[go: up one dir, main page]
More Web Proxy on the site http://driver.im/

CN113098565A - Stage carrier communication self-adaptive frequency hopping anti-interference technology based on deep network - Google Patents

Stage carrier communication self-adaptive frequency hopping anti-interference technology based on deep network Download PDF

Info

Publication number
CN113098565A
CN113098565A CN202110360575.4A CN202110360575A CN113098565A CN 113098565 A CN113098565 A CN 113098565A CN 202110360575 A CN202110360575 A CN 202110360575A CN 113098565 A CN113098565 A CN 113098565A
Authority
CN
China
Prior art keywords
stage
noise
carrier communication
communication
model
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
Application number
CN202110360575.4A
Other languages
Chinese (zh)
Other versions
CN113098565B (en
Inventor
荆俊明
石建明
王斌
韩旭润
祁虎强
刘晨
何鹏志
宋一凡
高原
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Big Stage Of Gansu Engineering Technology Engineering Co ltd
Original Assignee
Big Stage Of Gansu Engineering Technology Engineering Co ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Big Stage Of Gansu Engineering Technology Engineering Co ltd filed Critical Big Stage Of Gansu Engineering Technology Engineering Co ltd
Priority to CN202110360575.4A priority Critical patent/CN113098565B/en
Publication of CN113098565A publication Critical patent/CN113098565A/en
Application granted granted Critical
Publication of CN113098565B publication Critical patent/CN113098565B/en
Active legal-status Critical Current
Anticipated expiration legal-status Critical

Links

Images

Classifications

    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B3/00—Line transmission systems
    • H04B3/54—Systems for transmission via power distribution lines
    • H04B3/544—Setting up communications; Call and signalling arrangements
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/004—Artificial life, i.e. computing arrangements simulating life
    • G06N3/006—Artificial life, i.e. computing arrangements simulating life based on simulated virtual individual or collective life forms, e.g. social simulations or particle swarm optimisation [PSO]
    • 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
    • H—ELECTRICITY
    • H04—ELECTRIC COMMUNICATION TECHNIQUE
    • H04B—TRANSMISSION
    • H04B1/00—Details of transmission systems, not covered by a single one of groups H04B3/00 - H04B13/00; Details of transmission systems not characterised by the medium used for transmission
    • H04B1/69—Spread spectrum techniques
    • H04B1/713—Spread spectrum techniques using frequency hopping
    • H04B1/715—Interference-related aspects
    • 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

  • Engineering & Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Theoretical Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • General Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Artificial Intelligence (AREA)
  • Biomedical Technology (AREA)
  • Biophysics (AREA)
  • Computational Linguistics (AREA)
  • Signal Processing (AREA)
  • Evolutionary Computation (AREA)
  • Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • General Physics & Mathematics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • Computer Networks & Wireless Communication (AREA)
  • Power Engineering (AREA)
  • Cable Transmission Systems, Equalization Of Radio And Reduction Of Echo (AREA)

Abstract

The invention discloses a stage carrier communication self-adaptive frequency hopping anti-interference technology based on a deep network, which belongs to the technical field of carrier communication anti-interference and comprises the following steps: the method comprises the steps of firstly obtaining a large number of stage power line transmission channel parameters by using a model, then training a self-adaptive frequency hopping anti-interference model by using a DBN-BP deep network, obtaining a final self-adaptive frequency hopping anti-interference model by learning to obtain a relation matrix between a stage power line environment and an optimal frequency point, wherein the model can self-adaptively select a communication frequency band according to the stage power line communication environment, and the stability and the transmission efficiency of stage carrier communication are greatly improved. Experiments prove that the technology can quickly and effectively select a stable communication frequency band according to the interference condition of the stage communication power line environment, and further improves the reliability and stability of communication between stage equipment while improving the communication efficiency.

