Search

CN-122027412-A - On-line modulation identification method and system for channel robustness

CN122027412ACN 122027412 ACN122027412 ACN 122027412ACN-122027412-A

Abstract

The invention provides a channel steady on-line modulation identification method and system, which relate to the technical field of wireless communication and are used for carrying out modulation irrelevant blind synchronization pretreatment on an original received signal to obtain a received signal after blind synchronization, acquiring a transmission signal estimation with unknown modulation types by using a preset channel blind equalization module based on the received signal after blind synchronization and a channel blind equalizer, acquiring a modulation type estimation and a quantization estimation of the transmission signal by using a preset quantization identification module based on the transmission signal estimation with unknown modulation types and a standard constellation set of all possible modulation types, simultaneously training the quantization identification module and the channel blind equalization module, and updating the parameters of the quantization identification module and the parameters of the channel blind equalization module until the parameters of the quantization identification module and the parameters of the channel blind equalization module meet iteration update termination conditions. The invention can realize high-precision online identification of the stable modulation type of the channel under the multipath fading channel environment.

Inventors

  • LI YANGQING
  • YANG CAN

Assignees

  • 中国电子科技集团公司第七研究所

Dates

Publication Date
20260512
Application Date
20260212

Claims (10)

  1. 1. A method for channel robust on-line modulation identification, comprising the steps of: S1, performing blind synchronization pretreatment irrelevant to modulation on an original received signal to obtain a received signal after blind synchronization; s2, based on the received signals after blind synchronization and the constructed channel blind equalizer, acquiring a transmission signal estimation with unknown modulation type of blind equalization estimation by using a preset channel blind equalization module; S3, based on the unknown modulation type of the transmitted signal estimation and the standard constellation set of all possible modulation types, taking the minimum sum of the constellation distance statistical error and the constellation probability distribution entropy under the assumption of different modulation types as a target, acquiring the modulation type estimation and the quantization estimation of the transmitted signal by using a preset quantization identification module, training the quantization identification module and a channel blind equalization module, and updating the parameters of the quantization identification module and the parameters of the channel blind equalization module; And S4, judging whether the quantized identification module parameters and the channel blind equalization module parameters meet the iteration update termination conditions, if so, outputting modulation type estimation and quantized estimation of the transmitted signals, and otherwise, returning to the step S2.
  2. 2. The method for identifying on-line modulation of channel robustness according to claim 1, wherein in S1, the blind synchronization pretreatment irrelevant to modulation is performed on the original received signal to obtain a received signal after blind synchronization, the process is as follows: S11, estimating carrier frequency offset of an original received signal by using a modulation-independent blind frequency offset estimation algorithm, and performing frequency offset compensation; s12, estimating the symbol rate of the signal processed in the step S11 by using a blind symbol rate estimation algorithm irrelevant to modulation, and performing symbol sampling synchronization; s13, carrying out matched filtering on the signal processed in the step S12 by using a matched filter to obtain a matched and filtered signal; S14, correcting sampling deviation and sampling frequency deviation by using a blind symbol timing synchronization algorithm irrelevant to modulation based on the signal after matching and filtering to obtain a signal after timing synchronization; and S15, correcting carrier phase rotation deviation by using a modulation-independent blind carrier phase estimation method based on the signals after timing synchronization, and locking carrier phases to obtain received signals after blind synchronization.
  3. 3. The method for identifying on-line modulation of channel robustness according to claim 1, wherein in S2, based on the blind synchronized received signal and the constructed channel blind equalizer, a preset channel blind equalization module is used to obtain a transmission signal estimate with unknown modulation type of the blind equalization estimate, the process is as follows: Obtaining a residual vector of a tap vector of a channel blind equalizer, wherein the residual vector expression is as follows: Wherein, the Blind equalizer for channels Is used for initializing the estimation of the number; based on the received signal after blind synchronization and residual vector, obtaining the estimation of the transmission signal with unknown modulation type The expression is satisfied: Wherein, the Representing the received signal after the blind synchronization, Is provided with The transmitted signal vectors of the individual co-distributed modulation symbols, The transpose of the representation vector is performed, Taken from the modulation type Standard constellation set Modulation type Belonging to a set of all possible modulation types ; Is the equivalent channel vector of the channel, Is the total number of taps of the equivalent channel vector; is a noise vector; Is fixed phase shift; A first weight factor and a second weight factor respectively; A convolution operation is represented and is performed, Representing the complex domain.
  4. 4. A method of channel-robust on-line modulation identification as claimed in claim 3, wherein said constructed channel blind equalizer Has the following characteristics of The tap vectors are expressed as: And (2) and 。
  5. 5. The method for on-line modulation and identification of channel robustness according to claim 3, wherein in S3, based on the unknown transmission signal estimation of modulation type and the standard constellation set of all possible modulation types, with the minimum sum of the constellation distance statistical error and constellation probability distribution entropy under different modulation type assumptions as the target, the modulation type estimation and quantization estimation of the transmission signal are obtained by using a preset quantization and identification module, and simultaneously, the quantization and identification module and the channel blind equalization module are trained, and the quantization and identification module parameters and the channel blind equalization module parameters are updated, and the process is as follows: s31, estimating the transmission signal based on the unknown modulation type And standard constellation set for all possible modulation types Statistical error by minimizing constellation distance Entropy of constellation probability distribution Summing, obtaining a quantized estimate of the transmitted signal And modulation type estimation The expression is satisfied: Wherein, the To utilize channel blind equalization module pairs Blind equalization to obtain a transmission signal The posterior distribution of the estimates is calculated, Representing the information entropy; Representing modulation type The corresponding standard constellation set is used to determine, For the modulation type The number of elements of the standard constellation set, Representation of Is the first of (2) Each symbol point Equal to per modulation type Constellation point after phase offset correction Probability of (2); s32, constructing a loss function, estimating based on the received signal after blind synchronization and the transmission signal with unknown modulation type And quantized estimation of transmitted signal Calculating a loss function; And S33, updating the learning rate according to the loss function, and updating the quantized identification module parameters and the channel blind equalization module parameters based on the learning rate for iterative updating.
  6. 6. The method for channel-robust on-line modulation identification of claim 5, wherein said loss function is expressed as: Wherein, the In order to lose the weight factor of the function, To transmit signals With respect to Is used for the true posterior distribution of (c), To utilize channel blind equalization module pairs Blind equalization to obtain a transmission signal The posterior distribution of the estimates is calculated, Representation of And (3) with KL divergence of (c).
  7. 7. The method for channel-robust on-line modulation identification of claim 6, it is characterized in that the method comprises the steps of, And (3) with KL divergence of (2) The expression is: Wherein, the Is a desire for approximation of variation.
  8. 8. An electronic device comprising a memory, a processor and a program stored on the memory and executable on the processor, characterized in that the processor is arranged to read the program in the memory to implement the steps in the channel robust on-line modulation identification method according to any of claims 1 to 7.
  9. 9. A computer storage medium storing a program, wherein the program when executed by a processor implements the steps of the channel robust on-line modulation identification method according to any of claims 1 to 7.
  10. 10. A channel-robust on-line modulation identification system for implementing a channel-robust on-line modulation identification method according to any of claims 1-7, comprising: the preprocessing unit is used for carrying out blind synchronization preprocessing irrelevant to modulation on the original received signal to obtain a received signal after blind synchronization; The channel blind equalization unit is used for acquiring a transmission signal estimation with unknown modulation type of blind equalization estimation by using a preset channel blind equalization module based on the received signal after blind synchronization and the constructed channel blind equalizer; The quantization identification and module training unit is used for acquiring the modulation type estimation and the quantization estimation of the transmission signal by using a preset quantization identification module based on the transmission signal estimation with unknown modulation types and a standard constellation set of all possible modulation types, taking the minimum sum of the constellation distance statistical error and the constellation probability distribution entropy under the assumption of different modulation types as a target, and simultaneously training the quantization identification module and the channel blind equalization module to update the parameters of the quantization identification module and the parameters of the channel blind equalization module; And the circulating unit is used for judging whether the parameter of the quantization identification module and the parameter of the channel blind equalization module meet the iteration update termination condition, if so, outputting the modulation type estimation and the quantization estimation of the transmitted signal, and if not, returning to the channel blind equalization unit.

Description

On-line modulation identification method and system for channel robustness Technical Field The present invention relates to the field of wireless communications, and in particular, to a method and system for identifying on-line modulation with robust channel. Background In the radio fields of cognitive radio, spectrum sensing, electronic investigation and the like, the modulation type of a wireless communication signal needs to be rapidly and adaptively identified so as to support and realize applications such as cognitive communication, detection decoding and the like. At present, modulation identification is mainly realized by a likelihood class and a feature class method. For likelihood class algorithms, probability density functions of received signals are typically used to evaluate the likelihood of each possible modulation pattern hypothesis, and recognition results are obtained by maximizing the likelihood functions, while feature class algorithms extract signal features based on domain expert knowledge or based on depth networks, etc., and make classification recognition based on the differences of signal features of different modulation types. However, when applied to an actual multipath fading channel environment, the above methods can have serious degradation of recognition accuracy due to influence of undesirable channel effects (such as intersymbol interference, amplitude fading, frequency offset, etc.) in the signal. Existing methods based on deep learning and the like are mainly aimed at simplifying the task of modulation identification under a channel (such as an ideal channel or an additive white gaussian noise channel). Due to the influence of multipath fading channels, these methods have lower recognition accuracy in an actual open scene. If the signal modulation type is unknown, it is generally difficult to perform channel blind equalization in advance by effectively utilizing the statistical characteristics (such as cyclostationarity, constant modulus characteristics, constellation diagram, etc.) of the signal to eliminate the multipath channel effect. In addition, the modulation recognition method based on deep migration learning can adapt to the influence caused by the uncertainty of the multipath fading channel in an actual scene to a certain extent, but the recognition accuracy of the method is different from that of the current deep learning modulation recognition method based on complete data set offline training. Disclosure of Invention In order to solve the problem of low accuracy of wireless communication signal modulation type identification in the prior art, the invention provides a channel-stable online modulation identification method and system, which are used for realizing high-accuracy and real-time modulation identification of wireless communication signal modulation types in a multipath fading channel environment. In order to achieve the technical effects, the technical scheme of the invention is as follows: The application provides a channel steady on-line modulation identification method, which comprises the following steps: S1, performing blind synchronization pretreatment irrelevant to modulation on an original received signal to obtain a received signal after blind synchronization; s2, based on the received signals after blind synchronization and the constructed channel blind equalizer, acquiring a transmission signal estimation with unknown modulation type of blind equalization estimation by using a preset channel blind equalization module; S3, based on the unknown modulation type of the transmitted signal estimation and the standard constellation set of all possible modulation types, taking the minimum sum of the constellation distance statistical error and the constellation probability distribution entropy under the assumption of different modulation types as a target, acquiring the modulation type estimation and the quantization estimation of the transmitted signal by using a preset quantization identification module, training the quantization identification module and a channel blind equalization module, and updating the parameters of the quantization identification module and the parameters of the channel blind equalization module; And S4, judging whether the quantized identification module parameters and the channel blind equalization module parameters meet the iteration update termination conditions, if so, outputting modulation type estimation and quantized estimation of the transmitted signals, and otherwise, returning to the step S2. Preferably, in S1, a blind synchronization preprocessing is performed on an original received signal, where the blind synchronization preprocessing is performed on the original received signal, so as to obtain a received signal after blind synchronization, where the process is as follows: S11, estimating carrier frequency offset of an original received signal by using a modulation-independent blind frequency offset estimati