A wireless signal recognition system training sample simulation method

Abstract

The application discloses a wireless signal recognition system training sample simulation method, which comprises the following steps: frame element-based baseband signal modeling, random interference variable-based intermediate frequency signal generation, and sample generation based on superposition of multiple intermediate frequency signals. The application adopts a frame element-based modular method to construct the baseband signal, uses uniformly distributed interference variables to give the sample sufficient diversity, and can realize superposition of the same type and different types of intermediate frequency signals. The application is based on a standard sample simulation process, can generate any number of simulation signals, and meets the requirements of the training set of the recognition system for sample diversity and the fact that a single sample contains multiple signals. The application can provide direct data support for intelligent wireless signal processing.

Description

TECHNICAL FIELD

[0001] The application belongs to the field of data engineering, and specifically designs a wireless signal recognition system training sample simulation method. BACKGROUND

[0002] The signal recognition system based on deep learning needs a large number of high-quality training samples, but due to the limitations of real-time acquisition, it is impossible to provide a sufficient number and quality of samples. On the other hand, wireless signals are naturally generated by computers and transmitted through intermediate frequency, radio frequency devices and antennas, and then propagated through channels, so the actual collected signal and the computer simulated signal have homology, so it is necessary and feasible to use computer simulation method to realize wireless signal sample generation.

[0003] China Telecom Co., Ltd. discloses a wireless signal simulation method and device, electronic equipment and medium, based on the signal strength of each position point in the test data of the antenna to be tested, generates an isostatic line and an isostatic surface, and simulates a wireless signal; The Chinese People's Liberation Army Bengbu Tank Academy discloses a wireless radio station analog communication electromagnetic environment simulation device, which has a noise and radiation simulation signal generation module, a signal synthesis module, a user input and display module, and a control module, which can realize that the voice signal output by the actual radio station is processed through the analog electromagnetic environment simulation, and then output to the audio terminal; The Chinese People's Liberation Army National University of Defense Technology discloses a wireless signal test simulation system, which comprises: an arbitrary waveform module for configuring a signal generator virtual instrument to generate different types of simulation signals and display the time domain waveform of the simulation signals; a closed field path module for configuring various types of radio frequency connectors, radio frequency connecting lines and attenuators to simulate a closed field path; a spectrum analysis module for configuring a spectrum analyzer virtual instrument to perform spectrum analysis on the simulation signals and display the results of the spectrum analysis; The Chinese People's Liberation Army discloses a general information processing platform suitable for wireless simulation equipment of various frequency bands, which adopts a hardware module to build a public platform, uses software instead of hardware, focuses on using software instead of hardware to simulate the functions and operations of various types of wireless communication equipment in actual application, and uses Ethernet technology to realize system upgrade of the wireless simulation equipment connected in the network on a terminal device.

[0004] However, the simulation signal generated by the above wireless signal simulation system cannot directly support the training of the signal recognition system based on deep learning. On the one hand, the lack of randomness of the variables in the simulation signal leads to the inability of the sample diversity to meet the training requirements, and on the other hand, the simulation signal is a single intermediate frequency frequency sampling signal, which does not match the mode of the received signal (wideband sampling signal) of the recognition system. SUMMARY

[0005] To address the problems existing in the prior art, this invention provides a simulation method for training samples in a wireless signal recognition system, specifically including the following steps:

[0006] Step 1: Baseband signal modeling based on frame elements;

[0007] Frame elements include a start flag, address information, data fields, time slot structure, and end flag;

[0008] The baseband signal of a frame element is assembled based on the frame element;

[0009] Step 1: Determine the baseband signal frame type. Baseband signal frame types are divided into continuous signals, periodic frame signals, and aperiodic frame signals. Determine the frame elements included, including frame header, synchronization header, special code, time slot interval, data segment, and frame tail.