Description

Stage carrier communication self-adaptive frequency hopping anti-interference technology based on deep network
Technical Field
The invention belongs to the technical field of carrier communication anti-interference, and particularly relates to a stage carrier communication self-adaptive frequency hopping anti-interference technology based on a deep network.
Background
Along with the development of social economy, the living standard of people is increasingly improved, the demand on cultural art is also increasingly high, and a theater stage is taken as a main medium and a carrier of the present art display and has an important position in enriching the artistic life of people. In recent years, with the continuous innovation of scientific technology, more and more new methods and new technologies are widely applied to stage deductions. Through the diligent efforts of many researchers, the functions realized by the stage are more and more complex, and more used equipment is provided, so that the problems of poor equipment synchronization, difficult control and the like in the operation process are caused. In order to better mobilize the synchronous operation of each device and ensure the smooth completion of the performance, a reliable communication control system needs to be established between each device and each module. The existing widely applied communication system utilizes a sensor to connect a special communication line to be connected with a collection device and a controller protocol bus, but causes the problems of disordered wiring, difficult maintenance, easy disturbance and the like when the devices are too many and the monitoring parameters are complex.
The application of the current low-voltage carrier technology brings a new solution for improving the stage communication efficiency, and the widely applied carrier anti-interference technologies include direct sequence spread spectrum, Chirp, Orthogonal Frequency Division Multiplexing (OFDM), Frequency Hopping (FH), Time Hopping (TH), various combined spread spectrum technologies and the like, but the interference and noise in the line are difficult problems for restricting the popularization of the technology. Therefore, the construction of a stable, reliable and efficient stage equipment communication system is of great significance to the development of deductive technology
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a stage carrier communication adaptive frequency hopping anti-interference technology based on a deep network, which learns the characteristics of communication signals through a deep neural network and then carries out communication frequency point updating through a trained adaptive frequency hopping model, thereby ensuring the communication accuracy, improving the communication stability, saving the communication time and further improving the communication efficiency among stage equipment.
In order to achieve the purpose, the invention adopts the technical scheme that:
a stage carrier communication self-adaptive frequency hopping anti-interference technology based on a deep network comprises the following steps:
s1: establishing a stage carrier communication system model: on the basis of a stage carrier communication system, interference factors influencing communication quality in the stage carrier communication system are analyzed, a port network is established to simulate a stage power line carrier communication process by combining the actual operation condition of the stage carrier communication system, and a mathematical expression model of the stage power line carrier communication process is as follows:
Figure BDA0003005366290000021
wherein g is a channel trade-off factor with a value less than 1, d is a channel length, and alpha is0And alpha1K is an exponential decay factor with the value of 0.5 to 1 as a decay parameter, vpFor signal propagation speed, the expression is:
vp＝d/τ (2)
wherein tau is the channel transmission time delay,
Figure BDA0003005366290000022
c0＝3×108m/s, epsilon are constants of insulating dielectric layers of communication power lines, and the impact of a channel model is correspondingly as follows:
h(t)＝kδ(t-τ) (3)
in the formula, k is a channel balance factor, and tau is channel transmission path time delay;
s2: stage carrier communication interference analysis: according to a stage carrier communication system, analyzing interference factors influencing communication quality in stage communication;
s3: stage carrier communication environment simulation sampling: setting parameter ranges of each model aiming at the same environment after disturbance analysis of interference factors, and then simulating a similar communication environment by adopting a random algorithm to obtain a plurality of groups of experimental data;
s4: optimizing optimal frequency point parameters by using a particle swarm algorithm: setting the frequency point parameter particle swarm in the multidimensional space as X, and respectively setting the position and speed of the particle i as XiAnd viThe optimal History bit is marked as PibeGlobal optimum position is noted as GbeThen, the velocity update formula and the position update formula of the particle are shown as follows:
Figure BDA0003005366290000031
Figure BDA0003005366290000032
wherein
Figure BDA0003005366290000033
Is the velocity of the kth iteration of the ith particle,
Figure BDA0003005366290000034
for the position of the kth iteration of the ith particle, c1、c2Respectively, a learning factor, adjusting a learning step length, R is [0,1 ]]With the followingA machine value function;
in the optimization process, according to an error rate updating algorithm of stage carrier communication, finding out an optimal frequency point design parameter most suitable for the current communication environment after repeated iteration, taking the obtained parameter as a label, taking communication environment sampling data as a training sample to form a training data set, and obtaining a large number of training data sets for training a self-adaptive frequency hopping anti-interference model after optimizing enough environments;
s5: training the adaptive model by using a DBN-BP deep network: the method comprises the steps of designing a DBN-BP network structure through experiments, gradually learning and extracting disturbance characteristics of a simulated engineering environment by utilizing a RBM layer, bringing output characteristics of the RBM into a BP network, carrying out fine adjustment by utilizing a back propagation algorithm, and obtaining a relation matrix between a stage power line environment and an optimal frequency point through learning to obtain a final self-adaptive frequency hopping anti-interference model.
The stage carrier communication system in step S1 is mainly composed of a carrier controller, an encoding device, a coupling device, a decoupling device, and a decoding device, where the carrier controller is configured to process the collected information of each stage device and each detection device, and control the encoder to encode the transmission data, the encoded information is accessed to the stage power line through the coupler, and the transmitted encoded signal is separated by the decoupling device at the receiving location, and then the transmitted data information is obtained by the decoding device.
The interference factors in the step S2 and the step S3 mainly include colored background noise, narrow-band noise, asynchronous noise with power frequency, synchronous noise with power frequency, and impulse noise.
The colored background noise is noise generated after a plurality of low-intensity noise sources in a stage communication system are fused, the colored background noise can be regarded as a stable random process, a model of the colored background noise can be simulated by a group of white noise sequences with the average value of 0 through AR filtering, and the process is shown as the following formula:
Figure BDA0003005366290000041
where a (i) is the AR coefficient, w (n) is a set of white noises with an average value of 0.
