CN112202529B - Wireless communication information recovery method and device and electronic equipment - Google Patents
Wireless communication information recovery method and device and electronic equipment Download PDFInfo
- Publication number
- CN112202529B CN112202529B CN202011034205.3A CN202011034205A CN112202529B CN 112202529 B CN112202529 B CN 112202529B CN 202011034205 A CN202011034205 A CN 202011034205A CN 112202529 B CN112202529 B CN 112202529B
- Authority
- CN
- China
- Prior art keywords
- neural network
- convolutional neural
- deep convolutional
- bit stream
- signal
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L1/00—Arrangements for detecting or preventing errors in the information received
- H04L1/0001—Systems modifying transmission characteristics according to link quality, e.g. power backoff
- H04L1/0006—Systems modifying transmission characteristics according to link quality, e.g. power backoff by adapting the transmission format
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G06N3/045—Combinations of networks
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/08—Learning methods
- G06N3/084—Backpropagation, e.g. using gradient descent
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Computing Systems (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Data Mining & Analysis (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- Molecular Biology (AREA)
- Artificial Intelligence (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Health & Medical Sciences (AREA)
- Quality & Reliability (AREA)
- Computer Networks & Wireless Communication (AREA)
- Signal Processing (AREA)
- Error Detection And Correction (AREA)
Abstract
The application discloses a wireless communication information recovery method and device and electronic equipment. The method comprises the following steps: acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or an actual acquisition mode; constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream; carrying out information recovery training on the deep convolutional neural network by using a plurality of signal samples to obtain a trained deep convolutional neural network; and inputting the signal to be recovered to the trained deep convolution neural network, and outputting and recovering the obtained bit stream sequence by the trained deep convolution neural network. The embodiment of the application recovers the wireless communication information based on the deep convolutional neural network, can improve the information recovery accuracy rate in a complex environment, can be widely suitable for information recovery of various communication systems, and has higher universality.
Description
Technical Field
The present application relates to the field of wireless communication technologies, and in particular, to a method and an apparatus for recovering wireless communication information, and an electronic device.
Background
In a wireless communication system, information to be transmitted generally needs to be subjected to steps of channel coding, modulation, pulse forming, up-conversion, amplification and the like, and then signals are radiated outside through an antenna, the signals are transmitted to a wireless communication receiving end through a wireless communication channel, and the signals received by the wireless communication receiving end are often distorted. In order to recover original information as accurately as possible, a conventional wireless communication receiving end usually elaborately designs various algorithms (including carrier synchronization, demodulation, channel decoding and the like) based on a certain theoretical model to realize information recovery. However, due to the influence of various non-linear factors of the rf circuit, and the unpredictable nature of dynamic changes of the factors such as channel noise, doppler shift, multipath fading, and interference, the performance of the conventional information recovery method is greatly affected. In addition, the conventional information recovery method is often designed for a specific transmission signal system, and the same algorithm is difficult to adapt to various communication systems. Therefore, an information recovery method that can improve the accuracy of information recovery in a complex environment and that can be widely applied to various communication systems is required.
Disclosure of Invention
The embodiment of the application provides a wireless communication information recovery method, a wireless communication information recovery device and electronic equipment, which can improve the information recovery accuracy rate in a complex environment, can be widely suitable for information recovery of multiple communication systems, and have universality.
According to a first aspect of the present application, there is provided a wireless communication information recovery method, including:
acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or in an actual acquisition mode;
constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream;
carrying out information recovery training on the deep convolutional neural network by using the plurality of signal samples to obtain a trained deep convolutional neural network;
and inputting a signal to be recovered to the trained deep convolution neural network, and outputting and recovering the obtained bit stream sequence by the trained deep convolution neural network.
According to a second aspect of the present application, there is provided a wireless communication information recovery apparatus including:
the system comprises a sample acquisition unit, a data acquisition unit and a data processing unit, wherein the sample acquisition unit is used for acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or an actual acquisition mode;
the network construction unit is used for constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and the final classification layer of the network is designed according to the length of an information bit stream;
the training unit is used for performing information recovery training on the deep convolutional neural network by using the plurality of signal samples to obtain a trained deep convolutional neural network;
and the recovery unit is used for inputting a signal to be recovered to the trained deep convolutional neural network and outputting a recovered bit stream sequence by the trained deep convolutional neural network.