[0010] Step 2: Define frame elements and determine parameters such as modulation type, symbol rate, coding type, and information sequence construction;

[0011] Step 3: Determine the timing of frame elements, assemble frame elements, and construct the basic frame structure;

[0012] Step 4: If the baseband signal is a continuous signal, then use one frame structure to construct the baseband signal; if the baseband signal is a periodic frame signal, then use three or more frame structures to construct the baseband signal, with a fixed time slot between frames; if the baseband signal is a non-periodic frame signal, then use three or more frame structures to construct the baseband signal, with a random time slot between frames.

[0013] Step 2: Generation of intermediate frequency signal based on random interference variables;

[0014] The parameters involved in intermediate frequency (IF) signal simulation include signal occurrence time, signal duration, average signal power, noise standard deviation, signal center frequency, and signal carrier phase. Since IF signal parameters are independent of signal type, these parameters are used to identify interference in the system. For an observation time of T... d For intermediate frequency signals, let the following interference variables follow a uniform distribution:

[0015] (1) Signal occurrence time T start :T start ~U[-T d ,T d ]

[0016] (2) Signal duration T c :T c ~U[0,2T d ]

[0017] (3) Average signal power P: P ~ U[P min ,Pmax ]

[0018] (4) Noise standard deviation σ N : σ N ~ U [0, σ Nmax ]

[0019] (5) Signal center frequency f c : f c ~ U [0, f s / 2)

[0020] (6) Signal carrier phase where ~ U [·] represents that the random variable obeys uniform distribution, P min and P max represent the maximum and minimum values of the possible average power of the simulation signal, σ Nmax represents the maximum value of the possible noise standard deviation, f s is the signal sampling frequency; without considering the addition of channel and interference factors, and only considering the noise type as Gaussian white noise;

[0021] The generation of the intermediate frequency signal is a random process, and let the baseband signal generated in step 1 be s(t), and the specific process is as follows:

[0022] Step 1: Randomization of signal occurrence time and duration;

[0023] In order to ensure the randomization of the intermediate frequency signal occurrence time and duration, two times of interception are performed:

[0024] First time: randomly intercepting a signal segment s(t|T c ) with a length of T c , and zero-padded before and after to construct an intermediate process signal sequence s1(t):

[0025]

[0026] Second time: randomly intercepting s1(t) to form a signal sequence s2(t) with a random occurrence time and duration of T d ;

[0027] Step 2: Randomization of signal average power;

[0028] After normalizing s2(t) by dividing the standard deviation and multiplying by the amplitude gain, a random power signal s3(t) is obtained:

[0029]

[0030] where std() represents the standard deviation of the signal;

[0031] Step 3: Randomization of noise variance;

[0032] Add random standard deviation Gaussian white noise to the random power signal s3(t) to obtain the noisy intermediate frequency signal s4(t):

[0033] s4(t) = s3(t) + σ N *randn

[0034] In the formula, randn represents a random variable obeying normal distribution with value range [0, 1], mean value 0.5, and variance 1;

[0035] Step 4: Randomization of signal center frequency and carrier phase;

[0036] Set the random center frequency and carrier phase of the noisy intermediate frequency signal s4(t) to become the intermediate frequency signal s5(t):

[0037]

[0038] The above process is described using the joint probability density function:

[0039]

[0040] Since the above random variables are not related, according to the uniform distribution of the above variables, we have:

[0041]

[0042] In the formula, f(T start ), f(T c ), f(P), f(σ N ), f(f c ) represent the probability density functions of variables T start , T c , P, σ N , f c , respectively;

[0043] Step 3: Sample generation based on multiple intermediate frequency signal superposition;

[0044] Define the intermediate frequency signal generated by one round of step 2 operation as Where I represents the number of intermediate frequency signals within the sampling bandwidth; multiple intermediate frequency signals are superimposed to obtain the simulation sample S(t):

[0045]

[0046] ​Judge whether the index i of the signal reaches the set signal number I, if yes, execute sample output, output S(t); if no, judge whether the next signal and the current signal are of the same type; if yes, jump to execute step 2, if no, jump to step 1.

