Data compression method and system for radio signals
By constructing and training neural network models to compress and decompress radio signals, the problem of low storage and use of radio signals is solved, effective data classification and identification is realized, and application needs of radio signal big data is met.
Patent Information
- Application Number
- CN202411097386.2
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-08-12
- Publication Date
- 2024-12-13
AI Technical Summary
The prior art is difficult to classify and identify signals before decompression of radio signal data, resulting in low data storage and use efficiency and cannot meet the application needs of radio signal big data.
A neural network consisting of an encoder and a decoder is constructed, trained through radio signal samples, and a neural network model that meets the error requirements is generated, which is used to compress and decompress radio signals, and to perform conditional mapping and retrieval in the database.
It realizes effective compression and decompression of radio signals, reduces storage requirements, and can be used for signal classification and identification before decompression of data, and has important application value.
Smart Images
Figure CN119136253A_ABST
Abstract
Description
Technical Field
[0001] The present invention relates to the technical field of signal processing, and in particular to a data compression method and system for radio signals. Background Art
[0002] In the field of radio spectrum management, it is usually necessary to collect a large number of radio signals and build a signal library. Based on the data in the signal library and with appropriate methods, the type of signal, the protocol carried by the signal, and the behavior of the frequency-using equipment can be determined.
[0003] With the advancement of technology, the radio spectrum is expanding, the radio communication rate is increasing, and the scale of collected radio signals is getting larger and larger, forming radio signal big data. Its characteristics are huge data volume and extremely fast growth rate, which brings huge pressure to the storage system.
[0004] In order to reduce the storage pressure brought to the system by radio signal big data, the radio signal data needs to be compressed before storage, and then decoded and restored when the radio signal data is used in the future. Using conventional file compression methods, the data is compressed into rar or zip files, and the data cannot be used before decompression. Using methods similar to audio compression, there is also the problem that the data cannot be used before decompression. If there is a compression method that can also be used for signal classification and identification before the data is decompressed, it can meet the application requirements of radio signal big data. The present invention is generated based on this practical need. Summary of the invention
[0005] In order to solve the above problems, the purpose of the present invention is to provide a radio signal data compression technology to meet the demand for radio signal big data.
[0006] In order to achieve the above technical objectives, the present application provides a data compression method for radio signals, comprising the following steps:
[0007] Construct a neural network consisting of an encoder and a decoder, train it with radio signal samples, and generate a neural network model that meets the error requirements;
[0008] In response to compressing the radio signal, calling the encoder of the neural network model to compress the radio signal and store it in a database;
[0009] In response to extracting the radio signal, the compressed radio signal is called from the database, decompressed by a decoder of the neural network model, and the radio signal is obtained.
[0010] Preferably, in the process of constructing the neural network, a fully connected approach is adopted to construct the neural network, wherein the number of layers of the encoder and the decoder are the same.
[0011] Preferably, in the process of constructing the neural network, the encoder has 5 layers, the number of nodes in the input layer is n, the number of nodes in the first layer is n / 2, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 8, and the number of nodes in the output layer is n / 16;
[0012] The decoder has 5 layers, the number of nodes in the input layer is n / 16, the number of nodes in the first layer is n / 8, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 2, and the number of nodes in the output layer is n.
[0013] Preferably, in the process of constructing the neural network model, m radio signals are collected, and the number of m is not less than n×n; the difference between the original radio signal X=(x(1), x(2), ..., x(n)) and the decoded radio signal X'=(x'(1), x'(2), ..., x'(n))
[0014]
[0015] Preferably, during the process of compressing the radio signal, the category of the radio signal is obtained according to the pre-examination classification standard, and the search path of the radio signal is generated according to the timestamp corresponding to the radio signal during the compression process, and conditional mapping is performed with the database.
[0016] Preferably, in the process of extracting the radio signal, based on the input search conditions and the search path, the compressed radio signal is obtained from the database, and the radio signal is extracted by decompressing it through the decoder of the neural network model.
[0017] The present invention discloses a data compression system for radio signals, comprising:
[0018] A model building module is used to build a neural network consisting of an encoder and a decoder, and train it through radio signal samples to generate a neural network model that meets the error requirements;
[0019] A compression module is used to call the encoder of the neural network model to compress the radio signal and store it in the database;
[0020] The decompression module is used to call the compressed radio signal from the database, decompress it through the decoder of the neural network model, and obtain the radio signal.
