CN115225441A - Unmanned aerial vehicle cluster communication waveform identification method in complex environment - Google Patents
Unmanned aerial vehicle cluster communication waveform identification method in complex environmentDownload PDFInfo
- Publication number
- CN115225441A CN115225441ACN202210858726.3ACN202210858726ACN115225441ACN 115225441 ACN115225441 ACN 115225441ACN 202210858726 ACN202210858726 ACN 202210858726ACN 115225441 ACN115225441 ACN 115225441A
- Authority
- CN
- China
- Prior art keywords
- signal
- unmanned aerial
- aerial vehicle
- cluster communication
- channel
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Images
Classifications
- 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
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
- G—PHYSICS
- G06—COMPUTING OR CALCULATING; 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
- H—ELECTRICITY
- H04—ELECTRIC COMMUNICATION TECHNIQUE
- H04B—TRANSMISSION
- H04B7/00—Radio transmission systems, i.e. using radiation field
- H04B7/14—Relay systems
- H04B7/15—Active relay systems
- H04B7/185—Space-based or airborne stations; Stations for satellite systems
- H04B7/18502—Airborne stations
- 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)
- General Physics & Mathematics (AREA)
- Biomedical Technology (AREA)
- Data Mining & Analysis (AREA)
- Signal Processing (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Artificial Intelligence (AREA)
- Software Systems (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- Computer Networks & Wireless Communication (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Aviation & Aerospace Engineering (AREA)
- Astronomy & Astrophysics (AREA)
- Mobile Radio Communication Systems (AREA)
Abstract
Description
Translated fromChinese技术领域technical field
本发明属于数字通信技术领域，具体涉及一种复杂环境下的无人机集群通信波形识别方法。The invention belongs to the technical field of digital communication, and in particular relates to a method for identifying a communication waveform of an unmanned aerial vehicle group in a complex environment.
背景技术Background technique
由于无人机集群具备高协调性、多功能性、强抗毁性的特点，其在侦察定位、电子对抗、通信情报等军事领域的地位愈加突出。无人机集群作战将成为未来战争的主要形式之一，为维护国家安全，研究无人机集群反制技术迫在眉睫，实现无人机集群通信波形识别是无人机集群反制技术研究的关键环节。Because UAV swarms have the characteristics of high coordination, versatility and strong invulnerability, their status in the military fields such as reconnaissance positioning, electronic countermeasures, and communication intelligence is becoming more and more prominent. UAV swarm operations will become one of the main forms of future warfare. In order to maintain national security, it is imminent to study the UAV swarm countermeasure technology. The realization of UAV swarm communication waveform identification is a key link in the research of UAV swarm countermeasure technology. .
非合作通信领域中，需要先截获敌方的通信信号，进行分析处理，获取敌方情报，而信号的自动调制识别是获得敌方情报的基础。调制识别发生在信号检测和解调之间，其基本任务是：对所截获的未知信号进行调制类型的分析、判决和分类。In the field of non-cooperative communication, it is necessary to intercept the enemy's communication signals, analyze and process them, and obtain enemy intelligence. The automatic modulation and identification of signals is the basis for obtaining enemy intelligence. Modulation identification occurs between signal detection and demodulation, and its basic task is to analyze, judge and classify the modulation type of the intercepted unknown signal.
目前的调制识别方法主要包括两大类：基于最大似然函数的调制识别和基于特征提取的调制识别。调制识别过程主要分为三步：信号预处理、特征提取和分类识别。信号预处理过程包括下变频、噪声抑制、载波频率或符号周期的估计等；特征提取是从信号中提取出能够表征信号调制类型的特征，如循环谱、高阶累积量、小波变换特征等；分类识别则是将所提取的特征与分类性能结合，选择合适的方式进行判决分类，如决策树、神经网络等。The current modulation identification methods mainly include two categories: modulation identification based on maximum likelihood function and modulation identification based on feature extraction. The modulation recognition process is mainly divided into three steps: signal preprocessing, feature extraction and classification recognition. The signal preprocessing process includes down-conversion, noise suppression, carrier frequency or symbol period estimation, etc. Feature extraction is to extract features from the signal that can characterize the type of signal modulation, such as cyclic spectrum, high-order cumulant, wavelet transform features, etc.; Classification and recognition is to combine the extracted features with the classification performance, and select an appropriate method for decision classification, such as decision tree, neural network, etc.
现有通信信号自动调制识别方法的研究大多考虑在高斯信道下，其研究所假定的信道条件过于理想。在实际通信中，特别是战场环境，信号在传输过程中不可避免的会发生多径效应和多普勒频移，同时还可能存在用户干扰。Most of the existing research on automatic modulation and identification methods for communication signals considers Gaussian channels, and the assumed channel conditions are too ideal. In actual communication, especially in the battlefield environment, multipath effect and Doppler frequency shift will inevitably occur in the signal transmission process, and there may also be user interference.
现有调制识别算法信道条件过于理想、低信噪比鲁棒性差，难以满足无人机集群通信波形识别的要求。The channel conditions of the existing modulation identification algorithms are too ideal, and the robustness of the low signal-to-noise ratio is poor, so it is difficult to meet the requirements of the UAV swarm communication waveform identification.
发明内容SUMMARY OF THE INVENTION
本发明所要解决的技术问题是针对上述现有技术的不足，提供一种复杂环境下的无人机集群通信波形识别方法，充分考虑了无人机集群通信信号在传输过程中的小尺度衰落和用户间干扰问题，提取无人机通信信号的循环平稳特征，通过稀疏自编码器(SparseAutoEncoder,SAE)神经网络实现了无人机集群波形的识别。The technical problem to be solved by the present invention is to aim at the deficiencies of the above-mentioned prior art, and to provide a method for identifying the communication waveform of the swarm of unmanned aerial vehicles in a complex environment, which fully considers the small-scale fading and fading of the swarm communication signal of the unmanned aerial vehicle during the transmission process. To solve the problem of inter-user interference, the cyclostationary features of UAV communication signals were extracted, and the swarm waveform recognition of UAVs was realized through sparse autoencoder (Sparse AutoEncoder, SAE) neural network.
