CN110798275A - Mine multimode wireless signal accurate identification method - Google Patents
Mine multimode wireless signal accurate identification method Download PDFInfo
- Publication number
- CN110798275A CN110798275A CN201910981429.6A CN201910981429A CN110798275A CN 110798275 A CN110798275 A CN 110798275A CN 201910981429 A CN201910981429 A CN 201910981429A CN 110798275 A CN110798275 A CN 110798275A
- Authority
- CN
- China
- Prior art keywords
- mine
- signal
- fading
- signals
- order cumulant
- 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.)
- Pending
Links
Images
Classifications
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B17/00—Monitoring; Testing
- H04B17/30—Monitoring; Testing of propagation channels
- H04B17/391—Modelling the propagation channel
- H04B17/3911—Fading models or fading generators
-
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04L—TRANSMISSION OF DIGITAL INFORMATION, e.g. TELEGRAPHIC COMMUNICATION
- H04L27/00—Modulated-carrier systems
- H04L27/0012—Modulated-carrier systems arrangements for identifying the type of modulation
Abstract
The invention belongs to the technical field of signal identification, and discloses a mine multimode wireless signal accurate identification method, which is used for mine communication signal identification: signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of a classifier; analyzing a mine wireless channel model: analyzing large-scale fading characteristics and analyzing small-scale fading characteristics; feature extraction based on high-order cumulant: and constructing second moment, fourth moment and second-order cumulant fourth-order cumulant normalized characteristic quantity of the identified mine communication signals, and analyzing the influence of a fading channel on the high-order cumulant. The invention solves the problem of low recognition rate of the common SVM classifier under low signal-to-noise ratio, and improves the modulation recognition performance to a certain extent; under the environment of three channels with signal-to-noise ratio of-5, the average recognition rate of the four signals can reach 80 percent; under the environment of three channels with signal-to-noise ratios larger than-3, the average recognition rate of four signals can reach 90%.
Description
Technical Field
The invention belongs to the technical field of signal identification, and particularly relates to a mine multimode wireless signal accurate identification method.
Background
Currently, the current state of the art commonly used in the industry is such that:
the mine tunnel is a special limited space, the environment of the mine tunnel is much more complex and harsh compared with civil air defense and railway tunnels, the tunnel is narrow, the ground is rough, and the periphery of the mine tunnel is surrounded by coal and carbon stones, a bracket, an air door, a power line and the like. Unlike wired communication, radio waves are severely refracted and reflected when propagating in an underground roadway, and thus severe multipath fading can occur.
With the development of wireless technologies, the widespread application of wireless technologies such as WiFi, 3G, 4G, and 5G has had a great impact on various aspects of people's life. The wireless technology is applied to information platform construction of the coal industry, and it is very necessary to construct an integral solution which is relatively uniform and fully considers the independence and safety of each system. WiFi is suitable for wireless transmission of videos and data in fixed places and stable environments, is simple to install and arrange and high in transmission speed, and is mainly applied to video monitoring of inclined roadway transportation, main and auxiliary well lifting, gas drilling sites and mining working faces. However, WiFi signals have small coverage radius and poor penetration capability, and have high requirements or are difficult to implement for wireless transmission of voice, video and data in complex, mobile, large-span and other environments [9 ]. With the large-scale mining of the 4G technology, underground video monitoring, data acquisition, information distribution, video conferencing and the like are all on a wireless communication system platform. The 5G technology will also be used for mine communication in the near future.
The multimode wireless communication technology under the complex environment of the coal mine is a necessary trend for the information development and application of the coal mine. The future coal mine underground wireless technology application development trend is to be the continuous fusion of various wireless technologies such as WiFi, zigbee, 3G, 4G, 5G and the like, and the fusion of the multi-mode self-adaptive underground wireless communication system is the direction of the future coal mine underground complex environment application development. The method adopts a multimode wireless self-adaptive communication technology, realizes the fusion of various networks by utilizing the characteristics of different wireless access technologies, needs to adapt the multimode wireless communication technology and quickly adapt and model the underground wireless network environment so as to ensure the stability, reliability and real-time performance of the underground communication system. Therefore, how to construct the adaptive matching and fusion of the multimode wireless technology in the complex mine environment is a problem to be solved urgently in multimode wireless communication.
In summary, the problems of the prior art are as follows: currently, most of researches on signal modulation identification are based on an ideal white gaussian noise environment, and the proposed method is only suitable for an ideal simulated channel environment and currently lacks researches on signal identification under low signal-to-noise ratio and complex channel environments. Especially, the narrow rock stratum of the mine underground roadway space seriously attenuates electromagnetic wave loss, and the accurate identification of the signal modulation mode under the channel environment is particularly important, but the research on the aspect is lacked at present.
Heterogeneous networks and complex wireless signals on underground spectrum resources coexist dynamically, the prior art cannot realize accurate multi-system modulation recognition fusion in automatic modulation recognition of mine environments, and a common SVM classifier has low recognition rate under low signal-to-noise ratio.
The difficulty of solving the technical problems is as follows:
the signal modulation mode identification method mainly comprises two methods, one method is based on the likelihood function and mainly takes Bayes theory and hypothesis test, and the other method is mainly based on the pattern identification of signal feature extraction. The first method requires a large amount of prior probability as support, and the computational complexity and spatial complexity of the method are high, which is difficult to be widely applied in practice. In addition, the error of parameter estimation such as frequency offset and phase offset also causes the deviation of probability distribution function, resulting in greatly impaired recognition performance. In response to these problems, feature-based modulation scheme identification methods have been developed and are receiving increasing attention from researchers under study make internal disorder or usurp. However, the selection of features in this method will determine the recognition effect, so the features with effective signals are always the concern of researchers. For a complex channel environment in a well, the high-order cumulant of white gaussian noise is zero, so that the high-order cumulant can be used as a strong characteristic. Analysis of the mine channel yields parameters that closely approximate the downhole Nakagami channel and the large shadow fading channel. Further theoretical reasoning finally deduces the influence on the high-order cumulant under the two channels in the well and presents the influence in a formula. This provides an accurate data sample and channel model for our method.
The significance of solving the technical problems is as follows:
through the selection of effective features and the analysis of the influence of a downhole channel model on the features, the method is of great importance to downhole signal identification. The former has conducted little research on the identification of signals in fading environment because of the failure to select the characteristics of effective signals and the lack of research on the influence of fading channels on the characteristics of signals. The invention provides the influence of a Nakagami fading channel and a large shadow fading channel under a mine on the high-order cumulant of a modulation signal and deduces a formula of the high-order cumulant under the two environments. This provides a theoretical simulation basis for the following signal identification work based on complex channels.
