A high-sensitivity adaptive detection method based on FH-TDMA signals

Abstract

The application discloses a high-sensitivity adaptive detection method based on FH-TDMA signals and relates to the technical field of wireless communication signal detection and spectrum monitoring, and solves the problems that the real-time performance is low, the hardware cost is too high and the method is not applicable to practical application in the prior art FH-TDMA signal detection and identification; the method comprises the following steps: performing initial frequency conversion filtering on an intermediate frequency signal, obtaining a frequency hopping pattern through ordered statistical constant false alarm frequency detection, performing secondary frequency conversion to obtain a baseband complex signal, detecting a frame starting moment through a frame synchronization header peak value ratio, extracting a first envelope signal and an amplitude and a frequency through envelope preliminary detection, and extracting a second envelope signal and an amplitude and a frequency through envelope complex detection and outputting; and the method realizes simple, high-precision, small resource cost and high real-time FH-TDMA signal detection.

Description

Technical Field

[0001] This invention relates to the field of wireless communication signal detection and spectrum monitoring technology, and in particular to a high-sensitivity adaptive detection method based on FH-TDMA signals. Background Technology

[0002] FH-TDMA is a wireless data broadcasting network that uses TDMA access. It divides user transmission time into multiple time slots, allowing users to transmit information within their allocated slots. Inter-pulse frequency hopping and other anti-interference measures are employed within these slots. FH-TDMA signals have a unique frame synchronization structure; each burst frame header typically includes a frame synchronization field to aid in frame synchronization. Due to its excellent security and anti-interference capabilities, it is widely used in radar, satellite communications, and other fields. Furthermore, FH-TDMA signals are frequently used as inter-network communication signals between UAV swarms. Therefore, the detection and identification of FH-TDMA signals is of significant practical importance.

[0003] The physical layer design of FH-TDMA systems presents multiple technical challenges for signal detection. Firstly, in practical applications, a wider frequency hopping range and more dispersed frequency points in an FH-TDMA signal significantly improve its anti-interference capabilities, but also necessitate a corresponding increase in the detection frequency band. Secondly, while a higher frequency hopping rate makes an FH-TDMA signal less susceptible to monitoring and interference, this also places higher demands on real-time detection. Current detection methods, such as the frequency-doubling correlation method, require receiving and accumulating sufficient pulses to achieve the desired detection result, resulting in low real-time performance. Another example is the dual-sliding-window detection algorithm based on multiple parallel channels, which performs parallel reception and detection across all frequency hopping points of the FH-TDMA signal. This incurs excessive hardware overhead when dealing with large frequency hopping bandwidths, making it unsuitable for practical applications. Summary of the Invention

[0004] This invention provides a high-sensitivity adaptive detection method based on FH-TDMA signals, which solves the problems of low real-time performance and high hardware overhead in existing technologies for FH-TDMA signal detection and recognition, making them unsuitable for practical applications. It achieves simple, high-precision, low-resource-consumption, and high-real-time FH-TDMA signal detection.

[0005] This invention provides a high-sensitivity adaptive detection method based on FH-TDMA signals, the method comprising: The input intermediate frequency signal to be detected is acquired, and the intermediate frequency signal to be detected is initially frequency-converted and filtered to obtain a preprocessed intermediate frequency signal; A frequency hopping frequency point carried by each pulse in the preprocessed intermediate frequency signal is detected in real time using an ordered statistical constant false alarm rate (CFAR) frequency detection algorithm to obtain the frequency hopping pattern corresponding to the current time slot. The preprocessed intermediate frequency signal is down-converted twice according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point of the current time slot; The frame start time of the current time slot is obtained by performing cross-correlation-autocorrelation peak ratio detection based on the frame synchronization header field of the baseband complex signal; Using the envelope detection module, the first envelope signal corresponding to each pulse is initially extracted based on the frame start time of the current time slot, and the first amplitude value corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the first envelope signal. Based on the first amplitude and the first frequency value, the preprocessed intermediate frequency signal is subjected to envelope re-examination, and the second envelope signal corresponding to each pulse is extracted for the second time. At the falling edge of the second envelope signal, the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are captured, and the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are output.