为实现上述技术目的，本发明采取的技术方案为：In order to realize the above-mentioned technical purpose, the technical scheme adopted in the present invention is:
一种复杂环境下的无人机集群通信波形识别方法，包括：A method for identifying the communication waveform of a drone swarm in a complex environment, comprising:
步骤1：建立Alpha噪声干扰下的无人机集群通信多径衰落信道，获取通过信道的接收信号；Step 1: Establish a multipath fading channel for UAV swarm communication under Alpha noise interference, and obtain the received signal through the channel;
步骤2：对接收信号进行预处理，包括非线性变换、下变频、带通采样；Step 2: preprocessing the received signal, including nonlinear transformation, down-conversion, and band-pass sampling;
步骤3：提取预处理后信号的广义循环均值和广义循环谱特征，并构造不同信噪比下的无人机集群通信波形特征矩阵；Step 3: Extract the generalized cyclic mean and generalized cyclic spectrum features of the preprocessed signal, and construct the UAV swarm communication waveform feature matrix under different signal-to-noise ratios;
步骤4：将特征矩阵输入SAE神经网络进行训练测试，输出无人机集群通信波形类型，实现在Alpha噪声干扰、多径衰落与频移存在的复杂环境下的无人机集群通信波形的识别。Step 4: Input the feature matrix into the SAE neural network for training and testing, output the UAV swarm communication waveform type, and realize the identification of the UAV swarm communication waveform in the complex environment where Alpha noise interference, multipath fading and frequency shift exist.
为优化上述技术方案，采取的具体措施还包括：In order to optimize the above technical solutions, the specific measures taken also include:
上述的无人机集群通信波形类型包括BPSK、QPSK、2FSK、4FSK、2ASK、MSK。The above-mentioned UAV swarm communication waveform types include BPSK, QPSK, 2FSK, 4FSK, 2ASK, and MSK.
上述的步骤1中，采用TDL模型和Alpha稳定分布噪声建立无人机集群通信多径衰落信道，其信道参数基于3GPP TR 901.38技术报告设置，则截获的无人机通信信号，即获取的通过信道的接收信号表示为：In the
其中，x(t)为发送的调制信号，n(t)为Alpha稳定分布噪声；Among them, x(t) is the transmitted modulation signal, and n(t) is the Alpha stable distributed noise;
hl(t)和τl分别为第l条多径所对应的信道系数和时延，且0≤l≤L-1，L为多径衰落信道的可分辨路径数。hl (t) and τl are the channel coefficient and time delay corresponding to the l-th multipath, respectively, and 0≤l≤L-1, L is the number of distinguishable paths of the multipath fading channel.
上述的信道系数hl(t)由L个平坦衰落信号发生器的输出与每个抽头的功率相乘获得。The above-mentioned channel coefficient hl (t) is obtained by multiplying the output of the L flat fading signal generators by the power of each tap.
上述的步骤2包括：
1)对接收信号r(t)进行非线性变换：1) Perform nonlinear transformation on the received signal r(t):
其中，Δ为正常数；Among them, Δ is a normal number;
2)将非线性变换后的信号下变频至140MHz；2) Down-convert the nonlinearly transformed signal to 140MHz;
3)对下变频后的信号进行带通采样、低通滤波，将信号中心频率搬移至4MHz。3) Band-pass sampling and low-pass filtering are performed on the down-converted signal, and the center frequency of the signal is moved to 4MHz.
上述的步骤3包括：
1)提取广义循环平稳特征，预处理后的信号r'(t)的广义循环均值
其中，ε＝k/T为循环频率，Mr'(t)为信号r'(t)的均值；Among them, ε=k/T is the cycle frequency,Mr' (t) is the mean value of the signal r'(t);
2)信号r'(t)的广义循环谱密度
其中，
3)选择广义循环均值和广义循环谱的离散峰值及其个数作为特征，则某混合信噪比下的特征矩阵表示为：3) Select the generalized cyclic mean and the discrete peaks of the generalized cyclic spectrum and their numbers as features, then the feature matrix under a certain mixed signal-to-noise ratio is expressed as:
其中，
上述的步骤4中，利用SAE神经网络识别无人机集群通信波形；In the above-mentioned
所述SAE神经网络具有稀疏特性，其前向传播方程如下：The SAE neural network has sparse characteristics, and its forward propagation equation is as follows:
Sin＝σ[U(u1,...,um)×Xt+a] (11)Sin =σ[U(u1 ,...,um )×Xt +a] (11)
Sout＝Oin＝σ(η1×Sin×Vt(v1,v2...vn)+b) (12)Sout =Oin =σ(η1 ×Sin ×Vt (v1 ,v2 ...vn )+b) (12)
Oout＝f(η2×Sout×Wt(w1,w2…wp)+c) (13)Oout =f(η2 ×Sout ×Wt (w1 ,w2 ···wp )+c) (13)
其中，Sin，Sout和Oout分别是隐藏层的输入值，隐藏层的输出值和最终的输出值，U、V、W分别为对应连接层的权重矩阵，Xt为步骤3所构造的特征矩阵；Among them, Sin , Sout and Oout are the input value of the hidden layer, the output value of the hidden layer and the final output value, respectively, U, V, W are the weight matrix of the corresponding connection layer, Xt is constructed in
σ，f为激活函数，σ为tanh函数或sigmoid函数，f为Softmax函数，η是隐藏层的稀疏系数，a、b、c为各层偏置。σ, f are the activation function, σ is the tanh function or sigmoid function, f is the Softmax function, η is the sparse coefficient of the hidden layer, and a, b, and c are the biases of each layer.