Disclosure of Invention
Aiming at the problems in the prior art, the invention provides a mine multimode wireless signal accurate identification method.
The invention is realized in this way, a mine multimode wireless signal accurate identification method, comprising:
step one, mine communication signal identification: performing signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of a classifier;
step two, analyzing a mine wireless channel model: analyzing large-scale fading characteristics and analyzing small-scale fading characteristics;
step three, feature extraction based on high-order cumulant: and constructing second moment, fourth moment and second-order cumulant fourth-order cumulant normalized characteristic quantity of the identified mine communication signals, and analyzing the influence of a fading channel on the high-order cumulant.
Further, the method for identifying the signals under the low signal-to-noise ratio of the classifier in the step one comprises the following steps:
firstly, preprocessing after receiving a signal at a receiver end, and performing variable frequency down filtering and noise reduction;
secondly, analyzing two models of fading channels under the mine and analyzing high-order cumulant of the two fading channels to different modulation modes;
the third step: based on the construction and extraction of the high-order cumulant features of the signals under two fading channels, a PSO-SVM classifier and a GA-SVM classifier are selected for classification.
And thirdly, classifying by using a PSO-SVM classifier and a GA-SVM classifier, which specifically comprises the following steps:
dividing a data sample set into a test data set and a training data set, carrying out optimization processing on punishment factors and kernel functions of SVM in the training data set by using a particle swarm algorithm and a genetic algorithm to obtain an optimized SVM model, carrying out test classification on the test set by using the SVM model, and carrying out simulation by using a Matlab platform environment.
Further, the particle swarm algorithm is operated on a PSO-SVM classifier, and the specific steps comprise:
① initializing each particle in the particle swarm space, including initial position information X and initial speed information Y, the setting mode adopts random setting;
② substituting the initial position and velocity into the iterative formula of velocity and position to obtain new position information P of each particlei；
③ calculating the fitness value of each particle;
④ for each particle, the difference between the fitness value and the optimal position P experienced by itself is comparedidIf the fitness value is better, the historical best position P of the particle is updatedidIs the current position of the particle;
⑤ for each particle, the fitness value is compared with the current global best position P of the population by the difference of the fitness valuegdIf the self fitness value is better, the global optimal position P is updatedgdIs the current position of the particle;
⑥ updating the new position and new velocity of each particle according to the particle position updating formula and velocity updating formula;
⑦ executing circularly until reaching best position or iteration number, and ending the calculation process when meeting the end condition to obtain PidAnd PgdOtherwise, the process jumps to step ③ to continue the iteration.
Further, the genetic algorithm is operated on a classifier of the GA-SVM, and specifically comprises the following steps:
selecting a fitness function when parameter searching is carried out:
F＝100r,0≤r≤1；
the fitness function classification accuracy rate is optimized in a mode of maximization:
and coding the parameter C sum of the support vector machine to search the optimal classifier model parameter for the classification accuracy of the SVM.
Further, in step two, the method for analyzing the large-scale fading characteristics comprises:
the average received power of a signal is logarithmically attenuated as the distance increases, and the path loss pl (d) for the distance d between any transmitting and receiving antennas is expressed by a random normal logarithmic distribution:
wherein, is the reference distance; pl (d) random variables that follow a lognormal distribution; shadow fading S (d) is a random variable following a normal distribution with a mean value of zero, and PL (d) is
Mean, random variable of normal distribution of variance.
The analysis of the small-scale fading characteristics comprises the following steps: nakagami fading channel analysis:
after the transmission signal is transmitted in a short distance or a short time, the intensity of the channel is changed sharply; on different multipath signals, random frequency modulation caused by time-varying Doppler frequency shift exists; propagation due to multipath propagation delay.
Further, in the third step, in the fourth-order cumulant normalized characteristic quantity of the second-order moment, the fourth-order moment and the second-order cumulant of the identified mine communication signals, for a zero-mean k-order stationary random process x (t), the definition of the k-order cumulant is as follows:
Ckx(f1,f2,…fk-1)＝Cum(x(t),x(t+f1),…x(t+fk-1))；
the cumulative amount of stationary random complex signals is defined as:
C20＝Cum(X,X)＝M20
C21＝Cum(X,X*)＝M21
in analyzing the effect of the Nakagami fading channel on the high order cumulant,
the probability density function for a Nakagami fading channel is:
the second, fourth and sixth order moments are expressed as follows:
E(R2)＝Ω；
further, in step three, the fading channel is a large-scale fading channel, and is caused by shadow fading, and a probability density function of the shadow fading is as follows:
the second, fourth and sixth order moments are expressed as follows:
another object of the present invention is to provide a method for accurately identifying a mine multimode wireless signal, which comprises:
mine communication signal identification module: the system is used for signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of the classifier;
the mine wireless channel model analysis module: the mine communication signal identification module is connected with the mine communication signal identification module and is used for large-scale fading characteristic analysis and small-scale fading characteristic analysis;
the feature extraction module based on the high-order cumulant: and the second-order moment, the fourth-order moment and the fourth-order cumulant normalized characteristic quantity of the identified mine communication signals are constructed, and the influence of a fading channel on the high-order cumulant is analyzed.
The invention also aims to provide an information data processing terminal for realizing the method for accurately identifying the mine multimode wireless signals.
Another object of the present invention is to provide a computer-readable storage medium, which includes instructions that, when run on a computer, cause the computer to execute the method for accurately identifying a mine multimode wireless signal.
In summary, the advantages and positive effects of the invention are:
the method is provided for the first time based on the calculation of the high-order cumulant under two channel environments in the mine. A Support Vector Machine (SVM) is used as a classifier, the fourth-order cumulant of the signal is used as the input of the SVM, and classification and identification of the BPSK, the OFDM, the 16QAM and the 64QAM are achieved. The method for optimizing SVM classification recognition solves the problem of low recognition rate of a common SVM classifier under low signal-to-noise ratio, and improves modulation recognition performance to a certain extent.
The method uses Matlab platform environment for simulation, and simulation results show that the average recognition rate of four signals can reach more than 80% in three channel environments with signal-to-noise ratio of-5; under the environment of three channels with the signal-to-noise ratio larger than-3, the average recognition rate of the four signals can reach more than 90%.