[0047] In one embodiment of the present application, in step 1, for the Link-4A baseband signal in the CAINS mode, the following steps are specifically included:

[0048] Step 1: Determine that the Link-4A baseband signal in the CAINS mode is a periodic frame signal, which includes five frame elements of synchronization pulse train, guard interval, start bit, data bit and unkeyed signal.

[0049] Step 2: The Link-4A frame element definition in the CAINS mode is as follows: the modulation type is 2FSK modulation; the frequency interval is 40 kHz, symbol 0 is negative frequency offset-20 kHz, and symbol 1 is positive frequency offset 20 kHz; the symbol rate is 10 kBd, the synchronization pulse train; the 5 kBd guard interval, start bit, data bit and unkeyed signal, the encoding type is even check, and the information sequence structure is shown in Table 1.

[0050] Table 1 Link-4A (CAINS mode) frame element information sequence structure

[0051]

[0052]

[0053] Step 3: Determine that the Link-4A basic frame structure in the CAINS mode is arranged in the order of synchronization pulse train, guard interval, start bit, data bit and unkeyed signal.

[0054] Step 4: The Link-4A baseband signal in the CAINS mode is a periodic frame signal, and more than three frame structures are used to construct the baseband signal, and the frame interval is a fixed time slot of 2 ms.

[0055] The present application adopts computer simulation to realize the generation of data sets in the field of deep learning, which can make up for the problems of small number and low quality of actual signal samples.

[0056] The present application has the following advantages:

[0057] (1) The baseband signal generation based on frame elements conforms to the standard computer simulation process, giving more freedom and expandability to signal generation;

[0058] (2) The randomness of the interference variable in the sample is introduced from the perspective of statistical distribution, which greatly expands the diversification of the sample.

[0059] (3) The sample is generated by superimposing multiple intermediate frequency signals, conforms to the signal receiving mode of an actual identification system, and supports the training of a signal identification system in multiple signal identification modes based on deep learning. BRIEF DESCRIPTION OF DRAWINGS

[0060] Figure 1 A frame element-based baseband signal modeling process is shown;

[0061] Figure 2 The construction of a Link-4A (CAINS mode) basic frame structure is shown;

[0062] Figure 3 The construction of a Link-4A (CAINS mode) baseband signal is shown;

[0063] Figure 4 An intermediate frequency signal implementation process considering the statistical distribution of interference variables is shown;

[0064] Figure 5 A sample example of superimposing multiple intermediate frequency signals is shown. DETAILED DESCRIPTION

[0065] The present application is described in detail below with reference to the accompanying drawings.

[0066] The present application proposes a wireless signal identification system training sample simulation method, specifically including the following steps:

[0067] Step 1: frame element-based baseband signal modeling;

[0068] Frame elements refer to the smallest unit of baseband signals with relatively independent functions and structures, including start flags, address information, data fields, time slot structures, end flags, etc.

[0069] The frame element-based baseband signal modeling method adopts modular design, can flexibly set the combination mode and order of frame elements, presents the combination of frame elements in a form convenient for interaction and customization, and thus constructs any required baseband signal.

[0070] The baseband signal is assembled based on frame elements, and the process is shown in Figure 1 The following takes the generation of a Link-4A (CAINS mode) baseband signal as an example to illustrate the specific process:

[0071] Step 1: Determine the baseband signal frame type, which can usually be divided into continuous signals, periodic frame signals, and aperiodic frame signals, and determine the frame element composition contained, including frame headers, synchronization headers, special codes, time slot intervals, data segments, and frame tails;

[0072] For example, the Link-4A (CAINS mode) baseband signal is a periodic frame signal, which contains five frame elements: synchronization pulse train, guard interval, start bit, data bit, and unkeyed signal.

[0073] Step 2: Define frame elements, determine modulation type, symbol rate, encoding type, information sequence construction, and other parameters.

