度化处理示意图。
图7本发明所述以频率维为基准维进行频谱矩阵体灰度化处理示意图。
图8本发明所述以位置维为基准维进行频谱矩阵体灰度化处理示意图。
具体实施方式
下面结合附图和实施例详细说明本发明的实施方式。
为使本发明实施例的目的、技术方案和优点更加清楚,下面将结合本发明实施例中附图,对本发明实施例中的技术方案进行清楚、完整地描述,显然,所描述的实施例仅仅是本发明一部分实施例,而不是全部的实施例。通常在此处附图中描述和示出的本发明实施例的组件可以以各种不同的配置来布置和设计。因此,以下对在附图中提供的本发明的实施例的详细描述并非旨在限制要求保护的本发明的范围,而是仅仅表示本发明的选定实施例。基于本发明的实施例,本领域技术人员在没有做出创造性劳动的前提下所获得的所有其他实施例,都属于本发明保护的范围。
本发明主要应用多台站频谱监测数据的压缩、传输,有利于多台站监测数据被更加合理和高效的利用。
第一方面
图1为本发明实施例所提供的一种频谱监测数据结构化表示方法应用场景示意图。本发明提供一种频谱监测数据结构化表示方法,包括如下实施步骤:
1)将单台站频谱监测数据在时间维、频谱维离散化形成二维频谱矩阵;
2)将一定区域内所有台站获得的频谱矩阵以某点为中心按照一定规则在位置维排列,构建三维频谱矩阵体。
上述的频谱监测数据结构化表示方法中,二维频谱矩阵的构建包括以下步骤:
步骤1:频谱监测台站根据某种同步或校准方法,获得和其它台站同步时钟,获取统一规定的采样时间t n(1…N)、频谱带宽B、频率采样点数M等信息;
步骤2:在采样时刻点t n,利用频谱带宽B、频率采样点数M等信息对监测数据进行频率维上的离散化,得到向量
Figure PCTCN2019075666-appb-000028
数列
Figure PCTCN2019075666-appb-000029
为1×M维;
步骤3:给定的监测时间段,在采样时间t n(1…N),获取不同采样时刻频谱监测向量
Figure PCTCN2019075666-appb-000030
步骤4:按照时间先后顺序将不同采样时刻频谱监测向量进行排列,组成二维频谱矩阵
Figure PCTCN2019075666-appb-000031
矩阵W为N×M维。
所述的频谱监测数据结构化表示方法,三维频谱矩阵体的构建包括以下步骤:
步骤1:获取一定区域内的所有K个频谱监测台站在给定时间段内的监测数据,即W 1,W 2,…W K
步骤2:按照K个频谱监测台站按照位置上的相互关系,对W 1,W 2,…W K进行排列,组成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维。
所述的频谱监测数据结构化表示方法,二维频谱矩阵可定义以下数学运算:
(1)假设二维频谱矩阵为W i、W j,均为N×M维,则
频谱矩阵加法:
Figure PCTCN2019075666-appb-000032
频谱矩阵的加法可应用于将多台站获得的频谱数据进行求和,从而获得某 区域的频谱使用状况的总体平均数据。
频谱矩阵减法:
Figure PCTCN2019075666-appb-000033
频谱矩阵的减法可应用于将多个台站获得的频谱数据进行相减,从而获得不同台站间的频谱使用状况的差异信息;也可以用于将台站监测数据与电磁背景数据相减,获得特定辐射源的监测数据。
(2)频谱矩阵频率维投影:
Figure PCTCN2019075666-appb-000034
频谱矩阵频率维投影可以获得单台站在一定时间段内的频谱平均使用情况。
(3)频谱矩阵时间维投影:
Figure PCTCN2019075666-appb-000035
频谱矩阵频率维投影可以获得单台站在给定频谱段内的时间占用度情况。
(4)m为实数,则频谱矩阵数乘:
Figure PCTCN2019075666-appb-000036
所述的频谱监测数据结构化表示方法,三维频谱矩阵体可定义以下数学运算:
(1)假设频谱矩矩为Q i、Q j,均为N×M×K维,则