[0006] One or more technical solutions provided in this invention have at least the following technical effects or advantages: This invention acquires the input intermediate frequency (IF) signal to be detected and performs an initial frequency conversion filter to obtain a preprocessed IF signal. Preprocessing removes out-of-band interference and noise, improving signal quality and providing a clean data foundation for subsequent frequency point detection and parameter extraction, effectively reducing the probability of false alarms and missed detections. An ordered statistical constant false alarm rate (CFAR) frequency detection algorithm is used to detect the frequency hopping points carried by each pulse in the preprocessed IF signal in real time, obtaining the frequency hopping pattern corresponding to the current time slot. The ordered statistical CFAR algorithm maintains a constant false alarm rate even under multi-target or non-uniform noise backgrounds, detects frequency hopping points in real time, accurately reconstructs the frequency hopping pattern, and enhances the system's anti-interference capability and environmental adaptability. The preprocessed IF signal is then down-converted a second time according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point in the current time slot. Precise down-conversion is performed using the identified frequency hopping pattern to shift each pulse of the frequency hopping signal to zero IF, obtaining the baseband complex signal, which facilitates subsequent synchronization and parameter estimation while reducing signal bandwidth and processing complexity. To prevent the detection of false frequency points in the presence of fixed-frequency interference, this invention employs multi-path parallel detection. This mechanism effectively reduces the negative impact of fixed-frequency interference on subsequent detection performance. Specifically, when the frequency detection stage outputs multiple candidate frequency points that meet the threshold requirements, an independent frequency conversion branch is set up for each candidate frequency point. Each branch performs down-conversion processing using its corresponding candidate frequency point to obtain a baseband complex signal. Compared to the traditional single-path frequency conversion method, which is easily misled by fixed-frequency interference and may select the wrong frequency point, leading to subsequent detection failures, the multi-path parallel structure of this invention ensures the effective preservation of the true signal branch, thereby fundamentally eliminating the destructive effect of fixed-frequency interference on frequency conversion and subsequent detection stages, significantly improving the system's detection reliability and robustness in complex electromagnetic environments. The cross-correlation-autocorrelation peak ratio (CPR) of the frame synchronization header field of the baseband complex signal is used to obtain the frame start time of the current time slot. Combining the CPR and autocorrelation peak ratio methods can resist frequency offset and timing deviation, accurately lock the frame start time, and improve the robustness of frame synchronization. Using the envelope preliminary detection module, the first envelope signal corresponding to each pulse is initially extracted based on the frame start time of the current time slot. The first amplitude corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the first envelope signal. The envelope detection describes the pulse envelope through a four-sliding window detection algorithm, realizes a coarse estimate of the pulse width, and captures the amplitude and frequency at the falling edge of the envelope, providing effective parameters for subsequent envelope fine detection.The preprocessed intermediate frequency signal is down-converted again based on the first frequency value and sent to the envelope fine detection module. At the same time, the first amplitude will be used as the adaptive threshold required for envelope fine detection. The second envelope signal corresponding to each pulse is extracted for the second time. The second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the second envelope signal, and the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are output. Using the initial detection result as prior information for secondary envelope re-detection can eliminate edge errors and noise disturbances in the coarse estimation, obtain more accurate amplitude and pulse width, and improve the accuracy of signal parameter extraction. Attached Figure Description

[0007] Figure 1 A flowchart illustrating the steps of a high-sensitivity adaptive detection method based on FH-TDMA signals provided in this embodiment of the invention; Figure 2 This is a schematic diagram illustrating the frequency detection channel division and the calculation of the power integral value of each channel provided in an embodiment of the present invention; Figure 3 The autocorrelation-cross-correlation correlation waveform diagram provided in the embodiments of the present invention; Figure 4 This is a schematic diagram of four sliding window envelope detection provided in an embodiment of the present invention. Detailed Implementation

[0008] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. All other embodiments obtained by those skilled in the art based on the embodiments of the present invention without creative effort are within the scope of protection of the present invention.

[0009] This invention provides a high-sensitivity adaptive detection method based on FH-TDMA signals, see [link to relevant documentation]. Figure 1 The method includes the following steps S101 to S106.

[0010] S101: Acquire the input intermediate frequency signal to be detected, and perform initial frequency conversion filtering on the intermediate frequency signal to be detected to obtain a preprocessed intermediate frequency signal; Specifically, in step S101, the intermediate frequency signal to be detected is initially converted and filtered to obtain a preprocessed intermediate frequency signal, including the following steps S1011 to S1012.

[0011] S1011, Perform preliminary down-conversion processing on the intermediate frequency signal to be detected to obtain a preliminary down-converted signal; S1012 inputs the initial down-conversion signal to the finite impulse response filter to obtain the pre-processed intermediate frequency signal; the finite impulse response filter is designed with passband and stopband frequencies based on the frequency hopping range of the FH-TDMA signal.