[0074] For example, the Link-4A (CAINS mode) frame element definition is as follows: the modulation type is 2FSK modulation, the frequency interval is 40 kHz (symbol 0 is negative frequency offset -20 kHz, symbol 1 is positive frequency offset 20 kHz), the symbol rate is 10 kBd (synchronization pulse train) and 5 kBd (guard interval, start bit, data bit, and unkeyed signal), the encoding type is even parity, and the information sequence construction is shown in Table 1.

[0075] Table 1 Link-4A (CAINS mode) frame element information sequence construction

[0076] Information type Number of bits (time slot) Code pattern Total length 78 bits (14 ms) Synchronization burst 16 bits (1.6 ms) 0101010101010101 Guard interval 4 bits (0.8 ms) 0000 Start bit 1 bit (0.2 ms) 1 Data bits 56 bits (11.2 ms) Random code pattern Non-keying signal 1 bit (0.2 ms) 0

[0077] Step 3: Determine the timing of the frame elements, assemble the frame elements, and construct the basic frame structure.

[0078] For example, the Link-4A (CAINS mode) basic frame structure is arranged in the order of synchronization pulse train, guard interval, start bit, data bit, and unkeyed signal, as shown in Figure 2 .

[0079] Step 4: If the baseband signal is a continuous signal, use one frame structure to construct the baseband signal; if the baseband signal is a periodic frame signal, use three or more frame structures to construct the baseband signal, with a fixed time interval between frames; if the baseband signal is a non-periodic frame signal, use three or more frame structures to construct the baseband signal, with a random time interval between frames.

[0080] For example, the Link-4A (CAINS mode) baseband signal is a periodic frame signal, which uses three or more frame structures to construct the baseband signal, with a fixed time interval of 2 ms between frames, as shown in Figure 3 .

[0081] Step 2: Generation of intermediate frequency signal based on random disturbance variable

[0082] The baseband signal parameters such as modulation type, symbol rate, frame interval are the sensitive parameters (parameters related to signal type) of the signal recognition system, and the parameters involved in the intermediate frequency signal simulation include signal occurrence time, signal duration, signal average power, noise standard deviation, signal center frequency and signal carrier phase. Since the intermediate frequency signal parameters are not related to the signal type, the above parameters are interference parameters in the recognition system. Therefore, the parameters in the baseband signal modeling process are set to a fixed value, and in the intermediate frequency signal implementation, the interference factors are considered as random variables conforming to a certain statistical distribution, and the recognition system learns this distribution from the samples during training and is not sensitive to the samples meeting this distribution. For the intermediate frequency signal with an observation time of T d , the following interference variables can be subjected to uniform distribution:

[0083] (1) Signal occurrence time T start : T start ~ U [-T d , T d ]

[0084] (2) Signal duration T c : T c ~ U [0, 2T d ]

[0085] (3) Signal average power P: P ~ U [P min , P max ]

[0086] (4) Noise standard deviation σ N : σ N ~ U [0, σ Nmax ]

[0087] (5) Signal center frequency f c : f c ~ U [0, f s / 2)

[0088] (6) Signal carrier phase where ~ U [·] represents that the random variable is subjected to uniform distribution, P min and P max represent the maximum and minimum values of the average power of the simulated signal, σ Nmax represents the maximum value of the noise standard deviation, and f s is the signal sampling frequency. It should be noted that since the channel and interference conditions are relatively complex, for the purpose of simplifying the description, the present application does not consider adding channel and interference factors, and only considers the noise type as Gaussian white noise.

[0089] Thus, the generation of the intermediate frequency signal is defined as a random process, which is realized by adding a uniformly distributed interference variable to the baseband signal, and the baseband signal generated in step 1 is denoted as s(t). The specific process is as follows:

[0090] Step 1: Randomization of signal occurrence time and duration;

[0091] To ensure the randomization of the intermediate frequency signal occurrence time and duration, two times of interception are needed:

[0092] First time: Randomly intercepting a signal segment s(t|T c ) with a length of T c , and zero-padded before and after to construct the intermediate signal sequence s1(t):

[0093]

[0094] Second time: Randomly intercepting s1(t) to form a signal sequence s2(t) with a random occurrence time and duration of T d .