频谱矩阵体加法:
Figure PCTCN2019075666-appb-000037
频谱矩阵体的加法可应用于将多个区域获得的频谱数据进行求和,从而获得更大区域范围的频谱使用状况的总体平均数据。
频谱矩阵体减法:
Figure PCTCN2019075666-appb-000038
频谱矩阵体的减法可应用于将多个区域获得的频谱数据进行相减,从而获得不同区域间的频谱使用状况的差异信息;也可以用于将区域监测数据与区域电磁背景数据相减,获得特定区域辐射源的监测数据。
(2)频谱矩阵体频率维投影:
FProj(Q)=[FProj(W 1),FProj(W 2),…FProj(W K)] T
频谱矩阵体频率维投影可以获得给定区域一定时间段内的频谱使用情况与地理位置间的关系。
(3)频谱矩阵时间维投影:
TProj(Q)=[TProj(W 1),TProj(W 2),…TProj(W K)] T
频谱矩阵体频率维投影可以获得给定区域一定时间段内的频谱使用时间与地理位置间的关系。
(4)m为实数,则频谱矩阵体数乘:
mQ=[mW 1,mW 2,…mW K] T
所述的三维频谱矩阵体的构建步骤中,K个频谱监测台站频谱矩阵W 1,W 2,…W K排列为频谱矩阵体可采用如下步骤和规则:
步骤1:根据K个频谱监测台站的经纬度位置V n(1…K)计算出地理分布上的几何中心点V 0:
Figure PCTCN2019075666-appb-000039
步骤2:计算每个频谱监测台站V n与几何中心点V 0的距离D n,其中‖·‖ 2为2阶范数运算:
D n=‖V n-V 02
步骤3:按照频谱监测台站与几何中心点V 0的距离D n从小到大的顺序,排列台站对应的频谱矩阵,构建频谱矩阵体Q=[W 1,W 2,…W K] T
对于单台站监测数据,本发明将其表示为频谱矩阵的形式,优点如下:首先有利于利用传统的频谱矩阵这种结构化表示和计算方法来进行后续数据处理;其次,频谱数据在时间—频率两个维度上的信息进行了底层的关联,类似于瀑 布图、时频图,有助于进一步的信息挖掘;最后,频谱监测数据的频谱矩阵表示形式,可以将频谱数据映射为灰度图像,并可以方面的采用图像的压缩方法进行压缩处理和传输。例如利用频谱矩阵表示可以辅助完成以下工作:
(1)台站位置处频谱使用随时间变化情况;
(2)利用矩阵加法和数乘可以完成台站位置处频谱使用统计情况;
(3)实现频谱数据格式统一和压缩传输。
对于多台站频谱监测数据,本发明提出了频谱矩阵体的概念和结构化表示方法,优点如下:首先,该方法将传统的矩阵形式扩展了一个新维度用来标识频谱监测数据的位置维,从而实现了将频谱数据的时间、频谱、空间、能量四个维度关联起来,形成符合频谱监测数据的结构化组织体系;其次,本发明针对频谱矩阵体进一步定义了数学运算,从而方便地对海量监测数据进行信息挖掘;(3)频谱矩阵体可将一定区域的多台站频谱数据转化为视频数据来进行压缩和传输。例如,利用频谱矩阵体表示和运算可以辅助完成以下工作:
(1)统计给定区域频谱使用随时间变化情况;
(2)统计给定时间频谱使用随空间变化情况;
(3)统计给定频谱在一定时间和空间的使用情况;
(4)实现频谱数据格式统一和压缩传输;
(5)实现对辐射源的自定位和跟踪;
(6)实现对某辐射源辐射强度的稳定度分析;
(7)一定频段和时间段内,空间电磁环境复杂度的计算。
另外,本发明提出的频谱监测数据结构化表示方法,更为适应频谱监测台站网系的建设和大数据处理的要求,使得频谱监测数据被更加合理和高效的利用。
图3为利用频谱分析仪在临潼某处55个点位获得的频谱监测数据表示为频谱矩阵体实例,时间域取了50个采样时刻,采样带宽是500MHz,频率域采样点数是501,频谱矩矩阵体Q为50×501×55维。图中显示的是将55个频谱矩阵映射为8阶灰度得到的频谱矩阵灰度图并组成频谱矩阵体。
第二方面
图4为本发明所述一种多台站频谱监测数据压缩处理方法步骤图,图5为 本发明实施例所提供的一种多台站频谱监测数据压缩处理方法应用场景示意图。