The stage narrow-band is noise caused by frequency domain interference, and the narrow-band noise can be simulated by superposing a plurality of sine functions, and the simulation process is shown as the following formula:
Figure BDA0003005366290000042
wherein f isiIs the frequency of the ith waveform, AiIn order to be the amplitude value,
Figure BDA0003005366290000043
is the phase.
The noise synchronous with the power frequency is periodic noise which is synchronous with the power frequency, long in duration and large in frequency coverage range and is generated when stage equipment with a thyristor works under the power frequency, the periodic noise can be described through pulse width, pulse existing time and pulse interval, a sine signal attenuated according to an exponential is selected for simulation, and the simulation process is shown as the following formula.
Figure BDA0003005366290000044
Where N is the number of pulses, τiIs the ith pulse time constant, tarriThe ith pulse epoch.
Both the synchronous noise and asynchronous noise at power frequency can be simulated by a single periodic noise pulse.
The impulse noise mainly refers to random noise which is generated by transient signals generated in the processes of stage equipment switching and start-stop control and has short duration but concentrated energy, can be described by pulse width, pulse existing time and pulse interval, and can be simulated by a single pulse universal function.
In the step S4, in order to ensure that the algorithm does not fall into the locally optimal solution, the particle swarm velocity is updated by the following formula:
Figure BDA0003005366290000045
in the formula NpThe method is characterized in that the particle similarity gain is adopted, when the similarity between model particles is high and the iteration is carried out for more than 5 times in a smaller optimal frequency point range, the current optimal frequency point parameters are recorded and a larger N is setpAnd (4) continuously searching other optimal frequency points by the model, and selecting the frequency point parameters with optimal performance when the iteration is finished.
Compared with the prior art, the invention has the following beneficial effects:
according to the invention, firstly, a large number of stage power line transmission channel parameters are obtained by using the model, then the DBN-BP deep network is used for training the self-adaptive frequency hopping anti-interference model, the model can self-adaptively select a communication frequency band according to the stage power line communication environment, and the stability and the transmission efficiency of stage carrier communication are greatly improved. Experiments prove that the technology can quickly and effectively select a stable communication frequency band according to the interference condition of the stage communication power line environment, and further improves the reliability and stability of communication between stage equipment while improving the communication efficiency.
Drawings
Fig. 1 is a schematic diagram of the frequency hopping technique of the present invention.
Fig. 2 is the principle of the adaptive frequency hopping anti-interference technology of the present invention.
Fig. 3 is a stage carrier communication model of the present invention.
Fig. 4 is a colored background noise Simulink model of the present invention.
Fig. 5 is a narrow-band noise Simulink model of the present invention.
Fig. 6 is a periodic impulse noise Simulink model of the present invention.
Fig. 7 is a random burst noise Simulink model of the present invention.
Fig. 8 shows the interfering signal and its debounce signal according to the invention.
Fig. 9 is a signal transmission bit error plot of the present invention.
FIG. 10 is a BP network training error of the present invention.
Detailed Description
The invention will be described in detail with reference to the drawings and specific embodiments, which are illustrative of the invention and are not to be construed as limiting the invention.
For a better understanding of the substance of the invention, the following description is made:
the Frequency Hopping technique (FHSS) in the prior art is one of the most commonly used spreading schemes, and the operation principle thereof is shown in fig. 1, in which carrier frequencies of transmission signals of both the transmitting and receiving parties are discretely changed according to a predetermined rule, and carrier frequencies used in communication are randomly hopped under the control of pseudo-random change codes, and the Frequency Hopping technique is a communication scheme using code sequences to perform multi-Frequency shift keying. Compared with fixed frequency communication, frequency hopping communication has good anti-interference capability, and normal communication can be carried out on other non-interfered frequency points even if some frequency points are interfered. However, different frequency points are required to be selected continuously according to the communication quality to achieve the expected communication effect, which seriously affects the communication efficiency and stability.
The self-adaptive frequency hopping anti-interference technology adopts a method of training a self-adaptive model by a deep neural network, and self-adaptively updates an optimal communication frequency point sequence according to the detected interference condition of the communication environment, so as to further save the communication time and improve the communication efficiency and stability. Firstly, obtaining a large amount of data simulating the actual engineering environment of a stage through simulation, continuously carrying out frequency hopping communication simulation on the engineering environment under the same parameter, taking the average error rate as a target function, optimizing through a particle swarm optimization algorithm to obtain an optimal frequency point updating parameter, solving the optimal frequency point updating parameter of all simulated engineering environment data, taking the simulated environment data as a sample, arranging a training data set by taking the obtained optimal frequency point parameter as a label, then training an adaptive frequency point updating model by utilizing a deep neural network, wherein the model can self-adaptively set the optimal frequency point updating parameter according to the actual engineering sampling, exporting the trained model, and replacing the traditional method for continuously and randomly updating the frequency point parameter according to the error rate with a method for determining the frequency point updating parameter by the adaptive model to carry out communication. The implementation principle of the stage low-voltage carrier communication adaptive frequency hopping anti-interference technology is shown in fig. 2.
The invention discloses a stage carrier communication self-adaptive frequency hopping anti-interference technology based on a deep network, which mainly comprises the following contents:
1. stage carrier communication disturbance analysis and model establishment
1.1 introduction and model analysis of stage Carrier communication
The stage control system mainly comprises an upper control system, a programmable controller, an execution controller, an information acquisition system, controlled equipment and the like, stage carrier communication is an important link of the information acquisition system, and the stage control system has the main function that acquired equipment operation information, signals of each encoder and a position switch are transmitted to the execution controller and then are uniformly submitted to the programmable controller for processing through a data bus. The stage carrier communication system mainly comprises a carrier controller, an encoding device, a coupling device, a decoupling device and a decoding device, wherein the carrier controller is used for processing collected information of each stage device and each detection device and controlling the encoder to encode transmission data, the encoded information is accessed into a stage power line through the coupler, and transmitted encoding signals are separated through the decoupling device at a receiving position and then transmitted data information is obtained through the decoding device. Aiming at the problems of complex stage equipment, unmatched impedance, more interference factors, signal attenuation and the like existing in a power transmission line, a 2-port network shown in figure 3 is established to simulate the stage power line carrier communication process by combining the actual operation condition of a stage carrier communication system.