[0095] Step 2: Randomization of signal average power;

[0096] The signal samples in the training set should have random average power, i.e., after normalizing s2(t) by dividing the standard deviation and multiplying by the amplitude gain, a random power signal s3(t) is obtained:

[0097]

[0098] where std() represents the standard deviation of the signal.

[0099] Step 3: Randomization of noise variance;

[0100] Add a random standard deviation Gaussian white noise to the random power signal s3(t) to obtain a noisy intermediate frequency signal s4(t):

[0101] s4(t) = s3(t) + σ N *randn

[0102] where randn represents a random variable following a normal distribution with a value range of [0, 1], a mean of 0.5, and a variance of 1.

[0103] Step 4: Randomization of signal center frequency and carrier phase;

[0104] The noisy intermediate frequency signal s4(t) still belongs to the baseband signal category, and setting a random center frequency and carrier phase for it makes it become an intermediate frequency signal s5(t):

[0105]

[0106] The above process can be described using a joint probability density function:

[0107]

[0108] Since the above random variables are uncorrelated, according to the above variables obeying uniform distribution, we have:

[0109]

[0110] In the formula, f(T start ), f(T c ), f(P), f(σ N ), f(f c ) respectively represent the probability density functions of variables T start , T c , P, σ N , f c , .

[0111] Step 3: generating based on the samples of multiple intermediate frequency signal superpositions;

[0112] It is worth noting that the intermediate frequency signal s5(t) is a sampled signal at a single frequency. However, the signal receiving bandwidth in the recognition system is usually (much) larger than the signal bandwidth, and there is a high probability that more than one intermediate frequency signal exists in the received signal, as shown in Figure 4 . In particular, the trunking communication signal usually has multiple intermediate frequency signals in the sampling spectrum, so the sample generation needs to consider multiple intermediate frequency signal superpositions.

[0113] Define the operation of step 2 to generate one intermediate frequency signal in one round as Where I represents the number of intermediate frequency signals in the sampling bandwidth. Multiple intermediate frequency signal superpositions obtain the simulation sample S(t):

[0114]

[0115] Next, it is determined whether the index i of the signal reaches the set signal number I, if yes, the sample output is executed, and S(t) is output; if not, it is determined whether the next signal and the current signal are of the same type; if yes, step 2 is executed, if not, step 1 is executed, and the flow chart is as shown in Figure 5 .​