[0021] Preferably, the model construction module is also used to construct a neural network in a fully connected manner, wherein the number of layers of the encoder and the decoder is the same; the encoder has 5 layers, the number of input layer nodes is n, the number of first layer nodes is n / 2, the number of second layer nodes is n / 4, the number of third layer nodes is n / 8, and the number of output layer nodes is n / 16; the decoder has 5 layers, the number of input layer nodes is n / 16, the number of first layer nodes is n / 8, the number of second layer nodes is n / 4, the number of third layer nodes is n / 2, and the number of output layer nodes is n.
[0022] Preferably, the compression module is further used to obtain the category of the radio signal according to the pre-examination classification standard, and generate a retrieval path of the radio signal according to the timestamp corresponding to the radio signal in the compression process, and perform conditional mapping with the database.
[0023] Preferably, the decompression module is also used to obtain the compressed radio signal in the database according to the input search conditions and the search path, and decompress it through the decoder of the neural network model to extract the radio signal.
[0024] The present invention discloses the following technical effects:
[0025] The present invention compresses the radio signal and then stores it, which is beneficial to reducing the storage demand of radio signal big data. The data is used for classification and identification before decompression, and has important application value in radio signal big data. BRIEF DESCRIPTION OF THE DRAWINGS
[0026] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.
[0027] Figure 1 It is a schematic diagram of the method steps described in the present invention. DETAILED DESCRIPTION
[0028] To make the purpose, technical scheme and advantages of the embodiments of the present application clearer, the technical scheme in the embodiments of the present application will be clearly and completely described below in conjunction with the drawings in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, rather than all of the embodiments. The components of the embodiments of the present application generally described and shown in the drawings here can be arranged and designed in various different configurations. Therefore, the following detailed description of the embodiments of the present application provided in the drawings is not intended to limit the scope of the application claimed for protection, but merely represents the selected embodiments of the present application. Based on the embodiments of the present application, all other embodiments obtained by those skilled in the art without making creative work belong to the scope of protection of the present application.
[0029] like Figure 1 As shown, the present invention provides a data compression method for radio signals, comprising the following steps:
[0030] Construct a neural network consisting of an encoder and a decoder, train it with radio signal samples, and generate a neural network model that meets the error requirements;
[0031] In response to compressing the radio signal, calling the encoder of the neural network model to compress the radio signal and store it in a database;
[0032] In response to extracting the radio signal, the compressed radio signal is called from the database, decompressed by a decoder of the neural network model, and the radio signal is obtained.
[0033] Further preferably, in the data compression method provided by the present invention, in the process of constructing the neural network, a fully connected manner is used to construct the neural network, wherein the number of layers of the encoder and the decoder are the same.
[0034] Further preferably, in the data compression method provided by the present invention, in the process of constructing a neural network, the encoder has 5 layers, the number of nodes in the input layer is n, the number of nodes in the first layer is n / 2, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 8, and the number of nodes in the output layer is n / 16;
[0035] The decoder has 5 layers, the number of nodes in the input layer is n / 16, the number of nodes in the first layer is n / 8, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 2, and the number of nodes in the output layer is n.
[0036] Further preferably, in the data compression method provided by the present invention, in the process of constructing the neural network model, m radio signals are collected, and the number of m is not less than n×n; the difference between the original radio signal X=(x(1), x(2), ..., x(n)) and the decoded radio signal X'=(x'(1), x'(2), ..., x'(n))
[0037] Further preferably, the data compression method provided by the present invention, during the process of compressing the radio signal, obtains the category of the radio signal according to the pre-examination classification standard, and generates a retrieval path of the radio signal according to the timestamp corresponding to the radio signal during the compression process, and performs conditional mapping with the database.
[0038] Further preferably, the data compression method provided by the present invention, in the process of extracting the radio signal, obtains the compressed radio signal in the database according to the input retrieval conditions and the retrieval path, decompresses it through the decoder of the neural network model, and extracts the radio signal.