The mathematical expression model is shown as formula 10.
Figure BDA0003005366290000071
Wherein g is a channel balance factor with a value less than 1, d is a channel length, and alpha is0And alpha1K is an exponential decay factor with the value of 0.5 to 1 as a decay parameter, vpFor the signal propagation speed, the expression is shown in equation 11.
vp＝d/τ (11)
Tau is the time delay of the channel transmission,
Figure BDA0003005366290000072
c0＝3×108m/s and epsilon are constants of insulating dielectric layers of the communication power lines.
The channel model impulse response is shown in equation 12.
h(t)＝kδ(t-τ) (12)
Where k is the channel trade-off factor and τ is the channel transmission path delay.
1.2 stage carrier communication interference analysis and simulation
The main interference affecting the communication quality in the stage communication is colored background noise, narrow-band noise, asynchronous noise with power frequency, synchronous noise with power frequency, impulse noise and the like. The colored background noise of the stage is noise generated after a plurality of low-intensity noise sources in a stage communication system are fused, the power spectral density of the noise is low, the noise characteristic is very similar to Gaussian white noise, the noise generally exists in the whole stage power line channel frequency band, and the frequency range of interference is as high as 30 MHz. The colored background noise of the stage can be regarded as a stationary random process, and the model can be simulated by a group of white noise sequences with the average value of 0 through AR filtering, and the process is shown as a formula 13.
Figure BDA0003005366290000081
Where a (i) is the AR coefficient and w (n) is a set of white noises with mean 0, a model as shown in fig. 4 was constructed in Simulink to simulate the stage colored background noise.
The stage narrow-band noise is noise caused by frequency domain interference with a narrow frequency band, and is mostly sinusoidal modulation, the noise range is generally within a frequency band of 1-22MHz, and the content of the noise in a channel is different at different times. The narrow-band noise in the stage carrier communication can be simulated by superposing a plurality of sine functions, and the simulation process is shown as formula 14.
Figure BDA0003005366290000082
Wherein f isiIs the frequency of the ith waveform, AiIn order to be the amplitude value,
Figure BDA0003005366290000083
for phase, a model as shown in fig. 5 was built in Simulink to simulate stage narrowband noise.
The stage and power frequency synchronous noise is periodic noise which is synchronous with the power frequency, long in duration and wide in frequency coverage range and is generated when some stage equipment containing thyristors works under the power frequency. The asynchronous noise with the power frequency refers to a plurality of periodic strong pulse interferences with the repetition frequency concentrated in 50kHz to 500kHz caused by the action of stage switch equipment. The periodic impulse noise can be described by pulse width, pulse duration and pulse interval, and is simulated by selecting a sinusoidal signal attenuated exponentially, and the simulation process is shown in equation 15.
Figure BDA0003005366290000084
Where N is the number of pulses, τiIs the ith pulse time constant, tarriThe ith pulse epoch. Noise synchronous with power frequency and noise asynchronous with power frequency can be simulated through a single periodic noise pulse, and a model shown in figure 6 is built in Simulink to simulate periodic impulse noise of a stage.
The stage random burst noise mainly refers to random noise which is generated by transient signals generated in the switching and starting and stopping control processes of stage equipment, has short duration and high energy concentration intensity, can greatly influence the voltage and the current on a communication line, and can seriously interfere the communication signals. Impulse noise can be described by pulse width, pulse duration and pulse interval, and can be simulated by using a single pulse general function, and a model shown in fig. 7 is built in Simulink to simulate stage random burst noise.
2. Adaptive frequency hopping anti-interference model training
2.1 stage carrier communication environment simulation sampling
The stage carrier communication technology is an application practice of low-voltage power line carrier communication in the field of stage deduction, can better simplify circuit design and save cost while improving communication efficiency between stage equipment, but because the power lines between the stage equipment are complex, the equipment types are various, equipment is started and stopped randomly and other factors, various interferences exist in communication channels, and the communication quality between the stage equipment is seriously influenced. In order to better accord with the actual application environment of the stage equipment, the actual operation environment of the stage equipment is sampled, noise disturbance analysis is carried out, then parameter ranges of various models are set for the same environment, and then a random algorithm is adopted to simulate a similar communication environment to obtain a large amount of experimental data.
2.2 optimizing optimal frequency Point parameters Using particle swarm optimization
Particle Swarm Optimization (PSO) is a search algorithm based on group cooperation, and is an intelligent type of cluster. Setting the frequency point parameter particle swarm in the multidimensional space as X, and respectively setting the position and speed of the particle i as XiAnd viThe optimal History bit is marked as PibeGlobal optimum position is noted as GbeThen, the velocity update formula and the position update formula of the particle are shown in equations 16 and 17:
Figure BDA0003005366290000091
Figure BDA0003005366290000092
wherein
Figure BDA0003005366290000101
Is the velocity of the kth iteration of the ith particle,
Figure BDA0003005366290000102
for the position of the kth iteration of the ith particle, c1、c2Respectively, a learning factor, adjusting a learning step length, R is [0,1 ]]A random value function between. In the optimization process, a fitness evaluation function is designed according to the error rate of stage carrier communication to update the algorithm, the optimal frequency point design parameter most suitable for the current communication environment is found after multiple iterations, and the particle group speed is updated by adopting a formula 18 so as to ensure that the algorithm does not fall into a local optimal solution.
Figure BDA0003005366290000103
In the formula NpThe method is characterized in that the particle similarity gain is adopted, when the similarity between model particles is high and the iteration is carried out for more than 5 times in a smaller optimal frequency point range, the current optimal frequency point parameters are recorded and a larger N is setpAnd (4) continuously searching other optimal frequency points by the model, and selecting the frequency point parameters with optimal performance when the iteration is finished. And (3) taking the obtained parameters as labels, taking the communication environment sampling data as training samples to form a training data set, and obtaining a large number of training data sets for training the self-adaptive frequency hopping anti-interference model after optimizing enough environments.