The underground environment is not considered in the prior art, the underground environment is analyzed for the first time, the field test is carried out by using hardware equipment, and a theoretically feasible solution is provided for underground signal identification.
Drawings
Fig. 1 is a flowchart of a method for accurately identifying a mine multimode wireless signal according to an embodiment of the present invention.
Fig. 2 is a schematic diagram of signal identification provided by an embodiment of the present invention.
Fig. 3 is a schematic diagram of a small-scale fading model Nakagami fading channel provided by an embodiment of the invention.
Fig. 4 is a graph illustrating the variation of the fourth-order cumulant with noise under Nakagami according to an embodiment of the present invention.
Fig. 5 is a graph illustrating the variation of fourth-order cumulant with noise under shadow fading according to an embodiment of the present invention.
Fig. 6 is a flow chart of a particle swarm method provided by the embodiment of the invention.
FIG. 7 is a schematic diagram of the actual classification and the predicted classification of the Nakagami test set provided by the embodiment of the invention.
Fig. 8 is a schematic diagram of actual classification and predicted classification of a shadow fading test set according to an embodiment of the present invention.
FIG. 9 is a flowchart of the GA-SVM algorithm provided by the embodiment of the present invention.
FIG. 10 is a schematic diagram of the actual classification and the predicted classification of the Nakagami test set provided by the embodiment of the invention.
Fig. 11 is a schematic diagram of actual classification and predicted classification of a shadow fading test set according to an embodiment of the present invention.
Fig. 12 is a comparison graph of the recognition rates of the PSO-SVM and the GA-SVM provided by the embodiment of the present invention under different channels.
Fig. 13 is a schematic diagram of a mine multimode wireless signal accurate identification system provided by an embodiment of the invention.
In the figure: 1. a mine communication signal identification module; 2. a mine wireless channel model analysis module; 3. and the feature extraction module is based on the high-order cumulant.
Detailed Description
In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is further described in detail with reference to the following embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.
Heterogeneous networks and complex wireless signals on underground spectrum resources coexist dynamically, the prior art cannot realize accurate multi-system modulation recognition fusion, and a common SVM classifier has low recognition rate under low signal-to-noise ratio. In the prior art, in the automatic modulation recognition based on PSO-SVM and GA-SVM in the mine environment, the recognition accuracy of the signal modulation mode is low due to the diversification and complication of the communication signal modulation mode.
To solve the above technical problems, the present invention will be described in detail with reference to the accompanying drawings.
As shown in fig. 1, the method for accurately identifying a mine multimode wireless signal provided in the embodiment of the present invention includes:
s101, mine communication signal identification: and performing signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of the classifier.
S102, analyzing a mine wireless channel model: large scale fading characteristic analysis and small scale fading characteristic analysis.
S103, feature extraction based on the high-order cumulant: and constructing second moment, fourth moment and second-order cumulant fourth-order cumulant normalized characteristic quantity of the identified mine communication signals, and analyzing the influence of a fading channel on the high-order cumulant.
The method for identifying the signals of the classifier under the condition of low signal-to-noise ratio in the step S101 comprises the following steps:
firstly, preprocessing is carried out after a signal is received at a receiver end, and variable frequency down-filtering and noise reduction are carried out.
And secondly, analyzing two models of fading channels under the mine and analyzing high-order cumulant of the two fading channels to different modulation modes.
The third step: based on the construction and extraction of the high-order cumulant features of the signals under two fading channels, a PSO-SVM classifier and a GA-SVM classifier are selected for classification.
And thirdly, classifying by using a PSO-SVM classifier and a GA-SVM classifier, which specifically comprises the following steps:
dividing a data sample set into a test data set and a training data set, carrying out optimization processing on punishment factors and kernel functions of SVM in the training data set by using a particle swarm algorithm and a genetic algorithm to obtain an optimized SVM model, carrying out test classification on the test set by using the SVM model, and carrying out simulation by using a Matlab platform environment.
The technical solution and technical effects of the present invention are further described below with reference to specific embodiments.
Example 1:
the method for accurately identifying the mine multimode wireless signals provided by the embodiment of the invention comprises the following steps:
1) mine communication signal identification
The modulation recognition is carried out on the signal, and basically, the signal is subjected to three parts of signal preprocessing, characteristic parameter extraction and classifier design. The basic flow diagram of the modulation identification algorithm is shown in fig. 2.
In the signal preprocessing stage, down-conversion processing of signals, elimination of in-phase and quadrature components, elimination of carrier components and the like are mainly carried out; the characteristic parameter extraction is mainly carried out according to different modulation modes of the signals and different characteristics of the signals according to different environments, so that the signals are classified and identified according to the characteristics; the general classifier design mainly includes mode recognition-based and prior probability-based, and currently, the commonly used classifiers are an SVM classifier, a BP classifier, a decision tree classifier and the like.
2) Mine wireless channel model analysis
① analysis of large scale fading characteristics
The large-scale fading of the mine roadway is caused by a shadow effect caused by the fact that electromagnetic waves are blocked by large obstacles such as roadway walls and related equipment and large-distance fluctuation between the transmitting and receiving antennas in the propagation process, and reflects the process of slow change of signal field intensity between the transmitting and receiving equipment at a large distance.
In general, large-scale fading is affected by conditions such as time, propagation distance, carrier frequency, and the like, while in practical process, the effects of distance and carrier are generally ignored, and in a transmission model of theoretical analysis and actual measurement, it can be found that the average received power of a signal is logarithmically attenuated with the increase of distance, and for the distance d between any transmitting and receiving antennas, the path loss pl (d) can be represented by a random normal log distribution:
wherein, the reference distance (generally taking 1m) is adopted; PL (d) is a random variable that follows a lognormal distribution. The shadow fading s (d) is a random variable following a normal distribution with a mean of zero and a variance of zero. This indicates that PL (d) is
Mean, random variable of normal distribution of variance.
② analysis of small scale fading characteristics
Small-scale fading is caused by multipath propagation of the wireless channel, and the main effects are: after the transmission signal is transmitted in a short distance or a short time, the intensity of the channel is changed sharply; on different multi-jing signals, random frequency modulation caused by time-varying Doppler frequency shift exists; propagation due to multipath propagation delay.
As shown in fig. 3, the most obvious small-scale fading model Nakagami fading channel is a typical example in a mine environment.