In accordance with a third aspect of the present application, there is provided an electronic device comprising: a processor, a memory storing computer-executable instructions,
the executable instructions, when executed by the processor, implement the aforementioned wireless communication information recovery method.
According to a fourth aspect of the present application, there is provided a computer-readable storage medium storing one or more programs which, when executed by a processor, implement the aforementioned wireless communication information recovery method.
The beneficial effect of this application is:
according to the method and the device for recovering the wireless communication information, the wireless communication information is recovered based on the deep convolutional neural network, deep learning is automatically performed from a plurality of signal samples, the method and the device are more matched with the complex situation actually experienced by a communication system, and the information recovery accuracy rate under the complex environment can be improved; meanwhile, the plurality of signal samples have a plurality of communication systems, the deep convolutional neural network is trained by using the signal samples, and the trained deep convolutional neural network can be used for carrying out information recovery on signals of the communication systems, so that the method and the device can be widely applied to information recovery of the plurality of communication systems and have universality.
The above description is only an overview of the technical solutions of the present application, and the present application may be implemented in accordance with the content of the description so as to make the technical means of the present application more clearly understood, and the detailed description of the present application will be given below in order to make the above and other objects, features, and advantages of the present application more clearly understood.
Drawings
The accompanying drawings, which are included to provide a further understanding of the application and are incorporated in and constitute a part of this application, illustrate embodiment(s) of the application and together with the description serve to explain the application and not to limit the application. In the drawings:
fig. 1 is a flowchart illustrating a method for recovering wireless communication information according to an embodiment of the present application;
FIG. 2 is a block diagram of a deep convolutional neural network constructed in accordance with one embodiment of the present application;
FIG. 3 is a block diagram of a deep convolutional neural network constructed in accordance with another embodiment of the present application;
FIG. 4 is a block diagram of a wireless communication information recovery apparatus according to an embodiment of the present application;
fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application.
Detailed Description
In order to make the objects, technical solutions and advantages of the present application more apparent, the technical solutions of the present application will be described in detail and completely with reference to the following specific embodiments of the present application and the accompanying drawings. It should be apparent that the described embodiments are only some of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present application.
Fig. 1 is a schematic flowchart of a method for recovering wireless communication information according to an embodiment of the present application, and referring to fig. 1, the method for recovering wireless communication information according to the embodiment includes the following steps:
step S101, a plurality of signal samples with a plurality of communication systems are obtained in a simulation mode and/or an actual collection mode.
In step S101, in order to obtain a plurality of sample data, a plurality of signal samples having a plurality of communication systems may be obtained in a simulation manner and/or in an actual acquisition manner.
When signal samples are generated in simulation, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The information bit stream is subjected to channel coding, modulation and pulse forming, then subjected to certain channel fading and added with certain distributed noise, and subjected to power normalization to obtain a signal IQ sequence y i The signal IQ sequence is also called complex baseband signal sequence, thus composed of (x) i ,y i ) Forming a signal sample, wherein i is a natural number; the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems. It should be noted that, according to a specific communication system, the channel coding manner may be a BCH code, a convolutional code, and the like, and the code rate is set as needed; the modulation mode can be 2ASK, 4ASK, 2FSK, 4FSK, 8FSK, MSK, BPSK, QPSK, 8PSK, 16QAM, 64QAM, 128QAM, 256QAM, 512QAM, 1024QAM and the like; the pulse forming mode can be low-pass filtering, root raised cosine filtering, raised cosine filtering and the like, and the roll-off coefficient is set according to the requirement; the channel fading can be a Rayleigh fading model, a Rice fading model and the like; the noise can be additive white Gaussian noise, and the noise power is set according to the range of the signal-to-noise ratio.
Wherein, when actually collecting signal samples, randomly selecting a certain communication system to generate an information bit stream x with a certain length i The IQ sequence y is obtained by radiating the IQ sequence into the air through a transmitter of a selected communication system, transmitting the IQ sequence through a wireless channel, reaching a receiver and digitizing the IQ sequence by the receiver i Is composed of (x) i ,y i ) Forming a signal sample, wherein i is a natural number; the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems. It should be noted that, in order to make the acquired signal samples experience various wireless channel environments, the positions of the transmitter and the receiver may be changed, or the transmitter and the receiver may be placed on a moving platform to acquire the signal samples under various position distributions.