2.3 training adaptive models with DBN-BP deep networks
The Deep Belief Network (DBN) is a network structure formed by stacking a plurality of Restricted Boltzmann Machines (RBMs) generating a neural network structure and trained by using a contrast Divergence algorithm (CD), and can realize a DBN-BP network structure by adding a layer of BP network behind the DBN and bring an obtained data set into the DBN-BP Deep network for training. During training, firstly, a reasonable DBN-BP network structure is designed through experiments, the disturbance characteristics of a simulated engineering environment are gradually learned and extracted by utilizing an RBM layer, then the output characteristics of the RBM are brought into a BP network to be finely adjusted by utilizing a back propagation algorithm, and a relation matrix between a stage power line environment and an optimal frequency point is obtained through learning to obtain a final self-adaptive frequency hopping anti-interference model.
3. Experimental analysis and stage actual measurement data verification
3.1 Experimental training data acquisition
The carrier communication frequency band of a normal stage is set between 40 and 500kHz, the simulation communication environment information is sampled after model parameters are set, the initial frequency point updating parameters of a particle swarm algorithm are set, 1000 groups of data prepared in advance are sequentially coded and then sent, then the error rate of received signals is calculated to update the speed and the position of particles, the optimal frequency point parameters are selected from three frequency bands respectively, different disturbance model parameters are randomly set according to the actual power line environment of the stage to obtain 30000 groups of data aiming at 3000 groups of different working conditions, the optimal frequency point parameters under each different environment are respectively solved by adopting the particle swarm algorithm, and a training data set is formed by training a self-adaptive frequency point updating model together with the original environment parameters. Data acquisition is realized through Matlab and Simulink models.
The average power of the signal is set to 1, the modulation order is set to 2 orders, the data bit period is set to 0.01, the frequency modulation factor is set to 0.5, the center frequency of the carrier frequency of the signal is set to 100kHz, the minimum single-frequency interval is set to 100, the bandwidth of the signal is set to 400, and the order of the filter is set to 48 orders. Fig. 8 shows a set of signals with interference and their results after de-hopping.
The signal transmission bit error curve when the algorithm iterates is shown in fig. 9.
Repeating the steps, setting different interference types, and simulating to obtain 30000 groups of data and frequency point updating parameters thereof to form a training data group.
3.2 adaptive frequency hopping anti-interference model training
Firstly, different DBN layers are set, the influence of different layer number settings on the DBN network to extract input signal characteristics is compared by respectively comparing the reconstruction errors of RBMs of each layer in DBNs with different layer number settings, and the results shown in Table 1 are obtained through comparison.
TABLE 1 DBN number of layers comparison table
Figure BDA0003005366290000111
After many attempts, on the premise of only considering the reconstruction error and the calculation time, when the number of DBN layers is set to be 4 and the parameter of each layer is set to be [430,320,210,100], the final reconstruction error is relatively small and the calculation time is short. The features extracted from the RBM layer are brought into a BP layer, which is set to 3 layers, training and fine-tuning are performed using a back propagation method, 500 groups are randomly extracted from 3000 groups of data at each period of training for iterative training, and the training results after 10 periods of training are shown in fig. 10.
It can be seen from fig. 10 that as the iteration error per phase becomes smaller, the training error almost coincides with the test error at the time of the phase 10 iteration. 500 sets of data were extracted for testing, and the model averaged test error was about 0.0186.
3.3 engineering environmental applications
In order to further verify the actual engineering application effect of the adaptive frequency hopping communication technology, a certain positive debugging engineering is selected to sample the actual engineering power line environment, and the feasibility of the model is further verified by training the adaptive frequency hopping model by using the actual engineering environment data. Sampling environment has stage jib control system, elevating platform control system, side platform control system and rotatory platform control system, through the different equipment of actual operation, and total sampling experiment data 10000 groups carry out actual engineering analysis to stage carrier communication self-adaptation frequency hopping anti-interference system. And setting a particle swarm algorithm until the error rate is 0.01, acquiring a training data set, and training a self-adaptive frequency hopping anti-interference model by using a deep network.
3.4 engineering data validation result analysis
Firstly, randomly extracting 150-plus-300 groups of data for each working condition, respectively sampling the traditional frequency hopping communication mode CFH and the self-adaptive frequency hopping communication mode AFH to repeatedly send 1000 groups of randomly generated communication data, setting the error rate of the traditional frequency hopping communication to be less than 0.01, and obtaining the engineering verification result shown in the table 2.
Table 2 test results of carrier communication adaptive anti-interference technique for stage under different working conditions
Figure BDA0003005366290000121
The table 2 shows that the stage carrier communication adaptive anti-interference technology also shows an ideal effect when the actual engineering environment data is used for testing, and the communication time is greatly reduced while the error rate is controlled in a lower range.
4. Conclusion
Through data experiments and analysis, compared with the traditional random frequency hopping technology in stage carrier communication, the self-adaptive frequency hopping technology can adaptively select the optimal communication frequency band aiming at the current stage power line environment, so that the tedious link of repeatedly selecting the communication frequency band according to the error rate of a transmission signal in the traditional frequency hopping communication is avoided, and the communication efficiency is greatly improved. The characteristics of the stage power line communication environment can be comprehensively and effectively learned by utilizing the deep neural network, compared with the method that the communication channel is updated for multiple times through feedback information according to the preset setting, more communication channel selection modes can be provided, the optimal communication channel can be found more quickly and accurately, more communication time is saved on the premise of effectively reducing the communication error rate, and further the operation efficiency of the stage carrier communication system is integrally improved.
The technical solutions provided by the embodiments of the present invention are described in detail above, and the principles and embodiments of the present invention are explained in the present document by using specific examples, and the descriptions of the embodiments are only used to help understanding the principles of the embodiments of the present invention.