The Nakagami distribution controls the amplitude fading condition of the multi-jing signal along with the change of the value of m, and the smaller m is, the deeper the fading degree is. When m is 0.5, it will be a single-sided gaussian case, and when m is 1, it can be converted to a rayleigh distribution. When m is more than 0.5 and less than 1, the multipath fading is more serious than the Rayleigh fading; when m is 1, the channel fading condition is rayleigh fading; when m is more than 1, the channel fading condition is better than Rayleigh fading
3) Feature extraction based on high-order cumulant
3.1) higher order cumulant theory analysis
Since the transmission signal and the white gaussian noise are independent of each other and the second or higher order accumulation amount of the gaussian signal is 0, the second or higher order accumulation amount of the reception signal will not be affected by the white gaussian noise, and because of this, more and more people use the higher order accumulation amount as a characteristic parameter in the identification algorithm of the modulation signal.
For a zero-mean stationary random process of order k, x (t), the cumulative quantity of order k is defined as:
Ckx(f1,f2,…fk-1)＝Cum(x(t),x(t+f1),…x(t+fk-1))。
the cumulative amount of stationary random complex signals is defined as:
C20＝Cum(X,X)＝M20
C21＝Cum(X,X*)＝M21
because normalization can improve the classification accuracy of the SVM and can avoid the influence of noise and the power of a received signal on identification, the embodiment of the invention constructs the following normalized characteristic quantities:
the second and fourth moments, second order cumulants, and fourth order cumulants for each signal are shown in the following table:
TABLE 1 second order cumulant of signals fourth order cumulant
| C20 | C21 | C40 | |
| BPSK | 1 | 1 | -2 |
| |
0 | 1 | -0.68 |
| |
0 | 1 | -0.619 |
| |
0 | 1 | 0 |
3.2) Effect of fading channel on higher order cumulants
3.2.1) Nakagami fading channel
The influence of the Nakagami fading channel on the second and fourth order cumulants of the signal, the probability density function of the Nakagami channel is:
the second, fourth and sixth order moments are expressed as follows:
E(R2)＝Ω。
according to previous research on the mine environment, based on the analysis of measured data, when m is 0.85 and Ω is 1, the measured data is closer to the mine environment, and the following is the variation of the fourth-order cumulant of the signal under the Nakagami fading channel with the noise ratio, which can be obtained from the simulation result: when the signal-to-noise ratio is greater than 0dB, the fluctuation of the fourth-order accumulative quantity of the signal is small, and the fourth-order accumulative quantity values of the signal are close to the theoretical value of analysis; when the signal-to-noise ratio is less than 0dB, the fourth-order cumulant simulation fluctuation of the signals is large, and the difference of the fourth-order cumulant of the 16QAM and 64QAM signals is small.
Table 2 fourth order cumulant of signals with m 0.85 and Ω 1
| BPSK | 16QAM | 64QAM | OFDM | |
| C40 | -0.824 | -1.48 | -1.35 | 0 |
A graph of the fourth order cumulative quantity versus noise under Nakagami is shown in FIG. 4.
3.2.2) Large Scale fading channel
The large-scale fading of the mine roadway is complex, the consideration is more, it is assumed that the fading is mainly caused by shadow fading, and the probability density function of the shadow fading is as follows:
the second, fourth and sixth order moments are expressed as follows:
according to the previous research on the mine environment, based on the analysis of the measured data, when μ is 0 and σ is 1.1, the measured data is closer to the mine environment, and the following is the variation of the fourth-order cumulant of the signal under the shadow fading channel with the noise ratio, which can be obtained from the simulation result: when the signal-to-noise ratio is greater than 0dB, the fluctuation of the fourth-order accumulative quantity of the signal is small, and the fourth-order accumulative quantity values of the signal are close to the theoretical value of analysis; when the signal-to-noise ratio is less than 0dB, the fourth-order cumulant simulation fluctuation of the signals is large, and the difference of the fourth-order cumulant of the 16QAM and 64QAM signals is small.
TABLE 3 fourth order cumulant of mu-0, sigma-1.1 signals
| BPSK | 16QAM | 64QAM | OFDM | |
| C40 | 15615.09 | -10876.26 | -9900.59 | 0 |
The graph of the fourth-order cumulant variation with noise under shadow fading is shown in fig. 5.
4) The invention is further described below in connection with simulation analysis.
Because the recognition effect of the common SVM classifier algorithm is not ideal under the condition of low signal to noise ratio, aiming at the problem of low recognition rate of signals under the condition of low signal to noise ratio, the invention introduces a particle swarm optimization algorithm and a genetic optimization algorithm to optimize SVM parameters, namely, carries out optimization processing on penalty factors and kernel function parameters, carries out verification through a simulation experiment, and finally carries out comparative analysis on the recognition effect of the two algorithms.
4.1) particle swarm optimization
4.1.1) PSO-SVM classifier design
The particle swarm algorithm is realized by the following steps:
① initializing each particle in the particle swarm space, including initial position information X and initial speed information Y, the setting mode adopts random setting;
② substituting the initial position and velocity into the iterative formula of velocity and position to obtain new position information P of each particlei；
③ calculating the fitness value of each particle;
④ for each particle, by difference of fitness value, compares its optimal position P experienced by itselfidIf its fitness value is better, the historical best position P of the particle is updatedidIs the current position of the particle;
⑤ for each particle, comparing it with the current global best position P of the population by the difference in fitness valuegdIf the self fitness value is more optimal, the global optimal position P is updatedgdIs the current position of the particle;
⑥ updating the new position and new velocity of each particle according to the particle position update formula and velocity update formula;
⑦, until the best position or iteration number is reached, i.e. the end condition is satisfied, the calculation process is ended, and P is obtainedidAnd PgdOtherwise, the process jumps to step ③ to continue the iteration.
The structural block diagram of the whole algorithm is shown in fig. 6.
4.1.2) simulation experiments and analysis of results
(1) Simulation conditions
Subject: BPSK signals, 16QAM signals, 64QAM signals, OFDM signals.
Setting simulation parameters: assuming that m under the Nakagami fading channel is 0.85 and 1; the value under shadow fading is 0 and the sigma value takes 1. The OFDM signal has 128 subcarriers, the cycle length is 32, and the modulation scheme adopted on each subcarrier is QPSK. The noise source is Gaussian white noise.
SVM parameter setting: selecting Gaussian radial basis kernels as kernel functions, and performing binary tree classification.
PSO parameter setting: maximum number of iterations 200, population size 20, initial weight ωmax0.9, final weight ωmin0.4, learning factor c1＝1.5、c2The search interval for penalty factor C and is 1,100, respectively, 1.7]And [0,0.1]And the fitness function is the identification accuracy of the SVM classifier.