本发明具有以下有益效果：The present invention has the following beneficial effects:
无人机集群通信电磁环境复杂，存在用户干扰、多径衰落与频移等现象，传统高斯信道下的波形识别算法在此场景下性能大幅降低，针对该问题，本发明提出了一种复杂环境下的无人机集群通信波形识别方法。本发明建立Alpha噪声干扰下的无人机集群通信多径衰落信道，获取通过信道的接收信号；对接收信号进行预处理，包括非线性变换、下变频、带通采样；提取预处理后信号的广义循环均值和广义循环谱特征，并构造不同信噪比下的无人机集群通信波形特征矩阵；将特征矩阵输入SAE神经网络进行训练测试，输出无人机集群通信波形类型，实现在Alpha噪声干扰、多径衰落与频移存在的复杂环境下的无人机集群通信波形的识别。仿真结果表明：本发明方法复杂环境下具有较强的鲁棒性，在信噪比为-10dB时仍能保证80％以上的识别准确率。The electromagnetic environment of UAV swarm communication is complex, and there are phenomena such as user interference, multipath fading and frequency shift. The performance of the waveform recognition algorithm under the traditional Gaussian channel is greatly reduced in this scenario. To solve this problem, the present invention proposes a complex environment. The identification method of the UAV swarm communication waveform. The invention establishes the multipath fading channel of the unmanned aerial vehicle cluster communication under the interference of Alpha noise, and obtains the received signal passing through the channel; preprocesses the received signal, including nonlinear transformation, down-conversion, and band-pass sampling; Generalized cyclic mean and generalized cyclic spectrum features, and construct the characteristic matrix of UAV swarm communication waveforms under different signal-to-noise ratios; input the feature matrix into the SAE neural network for training and testing, output the UAV swarm communication waveform type, and realize the Alpha noise Identification of UAV swarm communication waveforms in complex environments where interference, multipath fading and frequency shift exist. The simulation results show that the method of the present invention has strong robustness under complex environment, and can still ensure the recognition accuracy of more than 80% when the signal-to-noise ratio is -10dB.
附图说明Description of drawings
图1是集群通信场景图。Figure 1 is a diagram of a cluster communication scenario.
图2是无人机集群用户与侦收机的位置关系图。Figure 2 is a diagram of the positional relationship between the UAV swarm user and the reconnaissance aircraft.
图3是2FSK信号广义循环均值。Figure 3 is the generalized cyclic mean of the 2FSK signal.
图4是2ASK信号广义循环均值。Figure 4 is the generalized cyclic mean of the 2ASK signal.
图5是BPSK信号广义循环谱。Figure 5 is a generalized cyclic spectrum of a BPSK signal.
图6是QPSK信号广义循环谱。Figure 6 is a generalized cyclic spectrum of a QPSK signal.
图7是TDL-A模型的路径增益随信道路径和样本时间的变化关系。Fig. 7 shows the relationship between the path gain of the TDL-A model and the channel path and sample time.
图8是TDL-D模型的路径增益随信道路径和样本时间的变化关系。Fig. 8 shows the relationship between the path gain of the TDL-D model and the channel path and sample time.
图9是基于SAE的信号调制方式预测模型。FIG. 9 is a signal modulation method prediction model based on SAE.
图10是TDL-A信道下信号调制识别性能曲线。Figure 10 is a signal modulation identification performance curve under the TDL-A channel.
图11是TDL-D信道下信号调制识别性能曲线。Fig. 11 is the signal modulation identification performance curve under the TDL-D channel.
图12是TDL-A信道下不同场景信号调制识别性能曲线；Fig. 12 is the signal modulation recognition performance curve of different scenarios under the TDL-A channel;
图13是本发明方法流程图。Figure 13 is a flow chart of the method of the present invention.
具体实施方式Detailed ways
以下结合附图对本发明的实施例作进一步详细描述。The embodiments of the present invention will be described in further detail below with reference to the accompanying drawings.
本发明建立Alpha噪声干扰下的无人机集群通信多径衰落信道，设计无人机集群通信场景，截获无人机集群通信信号，然后对其进行预处理后，并提取其循环均值和循环累积量特征(即广义循环均值和广义循环谱特征)，建立无人机集群通信波形特征向量，生成不同信噪比下的多维特征矩阵，建立SAE神经网络作为无人机集群通信波形识别模型，实现6种无人机集群波形的识别，本发明所能识别的无人机集群通信波形类型包括BPSK、QPSK、2FSK、4FSK、2ASK、MSK。The invention establishes the multi-path fading channel of the unmanned aerial vehicle swarm communication under the interference of Alpha noise, designs the unmanned aerial vehicle swarm communication scene, intercepts the unmanned aerial vehicle swarm communication signal, preprocesses it, and extracts its cyclic mean and cyclic accumulation Quantitative features (i.e. generalized cyclic mean and generalized cyclic spectral features), establish the eigenvectors of the UAV swarm communication waveform, generate multi-dimensional feature matrices under different signal-to-noise ratios, and establish the SAE neural network as the UAV swarm communication waveform recognition model to realize The identification of 6 kinds of UAV swarm waveforms, the UAV swarm communication waveform types that can be identified by the present invention include BPSK, QPSK, 2FSK, 4FSK, 2ASK, and MSK.
如图13所示，本发明提出的无人机集群通信波形识别方法步骤如下：As shown in FIG. 13 , the steps of the method for identifying the communication waveform of the swarm communication of the UAV proposed by the present invention are as follows:
步骤1：根据3GPP TR 901.38建立Alpha噪声干扰下的无人机集群通信多径衰落信道，获取通过信道的接收信号；Step 1: According to 3GPP TR 901.38, establish a multipath fading channel for UAV swarm communication under Alpha noise interference, and obtain the received signal through the channel;
无人机集群通信过程不可避免的会存在多径效应、多普勒频移和用户干扰，本发明基于抽头延迟线(Tapped Delay Line，TDL)模型建立莱斯/瑞利衰落信道，并采用Alpha稳定分布噪声作为干扰。TDL信道的冲激响应可表示为：The multipath effect, Doppler frequency shift and user interference will inevitably exist in the communication process of the UAV swarm. The present invention establishes the Rice/Rayleigh fading channel based on the Tapped Delay Line (TDL) model, and adopts Alpha Stable distributed noise acts as a disturbance. The impulse response of a TDL channel can be expressed as:
其中，hl(t)和τl分别为第l条多径所对应的信道系数和时延，且0≤l≤L-1，L为多径信道的可分辨路径数。Among them, hl (t) and τl are the channel coefficients and time delays corresponding to the l-th multipath, respectively, and 0≤l≤L-1, L is the number of distinguishable paths of the multipath channel.