Furthermore, in order to obtain as large a sufficient number of signal samples as possible, the signal samples generated by the simulation may be combined with the actually acquired signal samples to form a complete set of signal samples.
And S102, constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream.
In the deep convolutional neural network constructed in step S102, a certain layer in the middle of the network adopts a global maximum pooling design, and the advantage of adopting the global maximum pooling is that the network can adapt to different input signal lengths, and the final classification layer of the network is designed according to the information bit stream length.
When the information bit stream lengths of all the signal samples are the same, the last classification layer of the constructed deep convolutional neural network A is M1 classifiers, wherein M1 is the information bit stream length. FIG. 2 is a block diagram of a deep convolutional neural network constructed in accordance with one embodiment of the present application, which includes three convolutional layers. In fig. 2, the I and Q paths of the IQ signal form a matrix of Nx 2as the network input, where N is the signal length (N =15 in fig. 2). "conv" denotes the convolution layer, the first number of which denotes the size of the convolution kernel and the last number of which denotes the number of convolution kernels. "Batch Normalization" refers to a Batch Normalization operation; "ReLU" means rectified linear activation; "Global MaxPool" indicates Global maximum pooling; "Dropout" represents a Dropout layer, where the number in parentheses represents the Dropout probability; "SoftMax" denotes the SoftMax layer, the following numbers of which denote the number of hidden nodes. The output of the ith softmax corresponds to the probability that the ith bit takes values of 0 and 1, i =1,2 1 。
When the lengths of the information bit streams of the signal samples are different, the last classification layer of the constructed deep convolutional neural network B is 1+ M2 classifiers, wherein the first classifier is used for judging the length of the information bit stream, the number of classes of the first classifier corresponds to the number of all possible information bit stream lengths, the next M2 classifiers are classifiers, and M2 is the maximum possible information bit stream length. FIG. 3 is a block diagram of a deep convolutional neural network constructed in accordance with another embodiment of the present application, which also includes three convolutional layers. In fig. 3, the I and Q paths of the IQ signal form a matrix of Nx 2as the network input, where N is the signal length (N =15 in fig. 3). "conv" means volumeA pad whose front number represents the size of the convolution kernel and whose rear number represents the number of convolution kernels; "Batch Normalization" refers to a Batch Normalization operation; "ReLU" means rectified linear activation; "Global Maxpool" represents Global maximal pooling; "Dropout" represents a Dropout layer, where the numbers in parentheses represent Dropout probabilities; "SoftMax" denotes the SoftMax layer, the numbers following it denote the number of hidden nodes; j denotes the number of possible information sequence length classes in the signal sample set. The output of the 1 st softmax corresponds to the probability that the information bit stream length takes a corresponding value, the output of the ith softmax corresponds to the probability that the i-1 st bit takes values of 0 and 1, i =2 2 。
It should be noted that step S101 and step S102 are in parallel, and there is no requirement in the execution order.
And S103, performing information recovery training on the deep convolutional neural network by using a plurality of signal samples to obtain the trained deep convolutional neural network.
In step S103, T (T is greater than 1 and less than L) samples are selected from L sample data as training data, the rest are used as verification data, the training data is used to perform information recovery training on the deep convolutional neural network, and the training method adopts a stochastic gradient descent method in a back propagation algorithm. The random gradient descent method is the most classical, basic and common method in the field, and has the main advantage of faster convergence rate.
After training for a certain time (a certain number of iterations) by using the training data, evaluating whether the output error of the trained deep convolutional neural network meets the requirement by using the verification data, for example, whether the output error of each time is smaller than a preset value, if so, stopping training and considering that the trained deep convolutional neural network is obtained.
And step S104, inputting the signal to be recovered to the trained deep convolutional neural network, and outputting and recovering the obtained bit stream sequence by the trained deep convolutional neural network.
In step S104, the signal to be recovered is input to the trained deep convolutional neural network, and the output of the trained deep convolutional neural network is the information sequence of the recovered signal.