参照图5,一种多台站频谱监测数据压缩处理方法,包括以下步骤:
单台站频谱监测数据结构化表示:根据统一的监测频率带宽、监测时间维采样间隔和频率维采样间隔,完成给定时刻、给定带宽的监测数据频率维离散化和给定时段时间维上的离散化,将给定监测时间段和频率带宽的频谱监测数据表示为二维频谱矩阵;
多台站频谱监测数据结构化表示:将给定区域内的多台站获得的所有频谱矩阵,按照台站位置上的特定规则排列成三维频谱矩阵体。
二维频谱矩阵的构建可采用以下步骤:
步骤1:频谱监测台站根据某种同步或校准方法,获得和其它台站同步时钟,获取统一规定的采样时间t n(1…N)、频谱带宽B、频率采样点数M等信息;
步骤2:在采样时刻点t n,利用频谱带宽B、频率采样点数M等信息对监测数据进行频率维上的离散化,得到向量
Figure PCTCN2019075666-appb-000040
数列
Figure PCTCN2019075666-appb-000041
为1×M维;
步骤3:给定的监测时间段,在采样时间t n(1…N),获取不同采样时刻频谱监测向量
Figure PCTCN2019075666-appb-000042
步骤4:按照时间先后顺序将不同采样时刻频谱监测向量进行排列,组成二维频谱矩阵
Figure PCTCN2019075666-appb-000043
矩阵W为N×M维。
三维频谱矩阵体的构建可采用以下步骤:
步骤1:获取一定区域内的所有K个频谱监测台站在给定时间段内的监测数据,即W 1,W 2,…W K
步骤2:按照K个频谱监测台站按照位置上的相互关系,对W 1,W 2,…W K进行排列,组成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维。
频谱监测数据灰度化处理:从频谱矩阵体的三个维度中选择一个维度,频谱矩阵体可看作该维度上一系列矩阵的排列,将每个矩阵进行灰度化处理转化为灰度图像,则频谱矩阵体可视为一段视频数据;
采用视频压缩方法对监测数据进行压缩:对给定区域内的多台站监测数据来说,在时间、频率、位置各维度上,均有大量的冗余信息,可利用传统视频压缩标准,对频谱矩阵体对应的视频数据进行压缩。通常的视频压缩标准有 H.264、H.265、MPEG4、MJPEG等。H.264具有较高的压缩比同时还保证了高质量流畅的图像,H.265更加适用于高清/超高清传输领域。具体的压缩方法和标准可根据实际需要来进行选择。
在提出的多台站频谱监测数据压缩处理方法的多台站频谱监测数据结构化表示步骤中,频谱矩阵排列为频谱矩阵体规则可采用以下两种方法:
选择某一监测台站为基准台站,以所有台站与基准台站间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据,位置维以台站与基准台站间的地理直线距离作为刻度;
选择某区域多台站地理位置经纬度的平均值对应的地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据,位置维以台站与基准点间的地理直线距离作为刻度。
在提出多台站频谱监测数据压缩处理方法的频谱监测数据灰度化处理步骤中,频谱矩阵体可采用以下三种方法来进行灰度化:
如图6所示,选择时间维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在时间维上,频谱矩阵体可视为一段视频数据。
如图7所示,选择频率维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在频率维上,频谱矩阵体可视为一段视频数据。
如图8所示,选择位置维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据。可对每个矩阵进行灰度化处理转化为灰度图像,则在位置维上,频谱矩阵体可视为一段视频数据。
对二维矩阵进行灰度化的步骤为:
获取二维矩阵的最大值和最小值,将二维矩阵映射为一定位数的灰度图。