信道系数hl(t)由L个平坦衰落信号发生器(通过Jakes模型实现)的输出与每个抽头的功率相乘获得。The channel coefficient hl (t) is obtained by multiplying the output of L flat fading signal generators (implemented by the Jakes model) by the power of each tap.
本发明采用Alpha稳定分布模拟高强度的脉冲噪声：The present invention adopts Alpha stable distribution to simulate high-intensity impulse noise:
其中，α为特征指数，γ为分散系数，β为偏斜参数，υ为位置参数，sgn(t)为符号函数。Among them, α is the characteristic index, γ is the dispersion coefficient, β is the skew parameter, υ is the position parameter, and sgn(t) is the sign function.
本发明采用α＝1.5,β＝0,γ＝1,υ＝0的对称Alpha稳定分布噪声，定义混合信噪比：The present invention adopts symmetrical Alpha stable distribution noise with α=1.5, β=0, γ=1, υ=0, and defines the mixed signal-to-noise ratio:
式中，
至此，接收信号可表示为：So far, the received signal can be expressed as:
其中，x(t)为发送的调制信号，n(t)为Alpha稳定分布噪声。Among them, x(t) is the transmitted modulation signal, and n(t) is the Alpha stable distributed noise.
步骤2：对接收信号进行预处理，包括非线性变换、下变频、带通采样，包括：Step 2: Preprocess the received signal, including nonlinear transformation, down-conversion, and band-pass sampling, including:
1)为抑制Alpha噪声的影响，首先对接收信号进行非线性变换：1) In order to suppress the influence of Alpha noise, first perform nonlinear transformation on the received signal:
其中，Δ为正常数；Among them, Δ is a normal number;
2)接着为降低射频采样成本，将非线性变换后的信号下变频至140MHz；2) Next, in order to reduce the RF sampling cost, the nonlinearly transformed signal is down-converted to 140MHz;
3)再对下变频后的信号进行带通采样、低通滤波，将信号中心频率搬移至4MHz。3) Band-pass sampling and low-pass filtering are performed on the down-converted signal, and the center frequency of the signal is moved to 4MHz.
步骤3：针对Alpha稳定分布噪声不存在高阶统计量的问题，提取预处理后信号的广义循环均值和广义循环谱特征，并构造不同信噪比下的无人机集群通信波形特征矩阵；Step 3: Aiming at the problem that Alpha stable distribution noise does not have high-order statistics, extract the generalized cyclic mean and generalized cyclic spectrum features of the preprocessed signal, and construct the UAV swarm communication waveform feature matrix under different signal-to-noise ratios;
提取预处理后的无人机集群通信波形特征。Extract the preprocessed UAV swarm communication waveform features.
截获的无人机集群通信波形经预处理后，提取其循环平稳特征，包括广义循环均值特征和广义循环谱特征。After the intercepted UAV swarm communication waveform is preprocessed, its cyclostationary features, including generalized cyclic mean feature and generalized cyclic spectrum feature, are extracted.
预处理后的信号r'(t)的广义循环均值定义为：The generalized circular mean of the preprocessed signal r'(t) is defined as:
其中，ε＝k/T为循环频率，Mr'(t)为信号r'(t)的均值。where ε=k/T is the cycle frequency, andMr' (t) is the mean value of the signal r'(t).
信号r'(t)的广义循环谱密度
其中，
当ε＝0时，循环谱密度退化为功率谱密度。When ε=0, the cyclic spectral density degenerates into a power spectral density.
步骤3：建立无人机集群通信波形特征向量。本发明假设某混合信噪比下无人机集群中存在两个用户采用BPSK波形进行通信，提取两用户的广义循环平稳特征，则该样本的特征向量如下：Step 3: Establish the characteristic vector of the UAV swarm communication waveform. The present invention assumes that there are two users in the UAV cluster under a certain mixed signal-to-noise ratio using BPSK waveform to communicate, and the generalized cyclostationary characteristics of the two users are extracted, and the characteristic vector of the sample is as follows:
其中，
根据表4，选择广义循环均值和广义循环谱的离散峰值及其个数作为特征，则某混合信噪比下，每种波形截获N个样本构造的特征矩阵表示为：According to Table 4, the discrete peak value and number of generalized cyclic mean and generalized cyclic spectrum are selected as features, then under a certain mixed signal-to-noise ratio, the feature matrix constructed by intercepting N samples of each waveform is expressed as:
其中，
步骤4：将特征矩阵输入SAE神经网络进行训练测试，输出无人机集群通信波形类型，实现在Alpha噪声干扰、多径衰落与频移存在的复杂环境下的6种无人机集群通信波形的识别。Step 4: Input the feature matrix into the SAE neural network for training and testing, output the UAV swarm communication waveform type, and realize the 6 kinds of UAV swarm communication waveforms under the complex environment of Alpha noise interference, multipath fading and frequency shift. identify.
建立SAE神经网络作为无人机集群通信波形识别模型。The SAE neural network is established as the waveform recognition model of UAV swarm communication.
将提取的无人机通信波形特征，形成不同混合信噪比下的多维特征矩阵输入至SAE神经网络训练。The extracted UAV communication waveform features are formed into multi-dimensional feature matrices under different mixed signal-to-noise ratios and input to the SAE neural network training.