For the deep convolutional neural network A, the cascade of bit values corresponding to the categories with higher confidence coefficient output by the M1 classifiers is the bit stream sequence obtained by recovery; for the deep convolutional neural network B, the class corresponding to the maximum confidence value output by the first classifier is the length P (P is less than or equal to M) of the bit stream sequence 2 ) The cascade of bit values corresponding to the larger confidence value output by the 2 nd to the P +1 th classifiers is the bit stream sequence obtained by recovery.
As shown in fig. 1, in the wireless communication information recovery method according to the embodiment of the present application, wireless communication information is recovered based on a deep convolutional neural network, deep learning is automatically performed from a plurality of signal samples, and the method is more matched with a complex situation actually experienced by a communication system, so that the information recovery accuracy in a complex environment can be improved; meanwhile, the plurality of signal samples have various communication systems, the deep convolution neural network is trained by using the signal samples, and the trained deep convolution neural network can be used for carrying out information recovery on signals of the communication systems, so that the method can be widely applied to the information recovery of the various communication systems and has universality.
The wireless communication information recovery method belongs to the same technical concept as the wireless communication information recovery method, and the embodiment of the application also provides a wireless communication information recovery device. Fig. 4 is a block diagram of a wireless communication information recovery apparatus according to an embodiment of the present application, and referring to fig. 4, the wireless communication information recovery apparatus according to the present embodiment includes:
a sample acquiring unit 401, configured to acquire a plurality of signal samples with a plurality of communication systems in a simulation manner and/or in an actual acquisition manner;
a network construction unit 402, configured to construct a deep convolutional neural network suitable for signal processing, where a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream;
the training unit 403 is configured to perform information recovery training on the deep convolutional neural network by using multiple signal samples, so as to obtain a trained deep convolutional neural network;
and a recovery unit 404, configured to input a signal to be recovered to the trained deep convolutional neural network, and output a recovered bit stream sequence by the trained deep convolutional neural network.
In an embodiment of the present application, the sample acquiring unit 401 is specifically configured to:
when signal samples are generated in simulation, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The information bit stream is subjected to channel coding, modulation and pulse forming, then subjected to certain channel fading and added with certain distributed noise, and subjected to power normalization to obtain a signal IQ sequence y i Is composed of (x) i ,y i ) Forming a signal sample, wherein i is a natural number; repeating the above process for multiple times to obtain multiple signal samples with multiple communication systems; and/or the presence of a gas in the atmosphere,
when actually collecting signal samples, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The IQ sequence y is obtained by radiating the IQ sequence into the air through a transmitter of a selected communication system, transmitting the IQ sequence through a wireless channel, reaching a receiver and digitizing the IQ sequence by the receiver i From (x) i ,y i ) Forming a signal sample, wherein i is a natural number; the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems.
In an embodiment of the present application, the training unit 403 is specifically configured to:
selecting T (T is more than 1 and less than L) from L sample data as training data, taking the rest as verification data, performing information recovery training on the deep convolutional neural network by using the training data, adopting a random gradient descent method in a back propagation algorithm by using the training method, evaluating whether the output error of the trained deep convolutional neural network meets the requirement by using the verification data, and stopping training if the output error meets the requirement to obtain the trained deep convolutional neural network.
In an embodiment of the present application, when the information bit stream lengths of all the signal samples are the same, the last classification layer of the deep convolutional neural network a constructed by the network construction unit 402 is M1 classifiers, where M1 is the information bit stream length; the recovery unit 404 inputs the signal to be recovered to the trained deep convolutional neural network a, and the cascade of the bit values corresponding to the class with the higher confidence output by the M1 classifiers is the recovered bit stream sequence.
In an embodiment of the present application, when the lengths of the information bit streams of all the signal samples are different, the last classification layer of the deep convolutional neural network B constructed by the network construction unit 402 is 1+ M2 classifiers, where the first classifier is used to determine the length of the information bit stream, the number of classes thereof corresponds to the number of all possible information bit stream lengths, the last M2 classifiers are binary classifiers, and M2 is the maximum possible information bit stream length; the recovery unit 404 inputs the signal to be recovered to the trained deep convolutional neural network B, and the class corresponding to the maximum confidence value output by the first classifier is the length P of the bit stream sequence (P is less than or equal to M) 2 ) The cascade of the bit values corresponding to the higher confidence values output by the 2 nd to the P +1 th classifiers is the bit stream sequence obtained by recovery.