Claims (19)

  1. 一种频谱监测数据结构化表示方法,其特征在于,包括如下步骤:
    S1:将单台站频谱监测数据在时间维、频谱维离散化形成二维频谱矩阵;
    S2:将一定区域内所有台站获得的二维频谱矩阵以某点为中心按照一定规则在位置维排列,构建三维频谱矩阵体。
  2. 根据权利要求1所述频谱监测数据结构化表示方法,其特征在于,在步骤S1中,二维频谱矩阵的构建包括以下步骤:
    S1.1:频谱监测台站根据同步或校准方法,获得和其它台站同步时钟,获取统一规定的采样时间t n、频谱带宽B以及频率采样点数M,其中n取值为1、2、3、……、N,N为正整数;
    S1.2:在采样时刻点t n,利用频谱带宽B、频率采样点数M对监测数据进行频率维上的离散化,得到向量
    Figure PCTCN2019075666-appb-100001
    数列
    Figure PCTCN2019075666-appb-100002
    为1×M维;
    S1.3:给定的监测时间段,在采样时间t n,获取不同采样时刻频谱监测向量
    Figure PCTCN2019075666-appb-100003
    S1.4:按照时间先后顺序将不同采样时刻频谱监测向量进行排列,组成二维频谱矩阵
    Figure PCTCN2019075666-appb-100004
    矩阵W为N×M维。
  3. 根据权利要求2所述频谱监测数据结构化表示方法,其特征在于,在步骤S1中,二维频谱矩阵定义以下数学运算:
    假设二维频谱矩阵为W i、W j,均为N×M维,
    Figure PCTCN2019075666-appb-100005
    Figure PCTCN2019075666-appb-100006
    频谱矩阵加法:
    Figure PCTCN2019075666-appb-100007
    频谱矩阵减法:
    Figure PCTCN2019075666-appb-100008
    频谱矩阵频率维投影:
    Figure PCTCN2019075666-appb-100009
    频谱矩阵时间维投影:
    Figure PCTCN2019075666-appb-100010
    m为实数,则频谱矩阵数乘:
    Figure PCTCN2019075666-appb-100011
  4. 根据权利要求1所述频谱监测数据结构化表示方法,其特征在于,在步骤S2中,三维频谱矩阵体的构建包括以下步骤:
    S2.1:获取一定区域内的所有K个频谱监测台站在给定时间段内的监测数据,即W 1,W 2,…W K
    S2.2:按照K个频谱监测台站在位置上的相互关系,对W 1,W 2,…W K进行排列,组成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维。
  5. 根据权利要求4所述频谱监测数据结构化表示方法,其特征在于,在步骤S2.2中,K个频谱监测台站频谱矩阵W 1,W 2,…W K排列为频谱矩阵体采用如下步骤和规则:
    S2.2.1:根据K个频谱监测台站的经纬度位置V n计算出地理分布上的几何中心点V 0,其中n取值为1、2、3、……、K;
    Figure PCTCN2019075666-appb-100012
    S2.2.2:计算每个频谱监测台站的经纬度位置V n与几何中心点V 0的距离D n,其中||·|| 2为2阶范数运算:
    D n=||V n-V 0|| 2
    S2.2.3:按照频谱监测台站与几何中心点V 0的距离D n从小到大的顺序,排列台站对应的频谱矩阵,构建频谱矩阵体Q=[W 1,W 2,…W K] T
  6. 根据权利要求4所述的频谱监测数据结构化表示方法,其特征在于,在步骤S2中,三维频谱矩阵体定义以下数学运算:
    假设频谱矩矩为Qi、Qj,均为N×M×K维,
    Figure PCTCN2019075666-appb-100013
    Figure PCTCN2019075666-appb-100014
    频谱矩阵体加法:
    Figure PCTCN2019075666-appb-100015
    频谱矩阵体减法:
    Figure PCTCN2019075666-appb-100016
    频谱矩阵体频率维投影:
    FProj(Q)=[FProj(W 1),FProj(W 2),…FProj(W K)] T
    频谱矩阵时间维投影:
    TProj(Q)=[TProj(W 1),TProj(W 2),…TProj(W K)] T
    m为实数,则频谱矩阵体数乘:
    mQ=[mW 1,mW 2,…mW K] T
  7. 一种多台站频谱监测数据压缩处理方法,其特征在于,包括以下步骤:
    S1:单台站频谱监测数据结构化表示:根据统一的监测频率带宽、监测时间维采样间隔和频率维采样间隔,完成给定时刻、给定带宽的监测数据频率维离散化和给定时段时间维上的离散化,将给定监测时间段和频率带宽的频谱监测数据表示为频谱矩阵;
    S2:多台站频谱监测数据结构化表示:将给定区域内的多台站获得的所有频谱矩阵,按照台站位置上的特定规则排列成频谱矩阵体;
    S3:频谱监测数据灰度化处理:从频谱矩阵体的三个维度中选择一个维度,频谱矩阵体可看做该维度上一系列矩阵的排列,将每个矩阵进行灰度化处理转化为灰度图像,则频谱矩阵体可视为一段视频数据;
    S4:采用视频压缩方法对监测数据进行压缩:对给定区域内的多台站监测数据来说,在时间、频率、位置各维度上,均有大量的冗余信息,利用传统视 频压缩标准,对频谱矩阵体对应的视频数据进行压缩。
  8. 根据权利要求7所述多台站频谱监测数据压缩处理方法,步骤S2中,频谱矩阵排列为频谱矩阵体的规则采用以下三种方法之一:
    选择某一监测台站为基准台站,以所有台站与基准台站间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;或
    选择某区域多台站地理位置经纬度的平均值对应的地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;或
    选择任意地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据。
  9. 根据权利要求7所述多台站频谱监测数据压缩处理方法,步骤S3中,频谱监测数据灰度化处理步骤选择以下三种方法之一:
    选择时间维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据;对每个矩阵进行灰度化处理转化为灰度图像,则在时间维上,频谱矩阵体可视为一段视频数据;或
    选择频率维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据;对每个矩阵进行灰度化处理转化为灰度图像,则在频率维上,频谱矩阵体可视为一段视频数据;或
    选择位置维作为基准维,则频谱矩阵体可看作每个时间点对应的一系列二维矩阵排列而成的数据;对每个矩阵进行灰度化处理转化为灰度图像,则在位置维上,频谱矩阵体可视为一段视频数据。
  10. 一种频谱监测数据的处理方法,其特征在于,包括:
    将多个台站的二维频谱矩阵,依据台站在位置上的相互关系,对所述多个台站的二维频谱矩阵进行排列以形成三维频谱矩阵体,
    其中,每个台站的二维频谱矩阵通过该台站的频谱监测数据在时间维和频谱维的离散化形成。
  11. 根据权利要求10所述的频谱监测数据的处理方法,其特征在于,所述每个台站的二维频谱矩阵通过如下步骤形成:
    S1.1:一台站获得和其它台站的同步时钟,获取统一规定的采样时间t n(1…N)、频谱带宽B以及频率采样点数M;
    S1.2:在采样时刻点t n,利用频谱带宽B和频率采样点数M对频谱监测数据进行频率维上的离散化,得到向量
    Figure PCTCN2019075666-appb-100017
    数列
    Figure PCTCN2019075666-appb-100018
    为1×M维;
    S1.3:在给定的监测时间段,在采样时间t n,获取不同采样时刻的频谱监测向量
    Figure PCTCN2019075666-appb-100019
    S1.4:按照时间先后顺序将不同采样时刻的频谱监测向量进行排列,组成二维频谱矩阵
    Figure PCTCN2019075666-appb-100020
    矩阵W为N×M维。
  12. 根据权利要求11所述的频谱监测数据的处理方法,其特征在于,
    当要获得台站的总体频谱监测数据时,按照如下方式对所述多个台站的二维频谱矩阵进行处理:
    Figure PCTCN2019075666-appb-100021
    当要获得台站的频谱监测数据的差异时,按照如下方式对所述多个台站的二维频谱矩阵进行处理:
    Figure PCTCN2019075666-appb-100022
    其中,
    Figure PCTCN2019075666-appb-100023
    当要获得台站的频谱监测数据在一定时间段的使用情况时,按照如下方式对该台站的二维频谱矩阵处理:
    Figure PCTCN2019075666-appb-100024