具有稀疏特性的SAE的前向传播方程如下：The forward propagation equation of SAE with sparse property is as follows:
Sin＝σ[U(u1,...,um)×Xt+a] (11)Sin =σ[U(u1 ,...,um )×Xt +a] (11)
Sout＝Oin＝σ(η1×Sin×Vt(v1,v2...vn)+b) (12)Sout =Oin =σ(η1 ×Sin ×Vt (v1 ,v2 ...vn )+b) (12)
Oout＝f(η2×Sout×Wt(w1,w2...wp)+c) (13)Oout =f(η2 ×Sout ×Wt (w1 ,w2 ...wp )+c) (13)
其中，Sin，Sout和Oout分别是隐藏层的输入值，隐藏层的输出值和最终的输出值，U、V、W分别为对应连接层的权重矩阵，Xt为所构造的特征向量。σ，f为激活函数，σ通常为tanh函数或sigmoid函数，f为Softmax函数，η是隐藏层的稀疏系数，a、b、c为各层偏置。Among them, Sin , Sout and Oout are the input value of the hidden layer, the output value of the hidden layer and the final output value, respectively, U, V, W are the weight matrix of the corresponding connection layer, Xt is the constructed feature vector. σ, f are the activation functions, σ is usually the tanh function or the sigmoid function, f is the Softmax function, η is the sparse coefficient of the hidden layer, and a, b, and c are the biases of each layer.
最后，将无人机集群波形特征输入训练好的神经网络，判别无人机集群的通信波形。Finally, input the waveform characteristics of the UAV swarm into the trained neural network to determine the communication waveform of the UAV swarm.
本发明已经完成了MATLAB软件仿真和验证。The present invention has completed MATLAB software simulation and verification.
下面给出本发明的具体实施步骤：The specific implementation steps of the present invention are given below:
(1)根据步骤1，采用TDL模型和Alpha稳定分布噪声建立无人机集群通信多径衰落信道。(1) According to
本发明中的信道参数设置基于第三代合作伙伴计划(The 3rd GenerationPartnership Project，3GPP)发布的0.5～100GHz的信道模型技术报告3GPP TR 901.38(V17.0.0)。The channel parameter setting in the present invention is based on the 0.5-100 GHz channel model technical report 3GPP TR 901.38 (V17.0.0) issued by The 3rd Generation Partnership Project (3GPP).
3GPP TR 38.901报告根据实测数据提供了5种TDL信道模型参数，其中，TDL-A、TDL-B和TDL-C适用于非视距(Non-line of Sight，NLOS)场景下，TDL-D和TDL-E则适用于视距(Line of Sight，LOS)场景。The 3GPP TR 38.901 report provides five TDL channel model parameters based on the measured data. Among them, TDL-A, TDL-B and TDL-C are suitable for Non-line of Sight (NLOS) scenarios, TDL-D and TDL-E is suitable for Line of Sight (LOS) scenarios.
该技术报告规定了TDL信道的衰落服从瑞利分布，并提供了每种模型参数下信道的多径数以及每条径对应的标准化时延、抽头功率，具体参数配置参考报告中表7.7.2-1至7.7.2-5。上述内容的时延均为标准化时延，即其均方根(Root Mean Square，RMS)时延扩展(Delay Spread，DS)为1。This technical report stipulates that the fading of the TDL channel obeys the Rayleigh distribution, and provides the multipath number of the channel under each model parameter and the normalized delay and tap power corresponding to each path. For specific parameter configuration, refer to Table 7.7.2 in the report. -1 to 7.7.2-5. The delays of the above content are all standardized delays, that is, their root mean square (Root Mean Square, RMS) delay spread (Delay Spread, DS) is 1.
3GPP TR 38.901中规定，可通过调节RMS时延扩展的方式获得指定场景下的抽头时延，不同场景下的RMS时延扩展参考报告中表7.7.3-1。It is stipulated in 3GPP TR 38.901 that the tap delay in a specified scenario can be obtained by adjusting the RMS delay extension. For the RMS delay extension in different scenarios, refer to Table 7.7.3-1 in the report.
同时，该报告规定了对于含LOS径的信道模型，如TDL-D和TDL-E，用户可通过调节K因子获得指定的Kdesired(dB)，3GPP TR 38.901提供了不同场景下的K因子均值和标准差，参考表7.5-6。At the same time, the report stipulates that for channel models with LOS paths, such as TDL-D and TDL-E, users can obtain the specified Kdesired (dB) by adjusting the K factor. 3GPP TR 38.901 provides the average K factor in different scenarios. and standard deviation, refer to Table 7.5-6.
在调节K因子后，该模型每个抽头对应的功率应作相应调节。时延和K因子的具体调节方式参考3GPP TR 38.901中7.7.2小节。After adjusting the K factor, the power corresponding to each tap of the model should be adjusted accordingly. Refer to Section 7.7.2 in 3GPP TR 38.901 for specific adjustment methods of delay and K factor.
(2)设置无人机集群通信场景参数。(2) Set the parameters of the UAV swarm communication scene.
考虑无人机集群内的多个用户利用正交资源(时域、频域、码域、空域)进行通信，如集群内的用户UA和UB进行通信，假定UA和UB在时/频/码/空域是正交的，侦收机可分别获得两用户发送的信号。Consider that multiple users in the UAV cluster communicate using orthogonal resources (time domain, frequency domain, code domain, air domain), such as usersUA andUB in the cluster to communicate, assuming thatUA andUB are at the same time /frequency/code/space are orthogonal, and the receiver can obtain the signals sent by the two users respectively.
本发明考虑集群内所有用户采用同种调制方式，可能采用的调制方式类型包括BPSK、QPSK、2FSK、4FSK、2ASK、MSK，集群内正在互相通信的用户为两架无人机。The present invention considers that all users in the cluster adopt the same modulation mode, the possible modulation modes include BPSK, QPSK, 2FSK, 4FSK, 2ASK, MSK, and the users communicating with each other in the cluster are two UAVs.
无人机的通信频率为5.8GHz时，其大气损耗极小，可认为电磁波在自由空间传播，其传播损耗LP(dB)定义为：When the communication frequency of the UAV is 5.8GHz, its atmospheric loss is extremely small, and it can be considered that electromagnetic waves propagate in free space, and its propagation loss LP (dB) is defined as:
其中，PT为发射功率，PR为接收功率，fc(GHz)为工作频率，d(km)为收发天线之间的距离。Among them, PT is the transmit power, PR is the receive power, fc (GHz) is the operating frequency, and d (km) is the distance between the transmitting and receiving antennas.