It should be noted that, the exemplary explanation about the functions performed by the units in the wireless communication information recovery apparatus shown in fig. 4 is consistent with the exemplary explanation in the foregoing method embodiment, and is not repeated here.
In summary, the method and the device for recovering wireless communication information in the embodiment of the application recover the wireless communication information based on the deep convolutional neural network, automatically perform deep learning from a plurality of signal samples, are more matched with the complex situation actually experienced by a communication system, and can improve the accuracy of information recovery in a complex environment; meanwhile, the plurality of signal samples have a plurality of communication systems, the deep convolutional neural network is trained by using the signal samples, and the trained deep convolutional neural network can be used for carrying out information recovery on signals of the communication systems, so that the method and the device can be widely applied to information recovery of the plurality of communication systems and have universality.
Fig. 5 is a schematic structural diagram of an electronic device according to an embodiment of the present application. Referring to fig. 5, at a hardware level, the electronic device includes a memory and a processor, and optionally further includes an interface module, a communication module, and the like. The Memory may include a Memory, such as a Random-Access Memory (RAM), and may also include a non-volatile Memory, such as at least one disk Memory. Of course, the electronic device may also include hardware required for other services.
The processor, the interface module, the communication module, and the memory may be connected to each other via an internal bus, which may be an ISA (Industry Standard Architecture) bus, a PCI (Peripheral Component Interconnect) bus, an EISA (Extended Industry Standard Architecture) bus, or the like. The bus may be divided into an address bus, a data bus, a control bus, etc. For ease of illustration, only one double-headed arrow is shown in FIG. 5, but this does not indicate only one bus or one type of bus.
A memory for storing computer executable instructions. The memory provides computer executable instructions to the processor through the internal bus.
A processor executing computer executable instructions stored in the memory and specifically configured to perform the following operations:
acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or an actual acquisition mode;
constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream;
carrying out information recovery training on the deep convolutional neural network by using a plurality of signal samples to obtain a trained deep convolutional neural network;
and inputting the signal to be recovered to the trained deep convolutional neural network, and outputting and recovering the obtained bit stream sequence by the trained deep convolutional neural network.
The functions performed by the wireless communication information recovery apparatus according to the embodiment shown in fig. 4 of the present application may be implemented in or by a processor. The processor may be an integrated circuit chip having signal processing capabilities. In implementation, the steps of the above method may be performed by integrated logic circuits of hardware in a processor or instructions in the form of software. The Processor may be a general-purpose Processor, including a Central Processing Unit (CPU), a Network Processor (NP), and the like; but also Digital Signal Processors (DSPs), application Specific Integrated Circuits (ASICs), field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components. The various methods, steps, and logic blocks disclosed in the embodiments of the present application may be implemented or performed. A general purpose processor may be a microprocessor or the processor may be any conventional processor or the like. The steps of the method disclosed in connection with the embodiments of the present application may be directly implemented by a hardware decoding processor, or implemented by a combination of hardware and software modules in the decoding processor. The software module may be located in ram, flash memory, rom, prom, or eprom, registers, etc. storage media as is well known in the art. The storage medium is located in a memory, and a processor reads information in the memory and completes the steps of the method in combination with hardware of the processor.
The electronic device may further perform the steps performed by the wireless communication information recovery method in fig. 1, and implement the functions of the wireless communication information recovery method in the embodiment shown in fig. 1, which are not described herein again.
An embodiment of the present application further provides a computer-readable storage medium, which stores one or more programs that, when executed by a processor, implement the foregoing non-linearity correction method for an analog-to-digital converter, and are specifically configured to perform:
acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or in an actual acquisition mode;
constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream;
carrying out information recovery training on the deep convolutional neural network by using a plurality of signal samples to obtain a trained deep convolutional neural network;
and inputting the signal to be recovered to the trained deep convolution neural network, and outputting and recovering the obtained bit stream sequence by the trained deep convolution neural network.
As will be appreciated by one skilled in the art, embodiments of the present application may be provided as a method, system, or computer program product. Accordingly, the present application may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, the present application may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) that include computer-usable program code.