    当要获得台站的频谱监测数据在一定频段的频谱评价使用情况时,按照如下方式对该台站的二维频谱矩阵处理:
    Figure PCTCN2019075666-appb-100025
    当要获得台站的频谱监测数据的加倍数据时,按照如下方式对该台站的二维频谱矩阵处理:
    Figure PCTCN2019075666-appb-100026
    其中,m为实数。
  13. 根据权利要求12所述的频谱监测数据的处理方法,其特征在于,通过如下步骤获得一个区域内的K个台站的地理分布上的几何中心点V 0
    Figure PCTCN2019075666-appb-100027
    n取值为1、2、3、……、K;V i表示第i个台站的经纬度位置。
  14. 根据权利要求13所述的频谱监测数据的处理方法,其特征在于,通过如下步骤获得每个台站的经纬度位置V n与几何中心点V 0的距离D n
    D n=||V n-V 0|| 2
    其中||·|| 2为2阶范数运算。
  15. 根据权利要求14所述的频谱监测数据的处理方法,其特征在于,通过如下步骤形成所述三维频谱矩阵体Q:
    按照所述区域内的台站与几何中心点V 0的距离D n从小到大的顺序,排列台站对应的二维频谱矩阵,形成三维频谱矩阵体Q=[W 1,W 2,…W K] T,矩阵体Q为N×M×K维,W 1,W 2,…W K表示一个区域内的K个台站在给定时间段内的二维频谱矩阵。
  16. 根据权利要求15所述的频谱监测数据的处理方法,其特征在于,
    当要获得多个区域的台站的总体频谱监测数据时,按照如下方式对每个区域的台站的三维频谱矩阵进行如下处理:
    Figure PCTCN2019075666-appb-100028
    当要获得两个区域的台站的频谱监测数据的差异时,按照如下方式对两个区域的台站的三维频谱矩阵进行如下处理:
    Figure PCTCN2019075666-appb-100029
    当要获得台站的频谱监测数据在一定频段的频谱评价使用情况时,按照如下方式对该区域的台站的二维频谱矩阵处理:
    TProj(Q)=[TProj(W 1),TProj(W 2),…TProj(W K)] T
    当要获得一个区域的台站的频谱监测数据在一定频段的使用情况时,按照如下方式对该区域的台站的三维频谱矩阵处理:
    FProj(Q)=[FProj(W 1),FProj(W 2),…FProj(W K)] T
    当要获得一个区域的台站的频谱监测数据的加倍数据时,按照如下方式对该区域的台站的三维频谱矩阵处理:
    mQ=[mW 1,mW 2,…mW K] T
    其中,Q i和Q j为一个区域内的台站的三维频谱矩阵,均为N×M×K维,
    Figure PCTCN2019075666-appb-100030
  17. 一种多台站频谱监测数据压缩处理方法,其特征在于,包括:
    S1:将多个台站的二维频谱矩阵,依据台站在位置上的相互关系,对所述多个台站的二维频谱矩阵进行排列以形成三维频谱矩阵体,其中,每个台站的二维频谱矩阵通过该台站的频谱监测数据在时间维和频谱维的离散化形成;
    S2:从所述三维频谱矩阵体的三个维度中选择一个维度,将每个矩阵进行灰度化处理转化为灰度图像;
    S3:采用视频压缩方法对该维度上的频谱监测数据进行压缩。
  18. 根据权利要求17所述的多台站频谱监测数据压缩处理方法,步骤S1中,所述二维频谱矩阵的排列如下:
    选择某一台站为基准台站,以所有台站与该基准台站之间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据;或
    选择某区域多台站地理位置经纬度的平均值对应的地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱 矩阵数据;或
    选择任意地理位置作为基准点,以所有台站与基准点间的地理直线距离远近作为标准来排列各台站对应的频谱矩阵数据。
  19. 根据权利要求17所述的多台站频谱监测数据压缩处理方法,步骤S2中,频谱监测数据灰度化处理步骤选择以下三种方法之一:
    选择时间维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像;或
    选择频率维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像;或
    选择位置维作为基准维,对每个矩阵进行灰度化处理转化为灰度图像。
PCT/CN2019/075666 2018-03-06 2019-02-21 一种频谱监测数据结构化表示方法、数据处理方法和压缩方法 WO2019170001A1 (zh)