Claims (9)

1. The stage carrier communication self-adaptive frequency hopping anti-interference technology based on the deep network is characterized by comprising the following steps of:
s1: establishing a stage carrier communication system model: on the basis of a stage carrier communication system, interference factors influencing communication quality in the stage carrier communication system are analyzed, a port network is established to simulate a stage power line carrier communication process by combining the actual operation condition of the stage carrier communication system, and a mathematical expression model of the stage power line carrier communication process is as follows:
Figure FDA0003005366280000011
wherein g is a channel trade-off factor with a value less than 1, d is a channel length, and alpha is0And alpha1K is an exponential decay factor with the value of 0.5 to 1 as a decay parameter, vpFor signal propagation speed, the expression is:
vp＝d/τ (2)
wherein tau is the channel transmission time delay,
Figure FDA0003005366280000012
c0＝3×108m/s, epsilon are constants of insulating dielectric layers of communication power lines, and the impact of a channel model is correspondingly as follows:
h(t)＝kδ(t-τ) (3)
in the formula, k is a channel balance factor, and tau is channel transmission path time delay;
s2: stage carrier communication interference analysis: according to a stage carrier communication system, analyzing interference factors influencing communication quality in stage communication;
s3: stage carrier communication environment simulation sampling: setting parameter ranges of each model aiming at the same environment after disturbance analysis of interference factors, and then simulating a similar communication environment by adopting a random algorithm to obtain a plurality of groups of experimental data;
s4: optimizing optimal frequency point parameters by using a particle swarm algorithm: setting the frequency point parameter particle swarm in the multidimensional space as X, and respectively setting the position and speed of the particle i as XiAnd viThe optimal History bit is marked as PibeGlobal optimum position is noted as GbeThen, the velocity update formula and the position update formula of the particle are shown as follows:
Figure FDA0003005366280000013
Figure FDA0003005366280000014
wherein
Figure FDA0003005366280000021
Is the velocity of the kth iteration of the ith particle,
Figure FDA0003005366280000022
for the position of the kth iteration of the ith particle, c1、c2Respectively, a learning factor, adjusting a learning step length, R is [0,1 ]]A random value function between;
in the optimization process, according to an error rate updating algorithm of stage carrier communication, finding out an optimal frequency point design parameter most suitable for the current communication environment after repeated iteration, taking the obtained parameter as a label, taking communication environment sampling data as a training sample to form a training data set, and obtaining a large number of training data sets for training a self-adaptive frequency hopping anti-interference model after optimizing enough environments;
s5: training the adaptive model by using a DBN-BP deep network: the method comprises the steps of designing a DBN-BP network structure through experiments, gradually learning and extracting disturbance characteristics of a simulated engineering environment by utilizing a RBM layer, bringing output characteristics of the RBM into a BP network, carrying out fine adjustment by utilizing a back propagation algorithm, and obtaining a relation matrix between a stage power line environment and an optimal frequency point through learning to obtain a final self-adaptive frequency hopping anti-interference model.
2. The stage carrier communication self-adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 1, wherein: the stage carrier communication system in step S1 is mainly composed of a carrier controller, an encoding device, a coupling device, a decoupling device, and a decoding device, where the carrier controller is configured to process the collected information of each stage device and each detection device, and control the encoder to encode the transmission data, the encoded information is accessed to the stage power line through the coupler, and the transmitted encoded signal is separated by the decoupling device at the receiving location, and then the transmitted data information is obtained by the decoding device.
3. The stage carrier communication self-adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 1, wherein: the interference factors in the step S2 and the step S3 mainly include colored background noise, narrow-band noise, asynchronous noise with power frequency, synchronous noise with power frequency, and impulse noise.
4. The stage carrier communication adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 3, wherein: the colored background noise is noise generated after a plurality of low-intensity noise sources in a stage communication system are fused, the colored background noise can be regarded as a stable random process, a model of the colored background noise is simulated by a group of white noise sequences with the average value of 0 through AR filtering, and the process is shown as the following formula:
Figure FDA0003005366280000031
where a (i) is the AR coefficient, w (n) is a set of white noises with an average value of 0.
5. The stage carrier communication adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 3, wherein: the stage narrow-band is noise caused by frequency domain interference, and the narrow-band noise is simulated by superposing a plurality of sine functions, and the simulation process is shown as the following formula:
Figure FDA0003005366280000032
wherein f isiIs the frequency of the ith waveform, AiIn order to be the amplitude value,
Figure FDA0003005366280000033
is the phase.
6. The stage carrier communication adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 3, wherein: the noise synchronous with the power frequency is periodic noise which is synchronous with the power frequency and has long duration and large frequency coverage range and is generated when stage equipment with a thyristor works under the power frequency, the periodic noise is described by pulse width, pulse existing time and pulse interval, a sine signal attenuated according to an exponential is selected for simulation, and the simulation process is shown as the following formula.
Figure FDA0003005366280000034
Where N is the number of pulses, τiIs the ith pulse time constant, tarriThe ith pulse epoch.
7. The stage carrier communication adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 6, wherein: the noise synchronous with the power frequency and the noise asynchronous with the power frequency are both simulated by a single periodic noise pulse.
8. The stage carrier communication self-adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 1, wherein: the impulse noise mainly refers to random noise which is generated by transient signals generated in the processes of stage equipment switching and start-stop control, has short duration and concentrated energy, is described by pulse width, pulse existing time and pulse interval, and is simulated by a single pulse general function.
9. The stage carrier communication self-adaptive frequency hopping anti-jamming technology based on the deep network as claimed in claim 1, wherein: in the step S4, in order to ensure that the algorithm does not fall into the locally optimal solution, the particle swarm velocity is updated by the following formula:
Figure FDA0003005366280000041
in the formula NpThe method is characterized in that the particle similarity gain is adopted, when the similarity between model particles is high and the iteration is carried out for more than 5 times in a smaller optimal frequency point range, the current optimal frequency point parameters are recorded and a larger N is setpAnd (4) continuously searching other optimal frequency points by the model, and selecting the frequency point parameters with optimal performance when the iteration is finished.
CN202110360575.4A 2021-04-02 2021-04-02 Stage carrier communication self-adaptive frequency hopping anti-interference method based on deep network Active CN113098565B (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN202110360575.4A CN113098565B (en) 2021-04-02 2021-04-02 Stage carrier communication self-adaptive frequency hopping anti-interference method based on deep network