(2) Simulation experiment
Experiment: the value interval of the signal-to-noise ratio is [ -5dB, 5dB ], the change step length is 2dB, each type of digital signals is respectively and randomly generated at 200 under each signal-to-noise ratio, and the data in the data set is normalized. Taking the fourth-order cumulant of the signals as a feature set, taking the fourth-order cumulant of the signals as the input of the SVM, selecting 100 data of each signal as a training set, namely 4 × 100 training data, and constructing a test data set to be 4 × 100 according to the same way, so as to obtain the accurate recognition rates of the signals under three different channels as shown in the following table:
TABLE 4 recognition rate under Nakagami fading channel
| Signal | -5dB | -3dB | -1dB | 1dB | ≥3dB |
| BPSK | 99.57％ | 100％ | 100％ | 100％ | 100％ |
| 16QAM | 67.33％ | 90.45％ | 96.07％ | 100％ | 100％ |
| 64QAM | 62.68％ | 89.57％ | 94.03％ | 99.51％ | 100％ |
| OFDM | 97.94％ | 100％ | 100％ | 100％ | 100％ |
| Average recognition rate | 81.88％ | 90.01％ | 95.05％ | 99.51％ | 100％ |
TABLE 5 recognition rate under shadow fading
| Signal | -5dB | -3dB | -1dB | ≥ |
| BPSK | ||||
| 100％ | 100％ | 100％ | 100％ | |
| 16QAM | 84.93％ | 91.78％ | 100％ | 100％ |
| 64QAM | 81.29％ | 90.28％ | 96.54％ | 100 |
| OFDM | ||||
| 100％ | 100％ | 100％ | 100％ | |
| Average recognition rate | 83.11％ | 91.03％ | 96.54％ | 100％ |
The Nakagami fading channel is based on a classification simulation diagram of the PSO-SVM classifier at-1 dB, as shown in FIG. 7.
The shadow fading is based on a classification simulation diagram of the PSO-SVM classifier at-1 dB and a fitness function as shown in fig. 8.
4.2) genetic Algorithm
4.2.1 classifier implementation of GA-SVM
The kernel function and the penalty factor of the support vector machine play an important role in constructing a classifier with good classification capability.
The kernel function used is a Gaussian kernel function
Thus, the SVM problem can be transformed into the following:
wherein Q is a semi-positive definite matrix whose elements are:
Qij＝yiyjK(xi-xj)
the decision function is then:
the basis for optimizing the genetic optimization algorithm is a fitness function, and the fitness function is a mapping F from an individual space χ to a real space R: . It can be demonstrated that genetic algorithms can converge probabilistically to an optimal solution in the solution space. Here, an individual represents a set of parameters of the support vector machine algorithm, and the fitness function corresponding to the individual is the algorithm performance under the set of parameters. Selecting a fitness function when a genetic algorithm is adopted for parameter search:
F＝100r,0≤r≤1
namely the classification accuracy of the fitness function, and the optimization target is in a maximization form:
s.t.Clow≤C≤Cup
γlow≤γ≤γup
and coding the parameter C sum of the support vector machine to search the optimal classifier model parameter for the classification accuracy of the SVM. Genetic algorithms, while not necessarily optimal, can achieve satisfactory solutions in a short time.
The algorithm flow of the GA-SVM is shown in FIG. 9.
4.2.2) simulation experiment and result analysis of GA
(1) Simulation conditions
Subject: BPSK signals, 16QAM signals, 64QAM signals, OFDM signals.
Setting simulation parameters: assuming that m under the Nakagami fading channel is 0.85 and 1; the value under shadow fading is 0 and the sigma value takes 1. The OFDM signal has 128 subcarriers, the cycle length is 32, and the modulation scheme adopted on each subcarrier is QPSK. The noise source is Gaussian white noise.
SVM parameter setting: selecting Gaussian radial basis kernels as kernel functions, and performing binary tree classification.
GA parameter setting: the evolution frequency is 200, the population scale is 20, the cross probability is 0.4, the variation probability is 0.01, the search intervals of the penalty factors C and C are respectively [1,100] and [0,0.1], and the fitness function is the identification accuracy of the SVM classifier.
(2) Simulation experiment
Experiment one: the value interval of the signal-to-noise ratio is [ -5dB, 20dB ], the change step length is 1dB, each type of digital signals is respectively and randomly generated at 200 under each signal-to-noise ratio, and the data in the data set is normalized. And taking the fourth-order cumulant of the signals as a feature set, taking the fourth-order cumulant of the signals as the input of the SVM, selecting 100 data of each signal as a training set, namely 4 x 100 training data, and constructing a test data set of 4 x 100 according to the same mode. The accurate identification rates of the signals obtained under three different channels are shown in the following table:
TABLE 6 recognition rate under Nakagami fading channel
| Signal | -5dB | -3dB | -1dB | 1dB | ≥3dB |
| BPSK | 99.5％ | 100％ | 100％ | 100％ | 100％ |
| 16QAM | 65.52％ | 90.06％ | 97.67％ | 99.5％ | 100％ |
| 64QAM | 62.24％ | 88.06％ | 91.43％ | 98.52％ | 100％ |
| OFDM | 98.02％ | 100％ | 100％ | 100％ | 100％ |
| Average recognition rate | 81.32％ | 89.60％ | 94.55％ | 99.01％ | 100％ |
TABLE 7 recognition rate under shadow fading
| Signal | -5dB | -3dB | -1dB | ≥ |
| BPSK | ||||
| 100％ | 100％ | 100％ | 100％ | |
| 16QAM | 85.03％ | 91.58％ | 98.17％ | 100％ |
| 64QAM | 82.29％ | 89.6％ | 94.89％ | 100 |
| OFDM | ||||
| 100％ | 100％ | 100％ | 100％ | |
| Average recognition rate | 83.66％ | 90.59％ | 96.53％ | 100％ |
Under the condition that the signal is more than or equal to 10dB, the recognition rate of the signal reaches the recognition rate of hundreds, so that the advantage of the GA-SVM classifier is not fully embodied under the condition of high signal-to-noise ratio.
The Nakagami fading channel is shown in FIG. 11 based on a classification simulation diagram of the GA-SVM classifier at-1 dB and a fitness function.
The shadow fading is based on a classification simulation graph of the GA-SVM classifier at-1 dB and a fitness function as shown in fig. 12.