假定无人机集群内的任意两个用户，除了与侦收机的距离不同，其他参数均相同。设置无人机用户A距侦收机d1km，无人机用户B距侦收机d2km。It is assumed that any two users in the drone cluster have the same parameters except for the distance from the reconnaissance aircraft. Set the UAV user A is1 km away from the receiver d, and the UAV user B is2 km away from the receiver d.
定义集群内两无人机用户之间的路径损耗差ΔLp(dB)，表1是无人机集群中两用户之间的距离比和路径损耗差的关系。Define the path loss difference ΔLp (dB) between two UAV users in the cluster, and Table 1 shows the relationship between the distance ratio and the path loss difference between the two users in the UAV cluster.
表1集群两用户之间的距离比和路径损耗的关系Table 1 Relationship between distance ratio and path loss between two users in a cluster
集群内存在两无人机用户正在通信，通信频率5.8GHz，码元速率2M symbol/s。考虑其路径损耗差ΔLp分别为0dB和3dB的场景，对应地，两无人机用户相对于侦收机的距离比分别0.71和1，其他参数设置见表2。There are two UAV users in the cluster that are communicating, the communication frequency is 5.8GHz, and the symbol rate is 2M symbol/s. Considering the scenarios where the path loss differenceΔLp is 0dB and 3dB respectively, correspondingly, the distance ratios of the two UAV users to the receiver are 0.71 and 1, respectively. Other parameter settings are shown in Table 2.
本发明以2dB为步长，分别生成混合信噪比为-20～0dB的数据集。The present invention takes 2dB as a step, and respectively generates a data set with a mixed signal-to-noise ratio of -20-0dB.
根据表2设置无人机及信道参数，并生成每种调制方式下的无人机用户A和用户B的信号样本各1000。Set UAV and channel parameters according to Table 2, and generate 1000 signal samples of UAV user A and user B under each modulation mode.
表2无人机用户参数表Table 2 UAV user parameter table
无人机的飞行速度会影响最大多普勒频移的数值，从而影响无人机集群波形的识别准确率。The flying speed of the UAV will affect the value of the maximum Doppler frequency shift, thereby affecting the recognition accuracy of the UAV swarm waveform.
考虑两无人机用户的路径损耗差ΔLp为3dB的场景，采用TDL-A模型，无人机通信频率为5.8GHz，其他参数参考表3。Considering the scenario where the path loss differenceΔLp of two UAV users is 3dB, the TDL-A model is adopted, the UAV communication frequency is 5.8GHz, and other parameters refer to Table 3.
表3无人机用户参数表Table 3 UAV user parameter table
(3)根据步骤2预处理截获的无人机集群通信波形信号。(3) Preprocess the intercepted UAV swarm communication waveform signal according to
由于Alpha稳定分布噪声的二阶及以上各阶统计量趋于无穷，需对接收信号进行非线性变换，将噪声无穷的幅值限制在有限的区间内，以获得有效的信号特征。Since the second-order and above-order statistics of Alpha stable distributed noise tend to be infinite, it is necessary to perform nonlinear transformation on the received signal to limit the infinite amplitude of the noise to a limited interval to obtain effective signal characteristics.
经非线性变换后的信号为：The signal after nonlinear transformation is:
其中，Δ为正常数。where Δ is a positive constant.
信号经非线性处理后的特征称为广义循环均值和广义循环谱。The features of the signal after nonlinear processing are called generalized cyclic mean and generalized cyclic spectrum.
本发明设置无人机通信频率为5.8GHz，为降低射频采样成本，将信号下变频至140MHz，再进行带通采样，采样速率fs为16M sample/s。In the present invention, the communication frequency of the drone is set to 5.8GHz, in order to reduce the cost of radio frequency sampling, the signal is down-converted to 140MHz, and then band-pass sampling is performed, and the sampling rate fs is 16M sample/s.
经低通滤波后，信号中心频率搬移至4MHz。After low-pass filtering, the center frequency of the signal is shifted to 4MHz.
目前，有很多对信号载波频率fc和码元速率Rb等参数的估计方法，本发明中将无人机的通信频率和码元速率作为已知信息使用。At present, there are many estimation methods for parameters such as the signal carrier frequency fc and the symbol rate Rb . In the present invention, the communication frequency and symbol rate of the UAV are used as known information.
(4)根据步骤3提取无人机通信波形特征，构造无人机集群通信波形特征矩阵。(4) According to
表4为不同调制类型信号循环均值与循环谱特征。Table 4 shows the cyclic mean and cyclic spectrum characteristics of signals of different modulation types.
根据表4，选择神经网络的输入特征参数：对于不同调制信号的循环均值特征，以循环均值的离散峰值个数ρ1和平均循环均值ρ2作为特征参数；According to Table 4, the input characteristic parameters of the neural network are selected: for the cyclic mean characteristics of different modulation signals, the discrete peak number ρ1 of the cyclic mean and the average cyclic mean ρ2 are used as characteristic parameters;
对于不同调制信号的循环谱特征，以f＝0截面且ε＞0时的离散峰值ρ3～ρ7作为特征参数，即ε＝2fc，ε＝2fc±Rb和ε＝2fc±Rb/2处的循环谱密度峰值。For the cyclic spectrum characteristics of different modulated signals, the discrete peaks ρ3 to ρ7 when f=0 cross section and ε>0 are used as characteristic parameters, that is, ε=2fc , ε=2fc ±Rb and ε=2fc ± Cyclic spectral density peak at Rb /2.
表4不同调制信号循环特征Table 4 Cyclic characteristics of different modulation signals
图3至图6分别为部分调制信号的广义循环平稳特征。无人机通信信号经预处理后，中心频率搬移至4MHz。由图3和图4可知，2FSK信号的广义循环均值存在两个离散峰值，而2ASK信号的广义循环均值仅存在一个离散峰值；由图5和图6可知，BPSK信号的广义循环谱在f＝0截面的ε＝±(2fc±Rb)处存在离散峰值，而QPSK的f＝0截面不存在离散峰值。Figures 3 to 6 show the generalized cyclostationary characteristics of the partially modulated signal, respectively. After the UAV communication signal is preprocessed, the center frequency is moved to 4MHz. It can be seen from Figure 3 and Figure 4 that the generalized cyclic mean of the 2FSK signal has two discrete peaks, while the generalized cyclic mean of the 2ASK signal has only one discrete peak; it can be seen from Figures 5 and 6 that the generalized cyclic spectrum of the BPSK signal is There is a discrete peak at ε=±(2fc ±Rb ) in the 0 section, but there is no discrete peak in the f = 0 section of QPSK.