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN202110360575.4A CN113098565B (en) 2021-04-02 2021-04-02 Stage carrier communication self-adaptive frequency hopping anti-interference method based on deep network

Publications (2)

Publication Number Publication Date
CN113098565A true CN113098565A (en) 2021-07-09
CN113098565B CN113098565B (en) 2022-06-07

Family

ID=76673240

Family Applications (1)

Application Number Title Priority Date Filing Date
CN202110360575.4A Active CN113098565B (en) 2021-04-02 2021-04-02 Stage carrier communication self-adaptive frequency hopping anti-interference method based on deep network

Country Status (1)

Country Link
CN (1) CN113098565B (en)

Cited By (3)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114124153A (en) * 2021-11-05 2022-03-01 中通服咨询设计研究院有限公司 Wireless communication anti-interference method and wireless communication module
CN115372928A (en) * 2022-10-21 2022-11-22 成都信息工程大学 DCNN electromagnetic interference identification method and device based on I/Q data
CN115580510A (en) * 2022-12-08 2023-01-06 大尧信息科技(湖南)有限公司 Frequency hopping pattern generation method based on deep neural network and intelligent communication system

Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS62283735A (en) * 1986-05-30 1987-12-09 Kyocera Corp Spread spectrum system
TW201043056A (en) * 2009-05-22 2010-12-01 Primax Electronics Ltd Method of automatic frequency hopping for wireless device
CN103346989A (en) * 2013-05-06 2013-10-09 中国人民解放军重庆通信学院 Multi-path frequency hopping-based single-channel blind source separation anti-interference communication system and method thereof
EP2787664A1 (en) * 2011-12-01 2014-10-08 ZTE Corporation Time synchronization method and apparatus
CN109302262A (en) * 2018-09-27 2019-02-01 电子科技大学 A kind of communication anti-interference method determining Gradient Reinforcement Learning based on depth
CN110048742A (en) * 2019-05-05 2019-07-23 珠海市杰理科技股份有限公司 WLAN and the signal transmit-receive method of Bluetooth coexistence, device, system and storage medium
CN111726217A (en) * 2020-06-29 2020-09-29 中南大学 Deep reinforcement learning-based autonomous frequency selection method and system for broadband wireless communication