4.3) comparison of GA algorithm and PSO algorithm
Comparing the PSO-SVM classifier and the GA-SVM classifier under the same simulation environment, the following simulation results can be obtained, and can be obtained from the results: whether the PSO-SVM classifier or the GA-SVM classifier is adopted, the classification effect of the PSO-SVM classifier or the GA-SVM classifier is obviously improved compared with that of a conventional SVM classifier, and the advantages and the disadvantages of the PSO-SVM classifier or the GA-SVM classifier are compared and analyzed.
The same points are as follows: PSO-SVM and GA-SVM classifiers are all bionic algorithms. PSO is mainly proposed by simulating social behaviors such as bird foraging, human cognition and the like, and GA mainly depends on the rule of survival of a suitable person in biological evolution.
The PSO-SVM classifier and the GA-SVM classifier both belong to a global optimization algorithm. The initial population is randomly generated in the solution space, so that the algorithm searches in the global solution space, and the searching is mainly focused on the part with high performance.
The search is performed according to the individual adaptation information, and therefore is not limited by the functional constraint.
For the high-dimensional complex problem, the characteristics of early convergence and poor convergence are often encountered, and the convergence to the optimal point cannot be ensured.
The difference is as follows: PSO has memory properties that preserve the results of previous classification, whereas GA is memoryless, and the results of previous classification are destroyed as the population changes.
The particles in the PSO share information only by searching for the optimal point currently, so to a large extent this is a single information sharing mechanism. In GA, however, chromosomes share information with each other, so that the entire population moves to the most regions.
The encoding technology and genetic operation of GA are simpler, and PSO has no crossover and mutation operation compared with GA, and the particles are updated only by internal speed, so the principle is simpler, the parameters are fewer, and the implementation is easier.
The convergence rate of the GA algorithm is too slow and is easy to fall into a local optimal value, while the convergence rate of the PSO is much faster than that of the GA, which is the most important difference between the GA and the PSO.
Through comprehensive experimental demonstration and the analysis, the PSO algorithm in the invention has slightly better performance than the GA algorithm, and the training time and the testing time are less than those of the GA algorithm, and more accurate parameters can be obtained.
As shown in fig. 13, the present invention provides a system for accurately identifying multimode wireless signals in a mine, comprising:
mine communication signal identification module 1: the method is used for signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of the classifier.
The mine wireless channel model analysis module 2: and the mine communication signal identification module is connected with the mine communication signal identification module and is used for large-scale fading characteristic analysis and small-scale fading characteristic analysis.
The feature extraction module 3 based on the high-order cumulant: and the second-order moment, the fourth-order moment and the fourth-order cumulant normalized characteristic quantity of the identified mine communication signals are constructed, and the influence of a fading channel on the high-order cumulant is analyzed.
The present invention will be further described with reference to effects.
The invention mainly analyzes the automatic modulation recognition algorithm based on PSO-SVM and GA-SVM in the mine environment. The diversification and complication of communication signal modulation modes are particularly urgent and important for the identification of the signal modulation modes. The SVM belongs to a machine learning mode, can better handle complex nonlinear problems due to strong learning and recognition capabilities, has better robustness and potential fault tolerance, and is widely applied to modulation recognition. The algorithm based on the optimization of SVM parameters improves the modulation recognition performance to a certain extent.
The invention analyzes the influence of the mine environment on the signal characteristic parameters, uses a Support Vector Machine (SVM) as a classifier, uses the fourth-order cumulant of the signal as the input of the SVM, and realizes the classification and identification of the BPSK, OFDM, 16QAM and 64QAM signals. Aiming at the problem of low recognition rate of a common SVM classifier under low signal-to-noise ratio, a method for optimizing SVM classification recognition is provided. Dividing a data sample set into a test data set and a training data set, carrying out optimization processing on penalty factors and kernel functions of the SVM in the training data set by using a particle swarm algorithm and a genetic algorithm to obtain an optimized SVM model, and carrying out test classification on the test set by using the model. The Matlab platform environment is used for simulation, and simulation results show that the average recognition rate of four signals can reach more than 80% in three channel environments with the signal-to-noise ratio of-5 dB; under the environment of three channels with the signal-to-noise ratio larger than-3 dB, the average recognition rate of the four signals can reach more than 90 percent.
In the above embodiments, the implementation may be wholly or partially realized by software, hardware, firmware, or any combination thereof. When used in whole or in part, can be implemented in a computer program product that includes one or more computer instructions. When loaded or executed on a computer, cause the flow or functions according to embodiments of the invention to occur, in whole or in part. The computer may be a general purpose computer, a special purpose computer, a network of computers, or other programmable device. The computer instructions may be stored in a computer readable storage medium or transmitted from one computer readable storage medium to another, for example, the computer instructions may be transmitted from one website site, computer, server, or data center to another website site, computer, server, or data center via wire (e.g., coaxial cable, fiber optic, Digital Subscriber Line (DSL), or wireless (e.g., infrared, wireless, microwave, etc.)). The computer-readable storage medium can be any available medium that can be accessed by a computer or a data storage device, such as a server, a data center, etc., that includes one or more of the available media. The usable medium may be a magnetic medium (e.g., floppy Disk, hard Disk, magnetic tape), an optical medium (e.g., DVD), or a semiconductor medium (e.g., Solid State Disk (SSD)), among others.
The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.
Claims (10)
1. A mine multimode wireless signal accurate identification method is characterized by comprising the following steps:
step one, mine communication signal identification: performing signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of a classifier;
step two, analyzing a mine wireless channel model: analyzing large-scale fading characteristics and analyzing small-scale fading characteristics;
step three, feature extraction based on high-order cumulant: and constructing second moment, fourth moment and second-order cumulant fourth-order cumulant normalized characteristic quantity of the identified mine communication signals, and analyzing the influence of a fading channel on the high-order cumulant.
2. The method for accurately identifying mine multimode wireless signals as claimed in claim 1, wherein the step one classifier low signal-to-noise ratio signal identification method comprises the following steps:
firstly, preprocessing after receiving a signal at a receiver end, and performing variable frequency down filtering and noise reduction;
secondly, analyzing two models of fading channels under the mine and analyzing high-order cumulant of the two fading channels to different modulation modes;
the third step: based on the construction and extraction of the high-order cumulant features of the signals under two fading channels, a PSO-SVM classifier and a GA-SVM classifier are selected for classification.