无人机集群中存在两个正在通信的用户，所构造的特征矩阵如下：There are two communicating users in the UAV swarm, and the constructed feature matrix is as follows:
其中，
图1和图2描述某个无人机集群内存在两个正在通信的用户UA和UB，通信双方采用同种调制方式，侦收机可分别获得两用户发送的信号，无人机用户A距侦收机d1 km，无人机用户B距侦收机d2km。Figures 1 and 2 describe that there are two communicating usersUA and UB in a UAVswarm . Both communication parties use the same modulation method. The receivers can obtain the signals sent by the two users respectively. A is1 km from the reconnaissance machine d, and the UAV user B is2 km from the reconnaissance machine d.
(5)根据步骤4，采用SAE神经网络识别无人机集群波形。(5) According to
按照步骤4所描述的神经网络识别无人机集群波形，本发明所采用的SAE神经网络结构参数如表5。SAE神经网络中训练集与测试集大小比例为7：3，依据上述内容所设置的不同场景和用户参数，给出不同信噪比下无人机集群通信波形识别的性能曲线。According to the neural network described in
表5 SAE神经网络结构参数配置Table 5 SAE neural network structure parameter configuration
图7和图8分别为TDL-A模型和TDL-D的路径增益随信道路径和样本时间的变化关系。其中TDL-A信道为NLOS场景，TDL-D信道含LOS径。对于TDL-A考虑的信道参数：最大多普勒频移1160Hz，RMS时延扩展100ns，样本点数8000sample，该模型共有23条可分辨路径。对于TDL-D考虑的信道参数：最大多普勒频移1160Hz，RMS时延扩展30ns，样本点数8000sample，K因子为9dB，该模型共有13条可分辨路径，其中第一条径为LOS径，其增益相对其他径较大。Figures 7 and 8 show the relationship between the path gain of the TDL-A model and the TDL-D model as a function of the channel path and sample time, respectively. The TDL-A channel is an NLOS scenario, and the TDL-D channel includes an LOS path. For the channel parameters considered by TDL-A: the maximum Doppler frequency shift is 1160Hz, the RMS delay extension is 100ns, and the number of sample points is 8000sample. There are 23 distinguishable paths in this model. For the channel parameters considered by TDL-D: the maximum Doppler frequency shift is 1160Hz, the RMS delay extension is 30ns, the number of sample points is 8000sample, and the K factor is 9dB. There are 13 distinguishable paths in this model, of which the first path is the LOS path. Its gain is larger than other diameters.
图9为基于SAE神经网络的信号调制方式预测模型。模型左侧输入无人机集群通信波形特征，右侧输出无人机集群通信波形类型，SAE神经网络的结构参数配置参考表5。Fig. 9 is a signal modulation method prediction model based on SAE neural network. The left side of the model inputs the UAV swarm communication waveform characteristics, and the right side outputs the UAV swarm communication waveform type. Refer to Table 5 for the configuration of the structural parameters of the SAE neural network.
图10和图11分别为TDL-A和TDL-D信道下信号调制识别性能曲线。信道参数及无人机用户参数依照表2。由图10和图11可知：从信道分析，由于TDL-D信道存在LOS径，其无人机集群通信波形识别准确率高于TDL-A信道；从方法上分析，通过SAE神经网络的识别无人机集群波形的准确率高于基于决策树的方法；从路径损耗差分析，路径损耗差越大，其无人机集群波形的识别准确率越低。Figure 10 and Figure 11 are the signal modulation identification performance curves under TDL-A and TDL-D channels, respectively. The channel parameters and UAV user parameters are in accordance with Table 2. It can be seen from Figure 10 and Figure 11: from the channel analysis, due to the existence of LOS path in the TDL-D channel, the recognition accuracy of the UAV swarm communication waveform is higher than that of the TDL-A channel; from the method analysis, the identification through the SAE neural network has no The accuracy rate of human-machine swarm waveform is higher than that of the method based on decision tree; from the analysis of path loss difference, the larger the path loss difference, the lower the recognition accuracy of UAV swarm waveform.
图12为DL-A信道下不同场景信号调制识别性能曲线。无人机的飞行速度会影响最大多普勒频移的数值，从而影响无人机集群波形的识别准确率。考虑两无人机用户的路径损耗差ΔLp为3dB的场景，采用TDL-A模型，无人机通信频率5.8GHz，其他参数参考表5。由图12可知：随着无人机飞行速度和多普勒频移的增大，无人机集群通信波形识别的准确率下降。FIG. 12 is the performance curve of signal modulation recognition in different scenarios under the DL-A channel. The flying speed of the UAV will affect the value of the maximum Doppler frequency shift, thereby affecting the recognition accuracy of the UAV swarm waveform. Considering the scenario where the path loss differenceΔLp of two UAV users is 3dB, the TDL-A model is adopted, the UAV communication frequency is 5.8GHz, and other parameters refer to Table 5. It can be seen from Figure 12 that with the increase of UAV flight speed and Doppler frequency shift, the accuracy of UAV swarm communication waveform recognition decreases.
以上仅是本发明的优选实施方式，本发明的保护范围并不仅局限于上述实施例，凡属于本发明思路下的技术方案均属于本发明的保护范围。应当指出，对于本技术领域的普通技术人员来说，在不脱离本发明原理前提下的若干改进和润饰，应视为本发明的保护范围。The above are only preferred embodiments of the present invention, and the protection scope of the present invention is not limited to the above-mentioned embodiments, and all technical solutions that belong to the idea of the present invention belong to the protection scope of the present invention. It should be pointed out that for those skilled in the art, some improvements and modifications without departing from the principle of the present invention should be regarded as the protection scope of the present invention.