The present application is described in terms of flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the application. It will be understood that each flow and/or block of the flowchart illustrations and/or block diagrams, and combinations of flows and/or blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both permanent and non-permanent, removable and non-removable media, may implement the information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), static Random Access Memory (SRAM), dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), read Only Memory (ROM), electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), digital Versatile Disks (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium, which can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
The algorithms and displays presented herein are not inherently related to any particular computer, virtual machine, or other apparatus. Various general purpose devices may also be used with the teachings herein. The required structure for constructing an arrangement of this type will be apparent from the description above. In addition, this application is not directed to any particular programming language. It will be appreciated that a variety of programming languages may be used to implement the teachings of the present application as described herein, and any descriptions of specific languages are provided above to disclose the best modes of the present application.
It should be noted that the above-mentioned embodiments illustrate rather than limit the application, and that those skilled in the art will be able to design alternative embodiments without departing from the scope of the appended claims. In the claims, any reference signs placed between parentheses shall not be construed as limiting the claim. The word "comprising" does not exclude the presence of elements or steps not listed in a claim. The word "a" or "an" preceding an element does not exclude the presence of a plurality of such elements. The application may be implemented by means of hardware comprising several distinct elements, and by means of a suitably programmed computer.
Claims (8)
1. A method for recovering wireless communication information, the method comprising:
acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or in an actual acquisition mode;
constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and a final classification layer of the network is designed according to the length of an information bit stream;
performing information recovery training on the deep convolutional neural network by using the plurality of signal samples to obtain a trained deep convolutional neural network;
inputting a signal to be recovered to the trained deep convolutional neural network, and outputting a recovered bit stream sequence by the trained deep convolutional neural network; wherein the content of the first and second substances,
when the lengths of the information bit streams of all the signal samples are the same, the last classification layer of the constructed deep convolutional neural network A is M1 classifiers, wherein M1 is the length of the information bit stream; inputting a signal to be recovered to the trained deep convolutional neural network A, wherein the cascade connection of bit values corresponding to the categories with higher confidence coefficient output by the M1 classifiers is a recovered bit stream sequence;
when the lengths of the information bit streams of all the signal samples are different, the last classification layer of the constructed deep convolutional neural network B is 1+ M2 classifiers, wherein the first classifier is used for judging the length of the information bit stream, the number of classes of the first classifier corresponds to the number of all possible information bit stream lengths, the following M2 classifiers are two classifiers, and M2 is the maximum possible information bit stream length; inputting the signal to be recovered to the trained deep convolutional neural network B, and outputting the class corresponding to the maximum confidence coefficient value by the first classifier as the length P (P is less than or equal to M) of the bit stream sequence 2 ) The cascade of the bit values corresponding to the higher confidence values output by the 2 nd to the P +1 th classifiers is the bit stream sequence obtained by recovery.
2. The method of claim 1, wherein the obtaining the plurality of signal samples in the simulated manner having the plurality of communication regimes comprises:
when signal samples are generated in simulation, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The information bit stream is subjected to channel coding, modulation and pulse forming, then subjected to certain channel fading and added with certain distributed noise, and subjected to power normalization to obtain a signal IQ sequence y i Is composed of (x) i ,y i ) Forming a signal sample, wherein i is a natural number;
the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems.
3. The method of claim 1, wherein said obtaining a plurality of signal samples having a plurality of communication regimes in an actual acquisition manner comprises:
actual collectionWhen signal samples are taken, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The IQ sequence y is obtained by radiating the IQ sequence in the air by a transmitter of a selected communication system, transmitting the IQ sequence through a wireless channel, reaching a receiver and digitizing by the receiver i From (x) i ,y i ) Forming a signal sample, wherein i is a natural number;
the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems.
4. The method of claim 1, wherein the training of information recovery for the deep convolutional neural network using the plurality of signal samples, resulting in a trained deep convolutional neural network comprises:
selecting T (T is more than 1 and less than L) from L sample data as training data, taking the rest as verification data, utilizing the training data to carry out information recovery training on the deep convolutional neural network, adopting a random gradient descent method in a back propagation algorithm by the training method, utilizing the verification data to evaluate whether the output error of the deep convolutional neural network after training meets the requirement or not, and stopping training if the output error meets the requirement to obtain the trained deep convolutional neural network.