3. The method for accurately identifying the mine multimode wireless signals as recited in claim 2, wherein the third step selects a PSO-SVM classifier and a GA-SVM classifier for classification, and specifically comprises the following steps:
dividing a data sample set into a test data set and a training data set, carrying out optimization processing on punishment factors and kernel functions of SVM in the training data set by using a particle swarm algorithm and a genetic algorithm to obtain an optimized SVM model, carrying out test classification on the test set by using the SVM model, and carrying out simulation by using a Matlab platform environment;
the particle swarm algorithm is operated on a PSO-SVM classifier, and the specific steps comprise:
① initializing each particle in the particle swarm space, including initial position information X and initial speed information Y, the setting mode adopts random setting;
② substituting the initial position and velocity into the iterative formula of velocity and position to obtain new position information P of each particlei；
③ calculating the fitness value of each particle;
④ for each particle, the difference between the fitness value and the optimal position P experienced by itself is comparedidIf the fitness value is better, the historical best position P of the particle is updatedidIs the current position of the particle;
⑤ for each particle, the fitness value is compared with the current global best position P of the population by the difference of the fitness valuegdIf the self fitness value is better, the global optimal position P is updatedgdIs the current position of the particle;
⑥ updating the new position and new velocity of each particle according to the particle position updating formula and velocity updating formula;
⑦ executing circularly until reaching best position or iteration number, and ending the calculation process when meeting the end condition to obtain PidAnd PgdOtherwise, the process jumps to step ③ to continue the iteration.
4. The method for accurately identifying the mine multimode wireless signals as claimed in claim 3, wherein the genetic algorithm is operated in a classifier of a GA-SVM, and specifically comprises the following steps:
selecting a fitness function when parameter searching is carried out:
F＝100r,0≤r≤1；
the fitness function classification accuracy rate is optimized in a mode of maximization:
and coding the parameter C sum of the support vector machine to search the optimal classifier model parameter for the classification accuracy of the SVM.
5. The method for accurately identifying mine multimode wireless signals according to claim 1, wherein in the second step, the method for analyzing the large-scale fading characteristics comprises the following steps:
the average received power of a signal is logarithmically attenuated as the distance increases, and the path loss pl (d) for the distance d between any transmitting and receiving antennas is expressed by a random normal logarithmic distribution:
wherein, is the reference distance; pl (d) random variables that follow a lognormal distribution; shadow fading S (d) is a random variable following a normal distribution with a mean value of zero, and PL (d) is
Mean, random variable of normal distribution of variance.
The method for analyzing the small-scale fading characteristics comprises the following steps:
after the transmission signal is transmitted in a short distance or a short time, the intensity of the channel is changed sharply; on different multipath signals, random frequency modulation caused by time-varying Doppler frequency shift exists; propagation due to multipath propagation delay.
6. The method for accurately identifying mine multimode wireless signals as claimed in claim 1, wherein in the third step, in the second-order moment, fourth-order moment, second-order cumulant and fourth-order cumulant normalized characteristic quantities of the identified mine communication signals, for a zero-mean k-order stationary random process x (t), the k-order cumulant is defined as:
Ckx(f1,f2,…fk-1)＝Cum(x(t),x(t+f1),…x(t+fk-1))；
the cumulative amount of stationary random complex signals is defined as:
C20＝Cum(X,X)＝M20
C21＝Cum(X,X*)＝M21
in analyzing the effects of fading channels on high order cumulants,
the probability density function of a fading channel is:
the second, fourth and sixth order moments are expressed as follows:
E(R2)＝Ω；
7. the method for accurately identifying mine multimode wireless signals according to claim 1, characterized in that in step three, the fading channel is a large-scale fading channel, and is caused by shadow fading, and the probability density function of the shadow fading is as follows:
the second, fourth and sixth order moments are expressed as follows:
8. the mine multimode wireless signal accurate identification method according to claim 1, the mine multimode wireless signal accurate identification system, comprising:
mine communication signal identification module: the system is used for signal preprocessing, characteristic parameter extraction and signal identification under low signal-to-noise ratio of the classifier;
the mine wireless channel model analysis module: the mine communication signal identification module is connected with the mine communication signal identification module and is used for large-scale fading characteristic analysis and small-scale fading characteristic analysis;
the feature extraction module based on the high-order cumulant: and the second-order moment, the fourth-order moment and the fourth-order cumulant normalized characteristic quantity of the identified mine communication signals are constructed, and the influence of a fading channel on the high-order cumulant is analyzed.
9. An information data processing terminal for implementing the mine multimode wireless signal accurate identification method according to any one of claims 1 to 7.
10. A computer-readable storage medium comprising instructions that, when executed on a computer, cause the computer to perform the method of accurate identification of mine multimode wireless signals of any one of claims 1 to 7.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910981429.6A CN110798275A (en) | 2019-10-16 | 2019-10-16 | Mine multimode wireless signal accurate identification method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910981429.6A CN110798275A (en) | 2019-10-16 | 2019-10-16 | Mine multimode wireless signal accurate identification method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN110798275A true CN110798275A (en) | 2020-02-14 |
Family
ID=69439239
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910981429.6A Pending CN110798275A (en) | 2019-10-16 | 2019-10-16 | Mine multimode wireless signal accurate identification method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110798275A (en) |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111327554A (en) * | 2020-02-27 | 2020-06-23 | 电子科技大学 | Feature extraction method for digital modulation signal identification |
| CN111343115A (en) * | 2020-02-19 | 2020-06-26 | 北京邮电大学 | 5G communication modulation signal identification method and system |
| CN111835444A (en) * | 2020-06-17 | 2020-10-27 | 武汉大学 | Wireless channel scene identification method and system |