Patent Citations (7)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPS62283735A (en) * 1986-05-30 1987-12-09 Kyocera Corp Spread spectrum system
TW201043056A (en) * 2009-05-22 2010-12-01 Primax Electronics Ltd Method of automatic frequency hopping for wireless device
EP2787664A1 (en) * 2011-12-01 2014-10-08 ZTE Corporation Time synchronization method and apparatus
CN103346989A (en) * 2013-05-06 2013-10-09 中国人民解放军重庆通信学院 Multi-path frequency hopping-based single-channel blind source separation anti-interference communication system and method thereof
CN109302262A (en) * 2018-09-27 2019-02-01 电子科技大学 A kind of communication anti-interference method determining Gradient Reinforcement Learning based on depth
CN110048742A (en) * 2019-05-05 2019-07-23 珠海市杰理科技股份有限公司 WLAN and the signal transmit-receive method of Bluetooth coexistence, device, system and storage medium
CN111726217A (en) * 2020-06-29 2020-09-29 中南大学 Deep reinforcement learning-based autonomous frequency selection method and system for broadband wireless communication

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
刘猛: "基于深度学习的抗干扰决策技术研究", 《中国优秀硕士学位论文全文数据库》, 15 February 2020 (2020-02-15), pages 9 - 58 *

Cited By (6)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN114124153A (en) * 2021-11-05 2022-03-01 中通服咨询设计研究院有限公司 Wireless communication anti-interference method and wireless communication module
CN114124153B (en) * 2021-11-05 2023-05-26 中通服咨询设计研究院有限公司 Wireless communication anti-interference method and wireless communication module
CN115372928A (en) * 2022-10-21 2022-11-22 成都信息工程大学 DCNN electromagnetic interference identification method and device based on I/Q data
CN115372928B (en) * 2022-10-21 2023-02-28 成都信息工程大学 DCNN electromagnetic interference identification method and device based on I/Q data
CN115580510A (en) * 2022-12-08 2023-01-06 大尧信息科技(湖南)有限公司 Frequency hopping pattern generation method based on deep neural network and intelligent communication system
CN115580510B (en) * 2022-12-08 2023-04-07 大尧信息科技(湖南)有限公司 Frequency hopping pattern generation method based on deep neural network and intelligent communication system

Also Published As

Publication number Publication date
CN113098565B (en) 2022-06-07

Similar Documents

Publication Publication Date Title
CN113098565B (en) Stage carrier communication self-adaptive frequency hopping anti-interference method based on deep network
JP2021513251A (en) Fiber Optic Nonlinear Compensation Using Neural Networks
CN112733811B (en) Method for identifying underwater sound signal modulation modes based on improved dense neural network
CN114513278B (en) Intelligent interference method, device and system based on electromagnetic spectrum characteristic cognition
CN114520758B (en) Signal modulation identification method based on instantaneous characteristics
CN107947868A (en) A kind of more band hyperbolic frequency-modulation spread spectrum acoustic communications based on subband selection activation
CN102158254B (en) Method and device for receiving signal
Lazarovych et al. Software implemented enhanced efficiency BPSK demodulator based on perceptron model with randomization
Hu et al. Signal Classification in Real-time Based on SDR using Convolutional Neural Network
Liu et al. Research on discovery of radio communication relationship based on correlation analysis
CN113242201B (en) Wireless signal enhanced demodulation method and system based on generation classification network
CN102571656B (en) Method and device for eliminating interference from signal in short-range wireless network
CN102611470B (en) Anti-interference receiving machine in differential frequency hopping communication system and operation method of anti-interference receiving machine
CN106209283B (en) A kind of power line communication channel analysis method for variable working condition environment
CN112564830A (en) Deep learning-based dual-mode orthogonal frequency division multiplexing index modulation detection method and device
Ismail et al. Performance study of HF communication using NOMA over narrowband HF channel
Burns et al. Synthetic wireless signal generation for neural network algorithms
CN105262506A (en) Time-frequency overlapped Gaussian amplitude modulation communication signal separation method
CN109951245A (en) Multi-source end noise modeling method under the complex topology environment of multinode multirouting
Song et al. Simulation of ground-air integrated 5G networking adaptive anti-jamming technology
CN202551028U (en) Anti-jam receiver in differential frequency hopping communication system
Baishya et al. Emulating the Statistical Properties of Indoor Power Line Colored Background Noise for Development of a Power Line Noise Simulator
Zou et al. Frequency diversity of the power line time service system
CN117294364A (en) Frequency hopping channel simulation method based on broadband frequency response
Yuan et al. Active noise control method of power communication signal based on BP neural network

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