[0039] The present invention discloses a data compression system for radio signals, which is used to implement the data compression method provided by the present invention, and comprises:
[0040] A model building module is used to build a neural network consisting of an encoder and a decoder, and train it through radio signal samples to generate a neural network model that meets the error requirements;
[0041] A compression module is used to call the encoder of the neural network model to compress the radio signal and store it in the database;
[0042] The decompression module is used to call the compressed radio signal from the database, decompress it through the decoder of the neural network model, and obtain the radio signal.
[0043] Further preferably, the model building module of the data compression system disclosed in the present invention is also used to construct a neural network in a fully connected manner, wherein the number of layers of the encoder and the decoder is the same; the encoder has 5 layers, the number of input layer nodes is n, the number of first layer nodes is n / 2, the number of second layer nodes is n / 4, the number of third layer nodes is n / 8, and the number of output layer nodes is n / 16; the decoder has 5 layers, the number of input layer nodes is n / 16, the number of first layer nodes is n / 8, the number of second layer nodes is n / 4, the number of third layer nodes is n / 2, and the number of output layer nodes is n.
[0044] Further preferably, the compression module of the data compression system disclosed in the present invention is also used to obtain the category of the radio signal according to the pre-examination classification standard, and generate a retrieval path for the radio signal according to the timestamp corresponding to the radio signal in the compression process, and perform conditional mapping with the database.
[0045] Further preferably, the decompression module of the data compression system disclosed in the present invention is also used to obtain the compressed radio signal in the database according to the input search conditions and the search path, decompress it through the decoder of the neural network model, and extract the radio signal.
[0046] Embodiment: The present invention provides a technology for radio signal data compression, comprising the following steps:
[0047] (1) Neural network design steps: Design a neural network that includes an encoder and a decoder. The encoder is used to compress the signal, and the decoder is used to restore the signal. The encoder and decoder are cascaded, and the output of the encoder is the input of the decoder.
[0048] Specifically, if a radio signal is represented as a sequence X = (x(1), x(2), ..., x(n)), the neural network can be fully connected. For example, the encoder has 5 layers, the number of input layer nodes is n, the number of first layer nodes is n / 2, the number of second layer nodes is n / 4, the number of third layer nodes is n / 8, and the number of output layer nodes is n / 16. The decoder has 5 layers, the number of input layer nodes is n / 16, the number of first layer nodes is n / 8, the number of second layer nodes is n / 4, the number of third layer nodes is n / 2, and the number of output layer nodes is n.
[0049] (2) Neural network training steps: Collect radio signal samples to train the neural network, so that the radio signal samples are encoded by the encoder and then decoded by the decoder, and the difference between the radio signal obtained after decoding and the radio signal before encoding meets the error requirement.
[0050] In specific implementation, m radio signals need to be collected, and the number of m is not less than n×n. The difference between the original radio signal X=(x(1), x(2), ..., (n)) and the decoded (decompressed) radio signal X'=(x'(1), x'(2), ..., x'(n))
[0051] (3) Radio signal encoding and compression step: Encode the radio signal using the trained neural network encoder to obtain a compressed radio signal. Since the neural network is a feature extractor, the compressed radio signal represents the characteristics of the original radio signal. Other neural networks or machine learning methods can be used to classify and identify the compressed radio signal.
[0052] In a specific implementation, the radio signal X = (x(1), x(2), ..., x(n)) is outputted as Y = (y(1), y(2), ..., y(k)) after the encoder. According to step (1), k = n / 16. Y is the compressed radio signal. Y represents certain features of X and can be stored in a large radio signal database or used for signal recognition and classification.
[0053] (4) Radio signal decoding and recovery step: The compressed radio signal is decoded using a decoder of the trained neural network to obtain a recovered radio signal.
[0054] In specific implementation, the compressed radio signal Y=(y(1), y(2), ..., y(k)) is output as X'=(x'(1), x'(2), ..., x'(n)) after the decoder. The decoded (decompressed) radio signal X' is not much different from the original radio signal X.
[0055] The present invention compresses the radio signal and then stores it, which is beneficial to reducing the storage demand of radio signal big data. The data is used for classification and identification before decompression, and has important application value in radio signal big data.