| CN113240801A (en) * | 2021-06-08 | 2021-08-10 | 矿冶科技集团有限公司 | Three-dimensional reconstruction method and device for material pile, electronic equipment and storage medium |
| CN113655468A (en) * | 2021-07-16 | 2021-11-16 | 广州大学 | Unmanned aerial vehicle auxiliary positioning method and system, storage medium and terminal equipment |
| CN114826452A (en) * | 2022-03-19 | 2022-07-29 | 西安电子科技大学 | Nakagami-m parameter estimation method, system, equipment and terminal under low SNR |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040189525A1 (en) * | 2003-03-28 | 2004-09-30 | Beadle Edward R. | System and method for cumulant-based geolocation of cooperative and non-cooperative RF transmitters |
| CN103617147A (en) * | 2013-11-27 | 2014-03-05 | 中国地质大学（武汉） | Method for identifying mine water-inrush source |
| CN106027183A (en) * | 2016-05-30 | 2016-10-12 | 河海大学 | Method capable of quickly evaluating cumulative distribution performance of composite fading channel |
| CN106059972A (en) * | 2016-05-25 | 2016-10-26 | 北京邮电大学 | Modulation identification method under MIMO related channel based on machine learning algorithm |
| CN109787926A (en) * | 2018-12-24 | 2019-05-21 | 合肥工业大学 | A kind of digital signal modulation mode recognition methods |
-
2019
- 2019-10-16 CN CN201910981429.6A patent/CN110798275A/en active Pending
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20040189525A1 (en) * | 2003-03-28 | 2004-09-30 | Beadle Edward R. | System and method for cumulant-based geolocation of cooperative and non-cooperative RF transmitters |
| CN103617147A (en) * | 2013-11-27 | 2014-03-05 | 中国地质大学（武汉） | Method for identifying mine water-inrush source |
| CN106059972A (en) * | 2016-05-25 | 2016-10-26 | 北京邮电大学 | Modulation identification method under MIMO related channel based on machine learning algorithm |
| CN106027183A (en) * | 2016-05-30 | 2016-10-12 | 河海大学 | Method capable of quickly evaluating cumulative distribution performance of composite fading channel |
| CN109787926A (en) * | 2018-12-24 | 2019-05-21 | 合肥工业大学 | A kind of digital signal modulation mode recognition methods |
Non-Patent Citations (3)
| Title |
|---|
| WANG ANYI等: "Research on Digital Signal Recognition Based on Higher Order Cumulants", 《2019 INTERNATIONAL CONFERENCE ON INTELLIGENT TRANSPORTATION, BIG DATA & SMART CITY (ICITBS)》 * |
| 武金莉: "基于PSO优化LSSVM参数的矿井巷道场强预测", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
| 郗茜: "基于LS_SVM的半圆拱形巷道无线信道建模与预测", 《中国优秀硕士学位论文全文数据库信息科技辑》 * |
Cited By (11)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111343115A (en) * | 2020-02-19 | 2020-06-26 | 北京邮电大学 | 5G communication modulation signal identification method and system |
| CN111343115B (en) * | 2020-02-19 | 2021-06-29 | 北京邮电大学 | 5G communication modulation signal identification method and system |
| CN111327554A (en) * | 2020-02-27 | 2020-06-23 | 电子科技大学 | Feature extraction method for digital modulation signal identification |
| CN111835444A (en) * | 2020-06-17 | 2020-10-27 | 武汉大学 | Wireless channel scene identification method and system |
| CN111835444B (en) * | 2020-06-17 | 2021-04-27 | 武汉大学 | Wireless channel scene identification method and system |
| CN113240801A (en) * | 2021-06-08 | 2021-08-10 | 矿冶科技集团有限公司 | Three-dimensional reconstruction method and device for material pile, electronic equipment and storage medium |
| CN113240801B (en) * | 2021-06-08 | 2023-09-19 | 矿冶科技集团有限公司 | Material stack three-dimensional reconstruction method and device, electronic equipment and storage medium |
| CN113655468A (en) * | 2021-07-16 | 2021-11-16 | 广州大学 | Unmanned aerial vehicle auxiliary positioning method and system, storage medium and terminal equipment |
| CN113655468B (en) * | 2021-07-16 | 2023-05-23 | 广州大学 | Unmanned aerial vehicle auxiliary positioning method, unmanned aerial vehicle auxiliary positioning system, storage medium and terminal equipment |
| CN114826452A (en) * | 2022-03-19 | 2022-07-29 | 西安电子科技大学 | Nakagami-m parameter estimation method, system, equipment and terminal under low SNR |
| CN114826452B (en) * | 2022-03-19 | 2023-05-16 | 西安电子科技大学 | Nakagami-m parameter estimation method, system, equipment and terminal under low SNR |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110798275A (en) | Mine multimode wireless signal accurate identification method | |
| CN109547140B (en) | Double-medium mixed fading communication system performance analysis method based on AF protocol | |
| CN112924749B (en) | Unsupervised counterstudy electromagnetic spectrum abnormal signal detection method | |
| CN113543068B (en) | Forest area unmanned aerial vehicle network deployment method and system based on hierarchical clustering | |
| Cavalcanti et al. | A hybrid path loss prediction model based on artificial neural networks using empirical models for LTE and LTE-A at 800 MHz and 2600 MHz | |
| Feng et al. | A link quality prediction method for wireless sensor networks based on XGBoost | |
| CN108900269B (en) | Error rate performance analysis method of wireless and power line dual-medium cooperative communication system | |
| CN113473480A (en) | Improved reinforcement learning network coverage optimization method facing cellular network | |
| CN111464469A (en) | Hybrid digital modulation mode identification method based on neural network | |
| CN109787696A (en) | The cognitive radio resource allocation methods of case-based reasioning and cooperation Q study | |
| Zhang et al. | Edge computing resource allocation method for mining 5G communication system | |
| Wang et al. | Optimal DoS attack strategy for cyber-physical systems: A Stackelberg game-theoretical approach | |
| Zhang et al. | Cultivated Land Monitoring System Based on Dynamic Wake-Up UAV and Wireless of Distributed Storage. | |
| CN117319994A (en) | Design method for unmanned aerial vehicle auxiliary robust hidden communication against active inspector | |
| CN108599880B (en) | Civil aviation ground-air intercom system interference early warning method based on convolutional neural network | |
| CN108900325B (en) | Method for evaluating adaptability of power communication service and wireless private network technology | |
| CN113890633B (en) | Underwater acoustic communication system self-adaptive selection method based on deep neural network | |
| Xu et al. | Cyber Intrusion Detection Based on a Mutative Scale Chaotic Bat Algorithm with Backpropagation Neural Network | |
| CN112737678A (en) | Method and device for optimizing performance of free space optical communication system channel | |
| Kahriman et al. | Surface gateway placement optimization for underwater sensor networks | |
| Li et al. | Decision Model of Wireless Communication Scheme Evaluation via Interval Number | |
| CN115833991B (en) | Symbiotic radio intelligent interference method | |
| Fu et al. | Wireless D2D network link scheduling based on graph embedding | |
| Govindhan et al. | Improving High data rates in Milli meter Wave Communication networks via Long short term memory Technique | |
| CN113766599B (en) | Relay node selection method, device and terminal of pipe gallery wireless sensor 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 | ||
| WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20200214 |
|
| WD01 | Invention patent application deemed withdrawn after publication |