5. An apparatus for recovering wireless communication information, the apparatus comprising:
the system comprises a sample acquisition unit, a data acquisition unit and a data processing unit, wherein the sample acquisition unit is used for acquiring a plurality of signal samples with a plurality of communication systems in a simulation mode and/or an actual acquisition mode;
the network construction unit is used for constructing a deep convolutional neural network suitable for signal processing, wherein a certain layer in the middle of the network adopts global maximum pooling, and the final classification layer of the network is designed according to the length of an information bit stream;
the training unit is used for performing information recovery training on the deep convolutional neural network by using the plurality of signal samples to obtain a trained deep convolutional neural network;
a recovery unit, configured to input a signal to be recovered to the trained deep convolutional neural network, and output a recovered bit stream sequence by the trained deep convolutional neural network; wherein the content of the first and second substances,
when the lengths of the information bit streams of all the signal samples are the same, the last classification layer of the deep convolutional neural network A constructed by the network construction unit is M1 classifiers, wherein M1 is the length of the information bit stream; the recovery unit inputs a signal to be recovered to the trained deep convolutional neural network A, and the cascade of bit values corresponding to the categories with higher confidence degrees output by the M1 classifiers is a recovered bit stream sequence;
when the lengths of the information bit streams of all the signal samples are different, the last classification layer of the deep convolutional neural network B constructed by the network construction unit is 1+ M2 classifiers, wherein the first classifier is used for judging the length of the information bit stream, the number of classes of the first classifier corresponds to the number of all possible information bit stream lengths, the last M2 classifiers are binary classifiers, and M2 is the maximum possible information bit stream length; the recovery unit inputs a signal to be recovered to the trained deep convolutional neural network B, and the class corresponding to the maximum value of the output confidence of the first classifier is the length P (P is less than or equal to M) of the bit stream sequence 2 ) The cascade of the bit values corresponding to the higher confidence values output by the 2 nd to the P +1 th classifiers is the bit stream sequence obtained by recovery.
6. The apparatus of claim 5, wherein the sample acquisition unit is specifically configured to:
when signal samples are generated in simulation, a certain communication system is randomly selected to generate an information bit stream x with a certain length i The information bit stream is processed by channel coding, modulation and pulse forming, then processed by certain channel fading and added with certain distributed noise, and then processed by power normalization to obtain signal IQ sequence y i From (x) i ,y i ) Forming a signal sample, wherein i is a natural number; repeating the above process for multiple times to obtain multiple signal samples with multiple communication systems; and/or the presence of a gas in the gas,
when the signal samples are actually taken at the time of acquisition,randomly selecting a communication system to generate a bit stream x of information with a certain length i The IQ sequence y is obtained by radiating the IQ sequence in the air by a transmitter of a selected communication system, transmitting the IQ sequence through a wireless channel, reaching a receiver and digitizing by the receiver i Is composed of (x) i ,y i ) Forming a signal sample, wherein i is a natural number; the above process is repeated for a plurality of times to obtain a plurality of signal samples with a plurality of communication systems.
7. The apparatus according to claim 5, wherein the training unit is specifically configured to:
selecting T (T is more than 1 and less than L) from L sample data as training data, taking the rest as verification data, performing information recovery training on the deep convolutional neural network by using the training data, adopting a random gradient descent method in a back propagation algorithm by using the training method, evaluating whether the output error of the trained deep convolutional neural network meets the requirement by using the verification data, and stopping training if the output error meets the requirement to obtain the trained deep convolutional neural network.