Claims (7)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210858726.3ACN115225441B (en) | 2022-07-21 | 2022-07-21 | A method for identifying waveforms of UAV cluster communication in complex environments |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210858726.3ACN115225441B (en) | 2022-07-21 | 2022-07-21 | A method for identifying waveforms of UAV cluster communication in complex environments |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN115225441Atrue CN115225441A (en) | 2022-10-21 |
| CN115225441B CN115225441B (en) | 2024-09-03 |
Family
ID=83613776
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210858726.3AActiveCN115225441B (en) | 2022-07-21 | 2022-07-21 | A method for identifying waveforms of UAV cluster communication in complex environments |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN115225441B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115902804A (en)* | 2022-11-07 | 2023-04-04 | 南京航空航天大学 | Method and system for identifying type of unmanned aerial vehicle swarm |
| CN118972991A (en)* | 2024-08-01 | 2024-11-15 | 中国人民解放军空军工程大学 | A design method for adaptive rate transmission waveform of UAV measurement and control system |
Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160069994A1 (en)* | 2014-09-09 | 2016-03-10 | University Of Kansas | Sense-and-avoid systems and methods for unmanned aerial vehicles |
| CN109324244A (en)* | 2018-09-26 | 2019-02-12 | 北京航空航天大学 | A battlefield electromagnetic environment generation method for UAV survivability assessment |
| US20200336344A1 (en)* | 2019-04-16 | 2020-10-22 | Motorola Solutions, Inc. | Systems and methods for modulation classification of baseband signals using attention-based learned filters |
| CN113315727A (en)* | 2021-03-29 | 2021-08-27 | 中山大学 | Digital communication signal modulation identification method based on preprocessing noise reduction |
| CN114609597A (en)* | 2022-01-26 | 2022-06-10 | 南京航空航天大学 | Integrated radar waveform design method for interference and invasion for UAV swarm detective fusion |
- 2022
- 2022-07-21CNCN202210858726.3Apatent/CN115225441B/enactiveActive
Patent Citations (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20160069994A1 (en)* | 2014-09-09 | 2016-03-10 | University Of Kansas | Sense-and-avoid systems and methods for unmanned aerial vehicles |
| CN109324244A (en)* | 2018-09-26 | 2019-02-12 | 北京航空航天大学 | A battlefield electromagnetic environment generation method for UAV survivability assessment |
| US20200336344A1 (en)* | 2019-04-16 | 2020-10-22 | Motorola Solutions, Inc. | Systems and methods for modulation classification of baseband signals using attention-based learned filters |
| CN113315727A (en)* | 2021-03-29 | 2021-08-27 | 中山大学 | Digital communication signal modulation identification method based on preprocessing noise reduction |
| CN114609597A (en)* | 2022-01-26 | 2022-06-10 | 南京航空航天大学 | Integrated radar waveform design method for interference and invasion for UAV swarm detective fusion |
Non-Patent Citations (1)
| Title |
|---|
| 孔豫京;黄焱;马金全: "Alpha稳定分布噪声背景下衰落信号的调制识别方法", 信号处理, no. 008, 31 December 2015 (2015-12-31)* |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115902804A (en)* | 2022-11-07 | 2023-04-04 | 南京航空航天大学 | Method and system for identifying type of unmanned aerial vehicle swarm |
| CN115902804B (en)* | 2022-11-07 | 2024-01-05 | 南京航空航天大学 | Unmanned aerial vehicle cluster type identification method and system |
| CN118972991A (en)* | 2024-08-01 | 2024-11-15 | 中国人民解放军空军工程大学 | A design method for adaptive rate transmission waveform of UAV measurement and control system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN115225441B (en) | 2024-09-03 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN115225441B (en) | A method for identifying waveforms of UAV cluster communication in complex environments | |
| CN112787966B (en) | Method for demodulating antagonistic network signal based on end-to-end cascade generation | |
| CN101895494B (en) | Automatic Identification Method of Digital Modulation Mode Based on Stochastic Resonance Preprocessing | |
| CN111464469A (en) | Recognition method of mixed digital modulation mode based on neural network | |
| CN112215335B (en) | System detection method based on deep learning | |
| CN109547376A (en) | A kind of Modulation Identification method of the gauss hybrid models based on modulating characteristic | |
| CN112949846A (en) | Method for constructing generated deception jamming signal suitable for direct sequence spread spectrum system | |
| CN112232120B (en) | Radar radiation source signal classification system and method based on software radio | |
| Lin et al. | Deep unfolding network for target parameter estimation in OTFS-based ISAC systems | |
| CN107465472B (en) | A Multipath Delay Estimation Method Based on Path Synthesis | |
| CN118468125A (en) | A radar signal intelligent detection model construction method based on Channel-DeepIQ, radar signal intelligent detection method, computer device and storage medium | |
| CN116436491B (en) | Frequency hopping detection result supplementing method based on distributed multi-station monitoring equipment | |
| CN115499278B (en) | MIMO signal modulation identification method based on lightweight neural network | |
| Liu et al. | Specific emitter identification method based on deep ensemble learning | |
| Ping et al. | UAV swarm automatic modulation recognition with multiple interferences | |
| CN111796263B (en) | A distributed ranging method for low-altitude targets based on 5G base station signals | |
| Hu et al. | Signal classification in real-time based on SDR using convolutional neural network | |
| Zhai et al. | Waveform recognition of UAV swarm communication in multipath channel | |
| Yang et al. | SNR estimation of UAV control signal based on convolutional neural network | |
| Yang et al. | Deep-learning-based bouncing-order prediction for propagation channel characterization using graph modeling | |
| CN120090730A (en) | A wireless channel scene recognition method, system, device and medium | |
| Nerandžić et al. | Impact of AWGN estimation on classification performance of AMC algorithms based on higher order cumulants | |
| CN115664563B (en) | Passive cooperative spectrum sensing method based on geometrical symmetry of energy features | |
| Xiang et al. | Automatic modulation recognition based on SST and deep learning model | |
| Liu et al. | MIMO Signals Classification with Cauchy Noise in Space-air-ground Integrated Networks |
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 |