Priority Applications (1)

Application Number Priority Date Filing Date Title
US16/979,112 US20200400730A1 (en) 2018-03-06 2019-02-21 Frequency spectrum monitoring data structured representation method, and data processing method and compression method

Applications Claiming Priority (4)

Application Number Priority Date Filing Date Title
CN201810181414.7 2018-03-06
CN201810181415.1A CN108616720B (zh) 2018-03-06 2018-03-06 一种多台站频谱监测数据压缩处理方法
CN201810181415.1 2018-03-06
CN201810181414.7A CN108509506B (zh) 2018-03-06 2018-03-06 一种频谱监测数据结构化表示方法

Publications (1)

Publication Number Publication Date
WO2019170001A1 true WO2019170001A1 (zh) 2019-09-12

Family

ID=67846859

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/075666 WO2019170001A1 (zh) 2018-03-06 2019-02-21 一种频谱监测数据结构化表示方法、数据处理方法和压缩方法

Country Status (2)

Country Link
US (1) US20200400730A1 (zh)
WO (1) WO2019170001A1 (zh)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116953356A (zh) * 2023-09-20 2023-10-27 南京纳特通信电子有限公司 地空一体三维空间无线电频谱监测方法及系统

Families Citing this family (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115792137B (zh) * 2023-01-17 2023-06-06 河北先河环保科技股份有限公司 大气污染溯源方法及装置、终端
CN116192678B (zh) * 2023-04-26 2023-06-23 武汉能钠智能装备技术股份有限公司四川省成都市分公司 无线电频谱监测数据的压缩系统及方法
CN116527177B (zh) * 2023-04-28 2024-03-12 哈尔滨工程大学 一种基于相关性分析和图卷积网络的三维频谱预测方法

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103746750A (zh) * 2013-08-23 2014-04-23 西华大学 无线电监测电磁态势预测系统
CN106100776A (zh) * 2016-08-31 2016-11-09 成都九华圆通科技发展有限公司 基于无线台站网格监测系统的频谱感知方法
CN106844794A (zh) * 2017-04-12 2017-06-13 云南大学 一种基于数据多维特征的无线电监管的装置和方法
CN108509506A (zh) * 2018-03-06 2018-09-07 西安大衡天成信息科技有限公司 一种频谱监测数据结构化表示方法
CN108616720A (zh) * 2018-03-06 2018-10-02 西安大衡天成信息科技有限公司 一种多台站频谱监测数据压缩处理方法