8. An electronic device, comprising: a processor, a memory storing computer-executable instructions,
the executable instructions, when executed by the processor, implement the wireless communication information recovery method of any of claims 1 to 4.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011034205.3A CN112202529B (en) | 2020-09-27 | 2020-09-27 | Wireless communication information recovery method and device and electronic equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202011034205.3A CN112202529B (en) | 2020-09-27 | 2020-09-27 | Wireless communication information recovery method and device and electronic equipment |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN112202529A CN112202529A (en) | 2021-01-08 |
| CN112202529B true CN112202529B (en) | 2022-11-08 |
Family
ID=74008481
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202011034205.3A Active CN112202529B (en) | 2020-09-27 | 2020-09-27 | Wireless communication information recovery method and device and electronic equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN112202529B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN113300788B (en) * | 2021-04-19 | 2023-04-21 | 嘉兴学院 | Blind receiver method and device based on Capsule network |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6691073B1 (en) * | 1998-06-18 | 2004-02-10 | Clarity Technologies Inc. | Adaptive state space signal separation, discrimination and recovery |
| CN107171717A (en) * | 2017-05-31 | 2017-09-15 | 武汉光迅科技股份有限公司 | Recover the method and system of ideal signal in a kind of signal from distortion |
| CN110048980A (en) * | 2019-04-19 | 2019-07-23 | 中国电子科技集团公司第三十六研究所 | A kind of blind demodulation method of digital communication and device |
-
2020
- 2020-09-27 CN CN202011034205.3A patent/CN112202529B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US6691073B1 (en) * | 1998-06-18 | 2004-02-10 | Clarity Technologies Inc. | Adaptive state space signal separation, discrimination and recovery |
| CN107171717A (en) * | 2017-05-31 | 2017-09-15 | 武汉光迅科技股份有限公司 | Recover the method and system of ideal signal in a kind of signal from distortion |
| CN110048980A (en) * | 2019-04-19 | 2019-07-23 | 中国电子科技集团公司第三十六研究所 | A kind of blind demodulation method of digital communication and device |
Non-Patent Citations (2)
| Title |
|---|
| 一种基于星座图恢复的多进制相位调制信号识别算法;吴佩军，侯进，吕志良，桂梅书，张笑语，陈曾;《电讯技术》;20190531;第59卷(第5期);第60-63页 * |
| 畸变信号重建的神经网络方法;曹星平，朱炬波，易东云;《宇航计测技术》;19991031;第19卷(第5期);第549-555页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN112202529A (en) | 2021-01-08 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US8447797B2 (en) | MIMO system method and device using sorted QR-decomposition (SQRD) for detecting transmission signal with division detection | |
| CN100588193C (en) | Method and apparatus for calculating log-likelihood ratio for decoding in receiver for mobile communication system | |
| US11477060B2 (en) | Systems and methods for modulation classification of baseband signals using attention-based learned filters | |
| US11949544B2 (en) | Deep learning-based polymorphic platform | |
| EP3614637B1 (en) | Systems and methods for adjusting the sample timing of a gfsk modulated signal | |
| CN108540267B (en) | Multi-user data information detection method and device based on deep learning | |
| CN103166903B (en) | The soft solution preprocess method of constellation mapping and soft solution method | |
| US20230299872A1 (en) | Neural Network-Based Communication Method and Related Apparatus | |
| Alimohammad et al. | FPGA-based bit error rate performance measurement of wireless systems | |
| CN112202529B (en) | Wireless communication information recovery method and device and electronic equipment | |
| US10149181B2 (en) | Signal output apparatus, board, and signal output method | |
| CN107864029A (en) | A kind of method for reducing Multiuser Detection complexity | |
| CN109462457B (en) | Polar code decoding method, decoding device and decoder | |
| Huynh-The et al. | RanNet: Learning residual-attention structure in CNNs for automatic modulation classification | |
| JP5311469B2 (en) | MIMO receiver, demodulation circuit, and signal processing program | |
| CN109660473B (en) | Spherical decoding detection method and device and computer readable storage medium | |
| CN112350736A (en) | Dynamic correction factor configuration method in LDPC decoder | |
| CN113271177B (en) | Low-density parity check code decoding method, device and system and wireless receiving equipment | |
| JP2010130271A (en) | Decoder and decoding method | |
| US20230344685A1 (en) | A communication unit for soft-decision demodulation and method therefor | |
| CN111224741B (en) | BCH code decoding method and decoder for satellite navigation and satellite navigation receiver | |
| CN113971015B (en) | UIA2 computing circuit, data processing method, chip, electronic device and storage medium | |
| CN1561005B (en) | Quick double-error correction BCH code decoder | |
| JP2012124888A (en) | Decoder and decoding method | |
| Chen et al. | Grant-Free Sparse Code Multiple Access for Uplink Massive Machine-Type Communications and Its Real-Time Receiver Design |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |