CN115829016A - Method and device for adjusting wireless modem based on neural network - Google Patents
Method and device for adjusting wireless modem based on neural network Download PDFInfo
- Publication number
- CN115829016A CN115829016A CN202211613531.9A CN202211613531A CN115829016A CN 115829016 A CN115829016 A CN 115829016A CN 202211613531 A CN202211613531 A CN 202211613531A CN 115829016 A CN115829016 A CN 115829016A
- Authority
- CN
- China
- Prior art keywords
- neural network
- model
- adjustment
- wireless modem
- parameter
- 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.)
- Granted
Links
- 238000013528 artificial neural network Methods 0.000 title claims abstract description 308
- 238000000034 method Methods 0.000 title claims abstract description 20
- 238000012545 processing Methods 0.000 claims abstract description 200
- 238000012549 training Methods 0.000 claims abstract description 113
- 230000003044 adaptive effect Effects 0.000 claims description 43
- 230000010354 integration Effects 0.000 claims description 39
- 238000007619 statistical method Methods 0.000 claims description 17
- 230000006978 adaptation Effects 0.000 claims 3
- 230000009286 beneficial effect Effects 0.000 description 12
- 238000012795 verification Methods 0.000 description 8
- 238000007711 solidification Methods 0.000 description 4
- 230000008023 solidification Effects 0.000 description 4
- 238000010586 diagram Methods 0.000 description 2
- 238000013473 artificial intelligence Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
Images
Classifications
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D30/00—Reducing energy consumption in communication networks
- Y02D30/70—Reducing energy consumption in communication networks in wireless communication networks
Landscapes
- Digital Transmission Methods That Use Modulated Carrier Waves (AREA)
- Telephonic Communication Services (AREA)
Abstract
The invention provides a method and a device for adjusting a wireless modem based on a neural network, wherein the method comprises the following steps: performing large-scale neural network training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor; establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network; inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment; and (4) carrying out statistics on the adjusted parameter characteristics of the analysis model, continuously updating the embedded neural network processing self-adaptive adjustment model, and carrying out self-adaptive continuous updating intelligent adjustment on the wireless modem.
Description
Technical Field
The invention relates to the technical field of wireless communication modulation and demodulation and artificial intelligence, in particular to a method and a device for adjusting a wireless modem based on a neural network.
Background
At present, the efficiency of the wireless modulation and demodulation technology and the traditional chip is still low during the operation of the neural network; there are problems including: how to improve the neural network processor to train the neural network on a large scale, how to establish an adjustment model of the wireless modem, how to input and output model adjustment parameters, update and adjust the wireless modem model, and perform adaptive continuous update intelligent adjustment of the wireless modem, and the like, still remain to be further solved; therefore, there is a need for a method and an apparatus for adjusting a wireless modem based on a neural network, so as to at least partially solve the problems in the prior art.
Disclosure of Invention
In the summary section, a series of concepts in a simplified form are introduced, which will be described in further detail in the detailed description section; this summary of the invention is not intended to identify key features or essential features of the claimed subject matter, nor is it intended to be used as an aid in determining the scope of the claimed subject matter.
To at least partially solve the above problems, the present invention provides a neural network based tuning method of a wireless modem, comprising:
s100, performing large-scale neural network training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor;
s200, establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network;
s300, inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment;
s400, the parameter characteristics after the model adjustment are statistically analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment.
Preferably, S100 includes:
s101, carrying out system integration on an embedded neural network processing kernel and a standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
s102, performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
s103, performing large-scale neural network training through the embedded neural network processing standard modulation-demodulation integration subsystem and data driving parallel computation.
Preferably, S200 includes:
s201, taking the existing wireless modulation and demodulation system as an initial training model;
s202, using standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on an initial training model;
s203, setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem.
Preferably, S300 includes:
s301, acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise;
s302, inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model;
and S303, processing the parameter adjustment processing of the self-adaptive adjustment model through the embedded neural network, and outputting the parameter after the model adjustment.
Preferably, S400 includes:
s401, carrying out statistical processing on the feature classification data, and carrying out statistical analysis on the adjusted parameter features of the model;
s402, comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference;
and S403, continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
The invention relates to a wireless modem adjusting device based on a neural network, which comprises:
the parallel row-drive embedded neural network system adopts data-drive parallel computation through an embedded neural network processor to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem;
the wireless modulation and demodulation self-adaptive adjustment system establishes an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of an embedded neural network;
the parameter input model adjustment output system inputs the parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model and outputs the parameters after the model adjustment;
the characteristic classification statistical intelligent adjusting system is used for counting and analyzing the parameter characteristics after the model is adjusted, continuously updating the embedded neural network processing adaptive adjusting model and carrying out adaptive continuous updating intelligent adjustment on the wireless modem.
Preferably, the parallel row-drive embedded neural network system includes:
the embedded kernel modulation and demodulation integration unit is used for carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
the data driving parallel computing unit is used for carrying out data driving parallel computing according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
and the neural network large-scale training unit performs the neural network large-scale training through the embedded neural network processing standard modulation-demodulation integrated subsystem and the data driving parallel computation.
Preferably, the wireless modem adaptive adjustment system includes:
the existing wireless modulation and demodulation system importing unit takes the existing wireless modulation and demodulation system as an initial training model;
the embedded neural network large-scale training unit is used for performing embedded neural network large-scale training on the initial training model by taking the standard wireless modulation and demodulation parameters as model training target parameters;
and the parameter identification self-adaptive adjusting unit is used for setting Landau discrete time recursive parameter identification self-adaptive adjusting mechanism and establishing an embedded neural network processing self-adaptive adjusting model of the wireless modem.
Preferably, the parameter input model adjustment output system comprises:
the interference parameter acquisition unit to be adjusted acquires parameters to be adjusted of the wireless modem with wireless electromagnetic interference noise;
the parameter model input unit to be adjusted inputs the parameter to be adjusted into the embedded neural network processing self-adaptive adjustment model;
and the adjusting model parameter adjusting output unit is used for processing the parameter adjusting processing of the self-adaptive adjusting model through an embedded neural network and outputting the adjusted parameters of the model.
Preferably, the feature classification statistical intelligent adjustment system comprises:
the characteristic classification data statistical unit is used for carrying out statistical processing on the characteristic classification data and carrying out statistical analysis on the adjusted parameter characteristics of the model;
the parameter characteristic difference analysis unit is used for comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison differences;
and the modulation and demodulation self-adaptive intelligent adjusting unit continuously updates the embedded neural network processing self-adaptive adjusting model according to the parameter characteristic comparison difference and carries out self-adaptive continuous updating intelligent adjustment on the wireless modem.
Compared with the prior art, the invention at least comprises the following beneficial effects:
the beneficial effects of the above technical scheme are: the invention provides a method and a device for adjusting a wireless modem based on a neural network, which are used for carrying out neural network large-scale training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor; establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network; inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment; the parameter characteristics after the model adjustment are counted and analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment; carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; processing a standard modulation-demodulation integrated subsystem and data driving parallel computation through an embedded neural network, and performing large-scale neural network training; taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem; acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; processing parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network, and outputting parameters after model adjustment; performing statistical processing on the feature classification data, and performing statistical analysis on the parameter features after model adjustment; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; the processing of the wireless modem can be improved from the solidification setting to the intelligent self-adaptive adjustment, and the processing efficiency of the wireless modem is further greatly improved.
Additional advantages, objects, and features of the invention will be set forth in part in the description which follows and in part will become apparent to those having ordinary skill in the art upon examination of the following or may be learned from practice of the invention.
Drawings
The accompanying drawings, which are included to provide a further understanding of the invention and are incorporated in and constitute a part of this specification, illustrate embodiments of the invention and together with the description serve to explain the principles of the invention and not to limit the invention. In the drawings:
fig. 1 is a flowchart illustrating the steps of a method for adjusting a neural network-based wireless modem according to the present invention.
Fig. 2 is a diagram illustrating an embodiment of a method for adjusting a neural network-based wireless modem according to the present invention.
Fig. 3 is a system block diagram of an adjusting device of a neural network-based wireless modem according to the present invention.
Detailed Description
The present invention is further described in detail below with reference to the drawings and examples so that those skilled in the art can implement the invention with reference to the description; as shown in fig. 1 to 3, the present invention provides a method for adjusting a neural network-based wireless modem, including:
s100, performing large-scale neural network training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor;
s200, establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network;
s300, inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment;
s400, the parameter characteristics after the model adjustment are statistically analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment.
The working principle of the technical scheme is as follows: the invention provides a method for adjusting a wireless modem based on a neural network, which is characterized in that an embedded neural network processor adopts data drive parallel computation to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem; establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network; inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment; the parameter characteristics after the model adjustment are counted and analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment; carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; processing a standard modulation-demodulation integrated subsystem and data driving parallel computation through an embedded neural network, and performing large-scale neural network training; taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem; acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; processing parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network, and outputting parameters after model adjustment; performing statistical processing on the characteristic classification data, and performing statistical analysis on the adjusted parameter characteristics of the model; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; and continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
The beneficial effects of the above technical scheme are: the invention provides a method for adjusting a wireless modem based on a neural network, which comprises the following steps: performing large-scale neural network training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor; establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network; inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment; the parameter characteristics after the model adjustment are counted and analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment; carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; processing a standard modulation-demodulation integrated subsystem and data driving parallel computation through an embedded neural network, and performing large-scale neural network training; taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem; acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; processing parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network, and outputting parameters after model adjustment; performing statistical processing on the characteristic classification data, and performing statistical analysis on the adjusted parameter characteristics of the model; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; the processing of the wireless modem can be improved from the curing setting to the intelligent adaptive adjustment, and the processing efficiency of the wireless modem is further greatly improved.
In one embodiment, S100 includes:
s101, carrying out system integration on an embedded neural network processing kernel and a standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
s102, performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
s103, performing large-scale neural network training through the embedded neural network processing standard modulation-demodulation integration subsystem and data driving parallel computation.
The working principle of the technical scheme is as follows: s100 includes: s101, carrying out system integration on an embedded neural network processing kernel and a standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; s102, performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; s103, performing large-scale neural network training through the embedded neural network processing standard modulation-demodulation integration subsystem and data driving parallel computation.
The beneficial effects of the above technical scheme are: carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; and performing large-scale neural network training by an embedded neural network processing standard modulation-demodulation integrated subsystem and data driving parallel computation.
In one embodiment, S200 includes:
s201, taking the existing wireless modulation and demodulation system as an initial training model;
s202, using standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on an initial training model;
s203, setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem.
The working principle of the technical scheme is as follows: s200 comprises the following steps: s201, taking the existing wireless modulation and demodulation system as an initial training model; s202, using standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on an initial training model; s203, setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem.
The beneficial effects of the above technical scheme are: taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification self-adaptive adjustment mechanism, and establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem.
In one embodiment, S300 includes:
s301, acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise;
s302, inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model;
and S303, processing the parameter adjustment processing of the self-adaptive adjustment model through the embedded neural network, and outputting the parameter after the model adjustment.
The working principle of the technical scheme is as follows: s300 comprises the following steps: s301, acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; s302, inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; and S303, processing the parameter adjustment processing of the self-adaptive adjustment model through the embedded neural network, and outputting the adjusted parameters of the model.
The beneficial effects of the above technical scheme are: acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; and processing the parameter adjustment processing of the self-adaptive adjustment model through the embedded neural network, and outputting the adjusted parameters of the model.
In one embodiment, S400 includes:
s401, carrying out statistical processing on the feature classification data, and carrying out statistical analysis on the adjusted parameter features of the model;
s402, comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference;
and S403, continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
The working principle of the technical scheme is as follows: s400 includes: s401, performing statistical processing on the feature classification data, and performing statistical analysis on the adjusted parameter features of the model; s402, comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; s403, continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; performing wireless modem adaptive continuous update intelligent tuning comprises: a neural network embedded serial-parallel weight priority control module is arranged in a modulation-demodulation self-adaptive intelligent adjusting unit, a route is configured in the neural network embedded serial-parallel weight priority control module, and a plurality of neural network embedded serial-parallel weight adjusting sequences are created on the basis of parameter characteristic comparison difference information; carrying out self-adaptive continuous updating intelligent adjustment according to the adjustment condition; the self-adaptive continuous updating intelligent adjustment according to the adjustment condition comprises the following steps: when the routing state is not congested, a neural network embedded serial-parallel weight priority control module is used for respectively verifying the training results of the neural network for the highest adjusting sequence, the second highest adjusting sequence, the lowest adjusting sequence and the second lowest adjusting sequence in the neural network embedded serial-parallel weight adjusting sequence, if the verification is consistent, the verified serial-parallel weight adjusting sequence is executed, and if the verification is inconsistent, the second highest adjusting sequence is continuously verified and the serial-parallel weight adjusting sequence is sequentially updated; when the routing state is congested, performing parallel processing on the wireless modems; and performing adaptive continuous updating intelligent adjustment on the wireless modem.
The beneficial effects of the above technical scheme are: performing statistical processing on the characteristic classification data, and performing statistical analysis on the adjusted parameter characteristics of the model; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; performing wireless modem adaptive continuous update intelligent tuning comprises: a neural network embedded serial-parallel weight priority control module is arranged in a modulation-demodulation self-adaptive intelligent adjusting unit, a route is configured in the neural network embedded serial-parallel weight priority control module, and a plurality of neural network embedded serial-parallel weight adjusting sequences are created on the basis of parameter characteristic comparison difference information; carrying out self-adaptive continuous updating intelligent adjustment according to the adjustment condition; the self-adaptive continuous updating intelligent adjustment according to the adjustment condition comprises the following steps: when the routing state is not congested, a neural network embedded serial-parallel weight priority control module is used for respectively verifying the training results of the neural network for the highest adjusting sequence, the second highest adjusting sequence, the lowest adjusting sequence and the second lowest adjusting sequence in the neural network embedded serial-parallel weight adjusting sequence, if the verification is consistent, the verified serial-parallel weight adjusting sequence is executed, and if the verification is inconsistent, the second highest adjusting sequence is continuously verified and the serial-parallel weight adjusting sequence is sequentially updated; when the routing state is congested, performing parallel processing on the wireless modems; carrying out self-adaptive continuous updating intelligent adjustment on the wireless modem; the processing of the wireless modem can be improved from the solidification setting to the intelligent self-adaptive adjustment, and the processing efficiency of the wireless modem is further greatly improved.
The invention provides a device for adjusting a wireless modem based on a neural network, which comprises:
the parallel-row drive embedded neural network system adopts data drive parallel computation through an embedded neural network processor to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem;
the wireless modem adaptive adjustment system establishes an embedded neural network processing adaptive adjustment model of a wireless modem through large-scale training of an embedded neural network;
the parameter input model adjustment output system inputs the parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model and outputs the parameters after the model adjustment;
the characteristic classification statistical intelligent adjusting system is used for counting and analyzing the parameter characteristics after the model is adjusted, continuously updating the embedded neural network processing adaptive adjusting model and carrying out adaptive continuous updating intelligent adjustment on the wireless modem.
The working principle of the technical scheme is as follows: the invention provides a device for adjusting a wireless modem based on a neural network, which comprises: the parallel-row drive embedded neural network system adopts data drive parallel computation through an embedded neural network processor to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem; the wireless modem adaptive adjustment system establishes an embedded neural network processing adaptive adjustment model of a wireless modem through large-scale training of an embedded neural network; the parameter input model adjustment output system inputs the parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model and outputs the parameters after the model adjustment; the characteristic classification statistical intelligent adjustment system is used for performing statistical analysis on the parameter characteristics after the model adjustment, continuously updating the embedded neural network processing self-adaptive adjustment model and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; processing a standard modulation-demodulation integrated subsystem and data driving parallel computation through an embedded neural network, and performing large-scale neural network training; taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem; acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; processing parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network, and outputting parameters after model adjustment; performing statistical processing on the characteristic classification data, and performing statistical analysis on the adjusted parameter characteristics of the model; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; and continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
The beneficial effects of the above technical scheme are: the invention provides a wireless modem adjusting device based on a neural network, and a parallel row drive embedded neural network system, wherein a data drive parallel computation is adopted through an embedded neural network processor to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem; the wireless modem adaptive adjustment system establishes an embedded neural network processing adaptive adjustment model of a wireless modem through large-scale training of an embedded neural network; the parameter input model adjustment output system inputs the parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model and outputs the parameters after the model adjustment; the characteristic classification statistical intelligent adjustment system is used for counting the parameter characteristics after the analysis model is adjusted, continuously updating the embedded neural network processing adaptive adjustment model and performing adaptive continuous updating intelligent adjustment on the wireless modem; carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; performing data-driven parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem; processing a standard modulation-demodulation integrated subsystem and data driving parallel computation through an embedded neural network, and performing large-scale neural network training; taking the existing wireless modulation and demodulation system as an initial training model; taking the standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on the initial training model; setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem; acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise; inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model; processing parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network, and outputting parameters after model adjustment; performing statistical processing on the feature classification data, and performing statistical analysis on the parameter features after model adjustment; comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference; continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem; the processing of the wireless modem can be improved from the solidification setting to the intelligent self-adaptive adjustment, and the processing efficiency of the wireless modem is further greatly improved.
In one embodiment, a parallel digital-drive embedded neural network system includes:
the embedded kernel modulation and demodulation integration unit is used for carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
the data driving parallel computing unit is used for carrying out data driving parallel computing according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
and the neural network large-scale training unit performs the neural network large-scale training through the embedded neural network processing standard modulation-demodulation integrated subsystem and the data driving parallel computation.
The working principle of the technical scheme is as follows: the parallel row-drive embedded neural network system comprises: the embedded kernel modulation and demodulation integration unit is used for carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; the data driving parallel computing unit is used for carrying out data driving parallel computing according to the embedded neural network processing standard modulation-demodulation integrated subsystem; and the neural network large-scale training unit performs the neural network large-scale training through the embedded neural network processing standard modulation-demodulation integrated subsystem and the data driving parallel computation.
The beneficial effects of the above technical scheme are: the embedded kernel modulation and demodulation integration unit is used for carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem; the data driving parallel computing unit is used for carrying out data driving parallel computing according to the embedded neural network processing standard modulation-demodulation integrated subsystem; and the neural network large-scale training unit performs the neural network large-scale training through the embedded neural network processing standard modulation-demodulation integrated subsystem and the data driving parallel computation.
In one embodiment, a wireless modem adaptive adjustment system comprises:
the existing wireless modulation and demodulation system importing unit takes the existing wireless modulation and demodulation system as an initial training model;
the embedded neural network large-scale training unit is used for performing embedded neural network large-scale training on the initial training model by taking the standard wireless modulation and demodulation parameters as model training target parameters;
and the parameter identification self-adaptive adjusting unit is used for setting a Landau discrete time recursive parameter identification self-adaptive adjusting mechanism and establishing an embedded neural network processing self-adaptive adjusting model of the wireless modem.
The working principle of the technical scheme is as follows: the wireless modulation-demodulation self-adaptive adjusting system comprises: the existing wireless modulation and demodulation system importing unit takes the existing wireless modulation and demodulation system as an initial training model; the embedded neural network large-scale training unit is used for performing embedded neural network large-scale training on the initial training model by taking the standard wireless modulation and demodulation parameters as model training target parameters; and the parameter identification self-adaptive adjusting unit is used for setting a Landau discrete time recursive parameter identification self-adaptive adjusting mechanism and establishing an embedded neural network processing self-adaptive adjusting model of the wireless modem.
The beneficial effects of the above technical scheme are: the existing wireless modulation and demodulation system importing unit takes the existing wireless modulation and demodulation system as an initial training model; the embedded neural network large-scale training unit is used for performing embedded neural network large-scale training on the initial training model by taking the standard wireless modulation and demodulation parameters as model training target parameters; and the parameter identification self-adaptive adjusting unit is used for setting Landau discrete time recursive parameter identification self-adaptive adjusting mechanism and establishing an embedded neural network processing self-adaptive adjusting model of the wireless modem.
In one embodiment, the parametric input model adjustment output system comprises:
the interference parameter acquisition unit to be adjusted acquires parameters to be adjusted of the wireless modem with wireless electromagnetic interference noise;
the parameter model input unit to be adjusted inputs the parameter to be adjusted into the embedded neural network processing self-adaptive adjustment model;
and the adjusting model parameter adjusting output unit is used for processing the parameter adjusting processing of the self-adaptive adjusting model through the embedded neural network and outputting the adjusted parameters of the model.
The working principle of the technical scheme is as follows: the parameter input model adjustment output system comprises: the interference parameter acquisition unit to be adjusted acquires parameters to be adjusted of the wireless modem with wireless electromagnetic interference noise; the parameter model input unit to be adjusted inputs the parameter to be adjusted into the embedded neural network processing self-adaptive adjustment model; the adjustment model parameter adjustment output unit is used for processing the parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network and outputting the adjusted parameters of the model; calculating the error square sum value of the parameters after model adjustment:
wherein, PJQmin is the error square sum value of the parameter after model adjustment, R is the total number of neural network layers, S is the total number of self-adaptive circulation times, i is the neural network operation layer where the parameter to be adjusted is located, G (xi) is the adjustment weight of the ith layer of neural network of the parameter to be adjusted, wj (xi) is the jth self-adaptive circulation target adjustment value of the ith layer of neural network, hi is the initial weight coefficient of the ith layer of neural network, and F (yi) is the output adjusted parameter value of the ith layer of neural network; and calculating the error square sum value of the parameters after the model adjustment to minimize the error square sum value of the parameters after the model adjustment.
The beneficial effects of the above technical scheme are: the interference parameter acquisition unit to be adjusted acquires parameters to be adjusted of the wireless modem with wireless electromagnetic interference noise; the parameter model input unit to be adjusted inputs the parameter to be adjusted into the embedded neural network processing self-adaptive adjustment model; the adjustment model parameter adjustment output unit is used for processing the parameter adjustment processing of the self-adaptive adjustment model through an embedded neural network and outputting the adjusted parameters of the model; calculating an error square sum value of the parameters after model adjustment, wherein PJQmin is the error square sum value of the parameters after model adjustment, R is the total number of layers of the neural network, S is the total number of self-adaptive cycles, i is a neural network operation layer where the parameters to be adjusted are located, G (xi) is the adjustment weight value of the ith layer of the neural network of the parameters to be adjusted, wj (xi) is the jth self-adaptive cycle target adjustment value of the ith layer of the neural network, hi is the initial weight coefficient of the ith layer of the neural network, and F (yi) is the output adjusted parameter value of the ith layer of the neural network; calculating the error square sum value of the parameters after the model adjustment to ensure that the error square sum value of the parameters after the model adjustment is minimum; the parameter adjustment of the embedded neural network process adaptive adjustment model substantially reduces the process adjustment error of the wireless modem.
In one embodiment, the feature classification statistics intelligent adjustment system comprises:
the characteristic classification data statistical unit is used for carrying out statistical processing on the characteristic classification data and carrying out statistical analysis on the adjusted parameter characteristics of the model;
the parameter characteristic difference analysis unit is used for comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison differences;
and the modulation and demodulation self-adaptive intelligent adjusting unit continuously updates the embedded neural network processing self-adaptive adjusting model according to the parameter characteristic comparison difference and performs self-adaptive continuous updating intelligent adjustment on the wireless modem.
The working principle of the technical scheme is as follows: the intelligent adjusting system for the feature classification statistics comprises: the characteristic classification data statistical unit is used for carrying out statistical processing on the characteristic classification data and carrying out statistical analysis on the adjusted parameter characteristics of the model; the parameter characteristic difference analysis unit is used for comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison differences; the modulation-demodulation self-adaptive intelligent adjusting unit continuously updates the embedded neural network processing self-adaptive adjusting model according to the parameter characteristic comparison difference and carries out self-adaptive continuous updating intelligent adjustment on the wireless modem; performing wireless modem adaptive continuous update intelligent tuning comprises: a neural network embedded serial-parallel weight priority control module is arranged in a modulation-demodulation self-adaptive intelligent adjusting unit, a route is configured in the neural network embedded serial-parallel weight priority control module, and a plurality of neural network embedded serial-parallel weight adjusting sequences are created on the basis of parameter characteristic comparison difference information; carrying out self-adaptive continuous updating intelligent adjustment according to the adjustment condition; the self-adaptive continuous updating intelligent adjustment according to the adjustment condition comprises the following steps: when the routing state is not congested, a neural network embedded serial-parallel weight priority control module is used for respectively verifying the training results of the neural network for the highest adjusting sequence, the second highest adjusting sequence, the lowest adjusting sequence and the second lowest adjusting sequence in the neural network embedded serial-parallel weight adjusting sequence, if the verification is consistent, the verified serial-parallel weight adjusting sequence is executed, and if the verification is inconsistent, the second highest adjusting sequence is continuously verified and the serial-parallel weight adjusting sequence is sequentially updated; when the routing state is congested, performing parallel processing on the wireless modems; and performing adaptive continuous updating intelligent adjustment on the wireless modem.
The beneficial effects of the above technical scheme are: the characteristic classification data statistical unit is used for carrying out statistical processing on the characteristic classification data and carrying out statistical analysis on the adjusted parameter characteristics of the model; the parameter characteristic difference analysis unit is used for comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison differences; the modulation-demodulation self-adaptive intelligent adjusting unit continuously updates the embedded neural network processing self-adaptive adjusting model according to the parameter characteristic comparison difference and carries out self-adaptive continuous updating intelligent adjustment on the wireless modem; the adaptive continuous update intelligent adjustment of the wireless modem comprises the following steps: a neural network embedded serial-parallel weight priority control module is arranged in a modulation-demodulation self-adaptive intelligent adjusting unit, a route is configured in the neural network embedded serial-parallel weight priority control module, and a plurality of neural network embedded serial-parallel weight adjusting sequences are created on the basis of parameter characteristic comparison difference information; carrying out self-adaptive continuous updating intelligent adjustment according to the adjustment condition; the self-adaptive continuous updating intelligent adjustment according to the adjustment condition comprises the following steps: when the routing state is not congested, a neural network embedded serial-parallel weight priority control module is used for respectively verifying the training results of the neural network for the highest adjusting sequence, the second highest adjusting sequence, the lowest adjusting sequence and the second lowest adjusting sequence in the neural network embedded serial-parallel weight adjusting sequence, if the verification is consistent, the verified serial-parallel weight adjusting sequence is executed, and if the verification is inconsistent, the second highest adjusting sequence is continuously verified and the serial-parallel weight adjusting sequence is sequentially updated; when the routing state is congested, performing parallel processing on the wireless modems; carrying out self-adaptive continuous updating intelligent adjustment on the wireless modem; the processing of the wireless modem can be improved from the solidification setting to the intelligent self-adaptive adjustment, and the processing efficiency of the wireless modem is further greatly improved.
While embodiments of the invention have been disclosed above, it is not limited to the applications set forth in the description and the embodiments, which are fully applicable in various fields of endeavor to which the invention pertains, and further modifications may readily be made by those skilled in the art, it being understood that the invention is not limited to the details shown and described herein without departing from the general concept defined by the appended claims and their equivalents.
Claims (10)
1. The method for adjusting the wireless modem based on the neural network is characterized by comprising the following steps:
s100, performing large-scale neural network training on the existing large-scale parameter set of the wireless modem by adopting data-driven parallel computation through an embedded neural network processor;
s200, establishing an embedded neural network processing self-adaptive adjustment model of the wireless modem through large-scale training of the embedded neural network;
s300, inputting parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model, and outputting parameters after model adjustment;
s400, the parameter characteristics after the model adjustment are statistically analyzed, the embedded neural network processing self-adaptive adjustment model is continuously updated, and the wireless modem is subjected to self-adaptive continuous updating intelligent adjustment.
2. The neural network-based wireless modem tuning method of claim 1, wherein S100 comprises:
s101, carrying out system integration on an embedded neural network processing kernel and a standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
s102, performing data driving parallel computation according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
s103, performing large-scale neural network training through the embedded neural network processing standard modulation-demodulation integration subsystem and data driving parallel computation.
3. The neural network-based wireless modem tuning method of claim 1, wherein S200 comprises:
s201, taking the existing wireless modulation and demodulation system as an initial training model;
s202, using standard wireless modulation and demodulation parameters as model training target parameters, and carrying out embedded neural network large-scale training on an initial training model;
s203, setting Landau discrete time recursive parameter identification adaptive adjustment mechanism, and establishing an embedded neural network processing adaptive adjustment model of the wireless modem.
4. The tuning method of a neural network-based wireless modem according to claim 1, wherein S300 comprises:
s301, acquiring parameters to be adjusted of a wireless modem with wireless electromagnetic interference noise;
s302, inputting parameters to be adjusted into an embedded neural network processing self-adaptive adjustment model;
and S303, processing the parameter adjustment processing of the self-adaptive adjustment model through the embedded neural network, and outputting the adjusted parameters of the model.
5. The neural network-based wireless modem tuning method of claim 1, wherein S400 comprises:
s401, carrying out statistical processing on the feature classification data, and carrying out statistical analysis on the adjusted parameter features of the model;
s402, comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison difference;
and S403, continuously updating the embedded neural network processing self-adaptive adjustment model according to the parameter characteristic comparison difference, and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
6. An apparatus for adjusting a neural network based wireless modem, comprising:
the parallel-row drive embedded neural network system adopts data drive parallel computation through an embedded neural network processor to carry out neural network large-scale training on the existing large-scale parameter set of the wireless modem;
the wireless modem adaptive adjustment system establishes an embedded neural network processing adaptive adjustment model of a wireless modem through large-scale training of an embedded neural network;
the parameter input model adjustment output system inputs the parameters to be adjusted of the wireless modem into the embedded neural network processing self-adaptive adjustment model and outputs the parameters after the model adjustment;
the characteristic classification statistical intelligent adjusting system is used for performing statistical analysis on the adjusted parameter characteristics of the model, continuously updating the embedded neural network processing self-adaptive adjusting model and performing self-adaptive continuous updating intelligent adjustment on the wireless modem.
7. The neural network-based wireless modem adaptation device of claim 6, wherein the parallel row drive embedded neural network system comprises:
the embedded kernel modulation and demodulation integration unit is used for carrying out system integration on the embedded neural network processing kernel and the standard wireless modulation and demodulation unit to obtain an embedded neural network processing standard modulation and demodulation integration subsystem;
the data driving parallel computing unit is used for carrying out data driving parallel computing according to the embedded neural network processing standard modulation-demodulation integrated subsystem;
and the neural network large-scale training unit performs the neural network large-scale training through the embedded neural network processing standard modulation-demodulation integrated subsystem and the data driving parallel computation.
8. The neural network-based wireless modem adaptation device of claim 6, wherein the wireless modem adaptation adjustment system comprises:
the existing wireless modulation and demodulation system importing unit takes the existing wireless modulation and demodulation system as an initial training model;
the embedded neural network large-scale training unit is used for performing embedded neural network large-scale training on the initial training model by taking the standard wireless modulation-demodulation parameters as model training target parameters;
and the parameter identification self-adaptive adjusting unit is used for setting Landau discrete time recursive parameter identification self-adaptive adjusting mechanism and establishing an embedded neural network processing self-adaptive adjusting model of the wireless modem.
9. The neural network-based wireless modem adjustment apparatus according to claim 6, wherein the parameter input model adjustment output system comprises:
the interference parameter acquisition unit to be adjusted acquires parameters to be adjusted of the wireless modem with wireless electromagnetic interference noise;
the parameter model input unit to be adjusted inputs the parameter to be adjusted into the embedded neural network processing self-adaptive adjustment model;
and the adjusting model parameter adjusting output unit is used for processing the parameter adjusting processing of the self-adaptive adjusting model through the embedded neural network and outputting the adjusted parameters of the model.
10. The neural network-based wireless modem adjustment apparatus according to claim 6, wherein the feature classification statistical intelligence adjustment system comprises:
the characteristic classification data statistical unit is used for carrying out statistical processing on the characteristic classification data and carrying out statistical analysis on the adjusted parameter characteristics of the model;
the parameter characteristic difference analysis unit is used for comparing the parameter characteristics after model adjustment with the standard wireless modulation and demodulation parameter characteristics to obtain parameter characteristic comparison differences;
and the modulation and demodulation self-adaptive intelligent adjusting unit continuously updates the embedded neural network processing self-adaptive adjusting model according to the parameter characteristic comparison difference and performs self-adaptive continuous updating intelligent adjustment on the wireless modem.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211613531.9A CN115829016B (en) | 2022-12-15 | 2022-12-15 | Adjustment method and device of wireless modem based on neural network |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202211613531.9A CN115829016B (en) | 2022-12-15 | 2022-12-15 | Adjustment method and device of wireless modem based on neural network |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115829016A true CN115829016A (en) | 2023-03-21 |
| CN115829016B CN115829016B (en) | 2024-01-16 |
Family
ID=85545790
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202211613531.9A Active CN115829016B (en) | 2022-12-15 | 2022-12-15 | Adjustment method and device of wireless modem based on neural network |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115829016B (en) |
Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1961530A (en) * | 2003-12-07 | 2007-05-09 | 适应性频谱和信号校正股份有限公司 | Adaptive margin and band control in DSL system |
| CN101808173A (en) * | 2010-03-10 | 2010-08-18 | 杭州华三通信技术有限公司 | Method and apparatus for adaptively adjusting negotiable parameters of MODEM (Modulator-Demodulator) |
| CN110417516A (en) * | 2019-08-16 | 2019-11-05 | 北京小米移动软件有限公司 | The method of adjustment and device of radio modem neural network based |
| CN112417219A (en) * | 2020-11-16 | 2021-02-26 | 吉林大学 | Hyper-graph convolution-based hyper-edge link prediction method |
| CN113285902A (en) * | 2021-05-19 | 2021-08-20 | 南京航空航天大学 | Design method of OFDM system detector |
| CN113472713A (en) * | 2021-08-04 | 2021-10-01 | 上海无线电设备研究所 | High-order modulation signal demodulation method based on neural network and receiver |
| US20220004839A1 (en) * | 2020-07-02 | 2022-01-06 | Ahp-Tech Inc. | Artificial-intelligence decision-making core system with neural network |
| CN114169508A (en) * | 2021-12-14 | 2022-03-11 | 广东赛昉科技有限公司 | Method and system for processing data in neural network system |
| CN115314343A (en) * | 2022-08-05 | 2022-11-08 | 国网智能电网研究院有限公司 | Source-load-storage resource aggregation control gateway device and load and output prediction method |
-
2022
- 2022-12-15 CN CN202211613531.9A patent/CN115829016B/en active Active
Patent Citations (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1961530A (en) * | 2003-12-07 | 2007-05-09 | 适应性频谱和信号校正股份有限公司 | Adaptive margin and band control in DSL system |
| CN101808173A (en) * | 2010-03-10 | 2010-08-18 | 杭州华三通信技术有限公司 | Method and apparatus for adaptively adjusting negotiable parameters of MODEM (Modulator-Demodulator) |
| CN110417516A (en) * | 2019-08-16 | 2019-11-05 | 北京小米移动软件有限公司 | The method of adjustment and device of radio modem neural network based |
| US20220004839A1 (en) * | 2020-07-02 | 2022-01-06 | Ahp-Tech Inc. | Artificial-intelligence decision-making core system with neural network |
| CN112417219A (en) * | 2020-11-16 | 2021-02-26 | 吉林大学 | Hyper-graph convolution-based hyper-edge link prediction method |
| CN113285902A (en) * | 2021-05-19 | 2021-08-20 | 南京航空航天大学 | Design method of OFDM system detector |
| CN113472713A (en) * | 2021-08-04 | 2021-10-01 | 上海无线电设备研究所 | High-order modulation signal demodulation method based on neural network and receiver |
| CN114169508A (en) * | 2021-12-14 | 2022-03-11 | 广东赛昉科技有限公司 | Method and system for processing data in neural network system |
| CN115314343A (en) * | 2022-08-05 | 2022-11-08 | 国网智能电网研究院有限公司 | Source-load-storage resource aggregation control gateway device and load and output prediction method |
Non-Patent Citations (3)
| Title |
|---|
| MIN ZHANG ET AL.: "Enhanced Efficiency BPSK Demodulator Based on One-Dimensional Convolutional Neural Network", 《IEEE ACCESS》, vol. 6, pages 26939, XP011684856, DOI: 10.1109/ACCESS.2018.2834144 * |
| 钱付兰等: "基于路径相互关注的网络嵌入算法", 《南京大学学报(自然科学版)》, vol. 55, no. 4, pages 573 - 580 * |
| 韩桂华等: "电液伺服系统神经网络辨识及控制器设计", 《哈尔滨理工大学学报》, no. 05, pages 18 - 23 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115829016B (en) | 2024-01-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Saputra et al. | Analysis resilient algorithm on artificial neural network backpropagation | |
| CN111310915A (en) | Data anomaly detection and defense method for reinforcement learning | |
| CN109085469A (en) | A kind of method and system of the signal type of the signal of cable local discharge for identification | |
| CN111985601A (en) | Data identification method for incremental learning | |
| CN102469103B (en) | Trojan event prediction method based on BP (Back Propagation) neural network | |
| US7610250B2 (en) | Real-time method of determining eye closure state using off-line adaboost-over-genetic programming | |
| EP3502978A1 (en) | Meta-learning system | |
| CN110033081A (en) | A kind of method and apparatus of determining learning rate | |
| CN116384244A (en) | Electromagnetic field prediction method based on physical enhancement neural network | |
| CN114418117B (en) | Meta-learning method with adaptive learning rate for few-sample fault diagnosis | |
| CN116852665A (en) | Injection molding process parameter intelligent adjusting method based on mixed model | |
| CN111967308A (en) | Online road surface unevenness identification method and system | |
| CN112766292A (en) | Identity authentication method, device, equipment and storage medium | |
| CN111639680B (en) | Identity recognition method based on expert feedback mechanism | |
| CN115829016A (en) | Method and device for adjusting wireless modem based on neural network | |
| CN111898799B (en) | BFA-Elman-based power load prediction method | |
| CN116722541A (en) | Power system load prediction method and device based on convolutional neural network | |
| CN115412332B (en) | Internet of things intrusion detection system and method based on hybrid neural network model optimization | |
| US20240020531A1 (en) | System and Method for Transforming a Trained Artificial Intelligence Model Into a Trustworthy Artificial Intelligence Model | |
| CN116151409A (en) | Urban daily water demand prediction method based on neural network | |
| CN116048785A (en) | Elastic resource allocation method based on supervised learning and reinforcement learning | |
| CN113516163B (en) | Vehicle classification model compression method, device and storage medium based on network pruning | |
| CN115438786A (en) | Robust neural network training method based on sample-driven target loss function optimization | |
| CN114969148A (en) | System access amount prediction method, medium and equipment based on deep learning | |
| CN115294381A (en) | Small sample image classification method and device based on feature migration and orthogonal prior |
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 |