Family Cites Families (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US6243730B1 (en) * 1999-05-04 2001-06-05 Sony Electronics, Inc. Methods and systems for performing short integer chen IDCT algorithm with fused multiply/add
US7123655B2 (en) * 2001-08-09 2006-10-17 Sharp Laboratories Of America, Inc. Method for reduced bit-depth quantization
US7296045B2 (en) * 2004-06-10 2007-11-13 Hasan Sehitoglu Matrix-valued methods and apparatus for signal processing
US8719736B1 (en) * 2013-03-15 2014-05-06 Synopsys, Inc. Compact and accurate wafer topography proximity effect modeling for full chip simulation
CN108732423A (zh) * 2018-03-06 2018-11-02 西安大衡天成信息科技有限公司 一种频谱监测数据处理系统及方法

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103746750A (zh) * 2013-08-23 2014-04-23 西华大学 无线电监测电磁态势预测系统
CN106100776A (zh) * 2016-08-31 2016-11-09 成都九华圆通科技发展有限公司 基于无线台站网格监测系统的频谱感知方法
CN106844794A (zh) * 2017-04-12 2017-06-13 云南大学 一种基于数据多维特征的无线电监管的装置和方法
CN108509506A (zh) * 2018-03-06 2018-09-07 西安大衡天成信息科技有限公司 一种频谱监测数据结构化表示方法
CN108616720A (zh) * 2018-03-06 2018-10-02 西安大衡天成信息科技有限公司 一种多台站频谱监测数据压缩处理方法

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
LI, TIE ET AL.: "Research on Spectrum Management Model of Radio Communication", SHIPBOARD ELECTRONIC COUNTERMEASURE, vol. 28, no. 5, 31 October 2005 (2005-10-31), pages 27; 28, ISSN: 1673-9167 *

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN116953356A (zh) * 2023-09-20 2023-10-27 南京纳特通信电子有限公司 地空一体三维空间无线电频谱监测方法及系统

Also Published As

Publication number Publication date
US20200400730A1 (en) 2020-12-24

Similar Documents

Publication Publication Date Title
WO2019170001A1 (zh) 一种频谱监测数据结构化表示方法、数据处理方法和压缩方法
CN106682234B (zh) 一种基于空间插值的电磁频谱分布预测和动态可视化方法
CN108509506B (zh) 一种频谱监测数据结构化表示方法
Khan et al. A cloud-based architecture for citizen services in smart cities
CN112115198A (zh) 一种城市遥感智能服务平台
US20120001915A1 (en) Close-packed, uniformly adjacent multiresolutional, overlapping spatial data ordering
Wang et al. Edge federated learning via unit-modulus over-the-air computation
Do et al. Graph-deep-learning-based inference of fine-grained air quality from mobile IoT sensors
CN113128105B (zh) 突发性流域水污染事故监测方法及装置
CN111080501B (zh) 基于手机信令数据的真实人群密度时空分布估算方法
CN109462736B (zh) 一种基于bim+gis建模的路桥工程建设视频会议管理方法及系统
CN108235057A (zh) 一种大规模海洋洋流数据的网络传输及可视化方法
CN107220758A (zh) 一种电网规划辅助系统
CN108616720B (zh) 一种多台站频谱监测数据压缩处理方法
Zhou et al. Compressive sensing based distributed data storage for mobile crowdsensing
Xu et al. Spectragan: Spectrum based generation of city scale spatiotemporal mobile network traffic data
Kuzovkova et al. Matching of 6G Network Capabilities to Digital Services Requirements
Hohl et al. Spatiotemporal domain decomposition for massive parallel computation of space-time kernel density
Perifanis et al. Towards energy-aware federated traffic prediction for cellular networks
CN108732423A (zh) 一种频谱监测数据处理系统及方法
Chopade et al. Real-time large-scale big data networks analytics and visualization architecture
CN115860153B (zh) 基于个性化分组联邦学习的无线流量预测方法及系统
Lovén et al. Towards EDISON: An edge-native approach to distributed interpolation of environmental data
CN107995278A (zh) 一种基于城域级物联网感知数据的场景智能分析系统与方法
Gu et al. Predicting the network shift of large urban agglomerations in China using the deep-learning gravity model: A perspective of population migration

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19764963

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19764963

Country of ref document: EP

Kind code of ref document: A1