Claims

1. A simulation method for training samples in a wireless signal detection and identification system, characterized in that, Specifically, the following steps are included: Step 1: Baseband signal modeling based on frame elements; Frame elements include a start flag, address information, data fields, time slot structure, and end flag; The baseband signal of a frame element is assembled based on the frame element; Step 1: Determine the baseband signal frame type. Baseband signal frame types are divided into continuous signals, periodic frame signals, and aperiodic frame signals. Determine the frame elements included, including frame header, synchronization header, special code, time slot interval, data segment, and frame tail. Step 2: Define frame elements, determine modulation type, symbol rate, coding type, and information sequence construction parameters; Step 3: Determine the timing of frame elements, assemble frame elements, and construct the basic frame structure; Step 4: If the baseband signal is a continuous signal, then use one frame structure to construct the baseband signal; if the baseband signal is a periodic frame signal, then use three or more frame structures to construct the baseband signal, with a fixed time slot between frames; if the baseband signal is a non-periodic frame signal, then use three or more frame structures to construct the baseband signal, with a random time slot between frames. Step 2: Generation of intermediate frequency signal based on random interference variables; The parameters involved in intermediate frequency (IF) signal simulation include signal occurrence time, signal duration, average signal power, noise standard deviation, signal center frequency, and signal carrier phase. Since IF signal parameters are independent of signal type, these parameters are used to identify interference in the system. For an observation time of T... d For intermediate frequency signals, let the following interference variables follow a uniform distribution: (1) Signal occurrence time T start :T start ~U[-T d ,T d ] (2) Signal duration T c :T c ~U[0,2T d ] (3) Average signal power P: P ~ U[P min ,P max ] (4) Noise standard deviation σ N σ N ~U[0,σ Nmax ] (5) Signal center frequency f c :f c ~U[0,f s / 2) (6) Signal carrier phase Where ~U[·] represents that the random variable follows a uniform distribution, P min and P max σ represents the maximum and minimum values ​​of the average power of the simulated signal. Nmax f represents the maximum value of the noise standard deviation. s The signal sampling frequency is used; channel and interference factors are not considered, and only Gaussian white noise is considered. The generation of the intermediate frequency signal is a random process. Let the baseband signal generated in step 1 be s(t). The specific process is as follows: Step 1: Randomize the signal occurrence time and duration; To ensure the randomization of the occurrence time and duration of the intermediate frequency signal, two truncations are performed: First step: Randomly select a length of T. c signal segment s(t|T) c The signal sequence s1(t) for the intermediate process is constructed by padding with zeros at the beginning and end: Second: Randomly select s1(t) to form an observation time of T. d The randomized signal sequence s2(t) with the signal occurrence time and duration; Step 2: Randomization of average signal power; Normalizing s2(t) by dividing it by the standard deviation and then multiplying it by the amplitude gain yields the random power signal s3(t): Where std() represents the standard deviation of the signal; Step 3: Randomization of noise variance; By superimposing Gaussian white noise with random standard deviation onto a random power signal s3(t), we obtain a noisy intermediate frequency signal s4(t): s4(t) = s3(t) + σ N *randn In the formula, randn represents a random variable that takes values ​​in the range [0,1], has a mean of 0.5, and a variance of 1, and follows a normal distribution. Step 4: Randomization of signal center frequency and carrier phase; To transform a noisy intermediate frequency signal s4(t) into an intermediate frequency signal s5(t), a random center frequency and carrier phase are set: The above process is described using a joint probability density function: Since the above random variables are uncorrelated, and based on the fact that these variables follow a uniform distribution, we have: In the formula, f(T) start ), f(T) c f(P), f(σ) N ), f(f c ), They represent variables T respectively. start T c , P, σ N f c , The probability density function; Step 3: Sample generation based on the superposition of multiple intermediate frequency signals; Define the intermediate frequency signal generated in one round of operation step 2 as: Where I represents the number of intermediate frequency signals within the sampling bandwidth; the simulation sample S(t) is obtained by superimposing multiple intermediate frequency signals: Determine whether the signal index i has reached the set number of signals I. If yes, execute sample output and output S(t). If no, determine whether the next signal is of the same type as the current signal. If yes, jump to step 2. If no, jump to step 1.

2. The simulation method for training samples in a wireless signal detection and identification system as described in claim 1, characterized in that, Step 1, for the Link-4A baseband signal in CAINS mode, specifically includes the following steps: Step 1: Determine that the Link-4A baseband signal in CAINS mode is a periodic frame signal, which contains five frame elements: synchronization pulse train, guard interval, start bit, data bit, and no keying signal. Step 2: In CAINS mode, the Link-4A frame elements are defined as follows: the modulation type is 2FSK modulation; the frequency interval is 40kHz, symbol 0 is negative frequency offset -20kHz, symbol 1 is positive frequency offset 20kHz; the symbol rate is 10kBd, the synchronization pulse train; 5kBd guard interval, start bit, data bits, no keying signal, and the coding type is even parity. Step 3: Determine the basic frame structure of Link-4A in CAINS mode, arranging it in the order of synchronization pulse train, guard interval, start bit, data bit and non-keying signal; Step 4: In CAINS mode, the Link-4A baseband signal is a periodic frame signal. The baseband signal is constructed using more than 3 frame structures, with a fixed time slot of 2ms between frames.

Images