[0056] The present invention is described with reference to flowcharts and / or block diagrams of methods, devices (systems), and computer program products according to embodiments of the present invention. It should be understood that each process and / or block in the flowchart and / or block diagram, as well as the combination of processes and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing device to produce a machine, so that the instructions executed by the processor of the computer or other programmable data processing device generate instructions for implementing the processes in the flowchart and / or block diagram. Figure 1 A process or multiple processes and / or boxes Figure 1 A device that provides the functions specified in a block or multiple blocks.
[0057] In the description of the present invention, it should be understood that the terms "first" and "second" are used for descriptive purposes only and should not be understood as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include one or more of the features. In the description of the present invention, the meaning of "plurality" is two or more, unless otherwise clearly and specifically defined.
[0058] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.
Claims
1. A data compression method for radio signals, characterized in that: The following steps are involved: Construct a neural network consisting of an encoder and a decoder, train it with radio signal samples, and generate a neural network model that meets the error requirements; In response to compressing the radio signal, calling the encoder of the neural network model to compress the radio signal and store it in a database; In response to extracting the radio signal, the compressed radio signal is called from the database, decompressed by the decoder of the neural network model, and the radio signal is obtained.
2. A method for compressing radio signal data according to claim 1, characterized in that: In the process of constructing the neural network, the neural network is constructed in a fully connected manner, wherein the number of layers of the encoder and the decoder is the same.
3. A method for compressing radio signal data according to claim 2, characterized in that: In the process of constructing the neural network, the encoder has 5 layers, the number of nodes in the input layer is n, the number of nodes in the first layer is n / 2, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 8, and the number of nodes in the output layer is n / 16; The decoder has 5 layers, the number of nodes in the input layer is n / 16, the number of nodes in the first layer is n / 8, the number of nodes in the second layer is n / 4, the number of nodes in the third layer is n / 2, and the number of nodes in the output layer is n.
4. A method for compressing radio signal data according to claim 3, characterized in that: In the process of building the neural network model, m radio signals are collected, and the number of m is not less than n×n; the difference between the original radio signal X=(x(1), x(2), ..., x(n)) and the decoded radio signal X'=(x'(1), x'(2), ..., x(n)) 5. A method for compressing radio signal data according to claim 4, characterized in that: In the process of compressing the radio signal, the category of the radio signal is obtained according to the pre-examination classification standard, and the search path of the radio signal is generated according to the timestamp corresponding to the compression process of the radio signal, and conditional mapping is performed with the database.
6. A method for compressing radio signal data according to claim 5, characterized in that: In the process of extracting the radio signal, according to the input search conditions and the search path, the compressed radio signal is obtained from the database, and the radio signal is decompressed by the decoder of the neural network model to extract the radio signal.
7. A data compression system for radio signals, characterized in that include: A model building module is used to build a neural network consisting of an encoder and a decoder, and train it through radio signal samples to generate a neural network model that meets the error requirements; A compression module, used for calling the encoder of the neural network model to compress the radio signal and store it in a database; A decompression module is used to call the compressed radio signal from the database, decompress it through the decoder of the neural network model, and obtain the radio signal.
8. A data compression system for radio signals according to claim 7, characterized in that: The model construction module is also used to construct the neural network in a fully connected manner, wherein the encoder and the decoder have the same number of layers; the encoder has 5 layers, the number of input layer nodes is n, the number of first layer nodes is n / 2, the number of second layer nodes is n / 4, the number of third layer nodes is n / 8, and the number of output layer nodes is n / 16; the decoder has 5 layers, the number of input layer nodes is n / 16, the number of first layer nodes is n / 8, the number of second layer nodes is n / 4, the number of third layer nodes is n / 2, and the number of output layer nodes is n.
9. A data compression system for radio signals according to claim 8, characterized in that: The compression module is further used to obtain the category of the radio signal according to the pre-examination classification standard, and generate a search path for the radio signal according to the timestamp corresponding to the compression process of the radio signal, and perform conditional mapping with the database.
10. A data compression system for radio signals according to claim 9, characterized in that: The decompression module is also used to obtain the compressed radio signal in the database according to the input search conditions and the search path, decompress it through the decoder of the neural network model, and extract the radio signal.