[0012] S102, a frequency detection algorithm based on ordered statistical constant false alarm rate is used to detect the frequency hopping points carried by each pulse in the preprocessed intermediate frequency signal in real time, and obtain the frequency hopping pattern corresponding to the current time slot; Specifically, in step S102, a frequency hopping frequency point carried by each pulse in the preprocessed intermediate frequency signal is detected in real time using a frequency detection algorithm based on ordered statistical constant false alarm rate, and the frequency hopping pattern corresponding to the current time slot is obtained, including the following steps S1021 to S1024.

[0013] S1021, Window the preprocessed intermediate frequency signal to split the continuous preprocessed intermediate frequency signal into multiple short-time segment signals; S1022, perform Fast Fourier Transform on multiple short-time segment signals respectively to obtain the complex spectrum of the current time frame; S1023, map the complex spectrum onto D channels and calculate the power integral value of each channel; where the D channels are centered on the frequency hopping points of the FH-TDMA signal, dividing the working bandwidth of the FH-TDMA signal into a first bandwidth of... D channels; S1024. Using the dual-threshold channel discrimination method, the frequency hopping frequency of each pulse of the FH-TDMA signal in the preprocessed intermediate frequency signal is determined according to the power integral value of each channel, and the frequency hopping pattern corresponding to the current time slot is obtained.

[0014] For example, (1) the preprocessed intermediate frequency signal is windowed to split the continuous signal into short time segments; (2) Use short-time Fourier transform to extract the local spectral features of the preprocessed intermediate frequency signal, that is, perform fast Fourier transform on each windowed short-time segment, as shown in formula (1), to obtain the frequency hopping pattern in the corresponding short time.

[0015] (1); in, Represents the short-time Fourier transform; This indicates the sampling point number in the time domain of the short-time Fourier transform; Represents the spectral line index value in the frequency domain of the short-time Fourier transform; It is both the number of points in the FFT and the window function. The total number of sampling points is inversely proportional to the time resolution of the STFT in the time-frequency domain. The smaller the value, the higher the temporal resolution, but the smaller the number of FFT points; This indicates the sampling point number introduced in the windowed convolution operation; This represents the preprocessed intermediate frequency signal; Indicates the first Find the conjugate of the window function values ​​at each sampling point.

[0016] Based on the formula for calculating frequency resolution, the distance between two adjacent spectral lines in the final plotted spectrum is... This will increase, thus reducing the frequency resolution. Therefore, in order to simultaneously consider both the period of frequency transitions and the accuracy of frequency detection, the time-frequency resolution needs to meet the condition in formula (2): (2); in, express Window length, For the frequency hopping period of the FH-TDMA signal, if Greater than Therefore, the local spectrum of this short segment will contain two frequency hopping points, and it will be impossible to correctly convert each pulse of the FH-TDMA signal to zero frequency subsequently; Indicates the data sampling rate; This indicates the maximum interval between adjacent spectral lines that can be received during frequency detection. If this interval is exceeded, the accuracy of frequency detection will be greatly reduced. It is both the number of points in the FFT and the window function. The total number of sampling points; Indicates the distance between two adjacent spectral lines in the spectrum; (3) Taking each frequency hopping point of the FH-TDMA signal as the center, a total of D channels are divided, where the first bandwidth of each channel is B. According to formulas (2) and (3), a channel with a first bandwidth of B contains FFT points, The power integral value of the channel is obtained by summing the absolute values ​​of the spectral lines corresponding to each FFT point.

[0017] See Figure 2 Taking a 3MHz bandwidth, a 2048-point FFT, and a 320MHz data sampling frequency as an example, the specific channel division is explained; (3); in, Indicates the first bandwidth; Indicates the distance between two adjacent spectral lines in the spectrum; Indicates the data sampling rate; It is both the number of points in the FFT and the window function. The total number of sampling points.

[0018] (4) Perform channel decision, which involves two decision conditions: The first threshold channel discrimination method is to sort the D channel power integral values. A dynamic threshold value is selected based on the OS-CFAR algorithm: ,in, Indicates the first One power integral value, Indicates the standardization factor; The second threshold channel discrimination method uses a fixed noise floor threshold. When the channel power integral value is greater than both of the above thresholds simultaneously, the frequency point corresponding to the channel is determined to be the frequency point of the preprocessed intermediate frequency signal. The frequency points used within a given time period.

[0019] S103, perform a second down-conversion on the preprocessed intermediate frequency signal according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point of the current time slot; Specifically, in step S103, the preprocessed intermediate frequency signal is down-converted twice according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point of the current time slot, including the following steps S1031 to S1033.

[0020] S1031, the preprocessed intermediate frequency signal is divided into X sub-channels using a channelized receiver, and the second bandwidth and sampling rate of the X sub-channels are determined; S1032, according to the second bandwidth, maps the frequency points in the frequency hopping pattern to X sub-channels, and outputs the narrowband low-speed signal corresponding to each frequency point according to the sampling rate; S1033 performs secondary down-conversion processing on the narrowband low-speed signals corresponding to each frequency point to obtain the baseband complex signals corresponding to each frequency point in the current time slot.

[0021] For example, (1) in order to perform secondary downconversion on the acquired preprocessed intermediate frequency signal with less resources, the present invention uses a channelized receiver with low resource overhead and dual-channel parallel processing to receive the entire broadband, dividing the preprocessed intermediate frequency signal into X channels, and setting the second bandwidth of each sub-channel as: Simultaneously, the sampling rate of the output data from each sub-channel becomes the sampling rate of the preprocessed intermediate frequency signal. .

[0022] (2) Map the frequency points in the frequency hopping pattern to X sub-channels according to the second bandwidth, and output the narrowband low-speed signal corresponding to each frequency point according to the sampling rate; (3) Perform a second down-conversion on the above narrowband low-speed signal to convert it into a baseband complex signal. , for subsequent testing.

[0023] It is worth noting that when there is a fixed-frequency interference signal in the environment, in addition to detecting the current correct FH-TDMA signal frequency, S102 will also detect the frequency of the interference signal. Therefore, in order to eliminate the influence of the fixed-frequency interference signal on subsequent detection, the present invention implements multi-channel parallel detection, that is, the preprocessed intermediate frequency signal in S101 is divided into multiple channels, and the operation of S103 is performed simultaneously according to the multiple frequency points detected in S102. If S102 only detects one frequency point, then only the first channel will output valid data, and the remaining channels will not participate in subsequent detection.

[0024] S104, perform cross-correlation-autocorrelation peak ratio detection based on the frame synchronization header field of the baseband complex signal to obtain the frame start time of the current time slot; Specifically, in step S104, the cross-correlation-autocorrelation peak ratio is detected based on the frame synchronization header field of the baseband complex signal to obtain the frame start time of the current time slot, including the following steps S1041 to S1045.

[0025] S1041, Calculate the autocorrelation sequence and cross-correlation sequence of the baseband complex signal based on the frame synchronization header field of the baseband complex signal; Here, the formula for calculating the autocorrelation sequence is expressed as: (4); The formula for calculating cross-correlation sequences is expressed as: (5); in, Indicates the length of the relevant window; Indicates the first in the relevant window One sampling point; This represents the first autocorrelation sequence or cross-correlation sequence. One point; Indicates the first The value of the baseband complex signal at each sampling point; Indicates the first The conjugate value of the baseband complex signal at each sampling point; Indicates the first cross-correlation sequence. The value at each point; Indicates reference signal The conjugate signal; Indicates the th in the autocorrelation sequence The value at each point.

[0026] S1042, Calculate the peak ratio based on the autocorrelation sequence and the cross-correlation sequence, and determine the first decision condition based on the peak ratio; S1043, acquire the pulse period of the FH-TDMA signal, and determine the second decision condition based on the pulse period; S1044, if the baseband complex signal satisfies the first decision condition and the second decision condition, then the synchronization pulse in the baseband complex signal is captured, and the counter value is incremented. S1045, based on the state transition of the counter and status flag and the frame positioning operation, obtain the frame start time of the current time slot.

[0027] For example, (1) when relying solely on the cross-correlation detection algorithm for detection, its decision threshold is directly affected by the signal power, and the threshold value needs to be adjusted in real time according to the signal-to-noise ratio, which has a certain degree of instability. Therefore, in order to improve the adaptability, this invention adopts the autocorrelation-cross-correlation peak ratio detection method for detection. The formula for calculating the peak ratio is shown in formula (6): (6); in, The defined feasible region is the area at the sampling point. Only when the data falls within this range should we begin searching for the peak values ​​of both autocorrelation and cross-correlation. Indicates the first cross-correlation sequence. The value at each point; Indicates the th in the autocorrelation sequence The value at each point; See Figure 3 The simulated autocorrelation-cross-correlation waveforms show that, in a noise-free environment, the ratio of the cross-correlation peak to the autocorrelation peak at the synchronization head is approximately twofold. Considering the influence of noise in the actual environment, as long as the peak ratio meets the requirements... Under the given conditions, the current signal is considered to be the synchronization header of the FH-TDMA signal. Therefore, the second step requires obtaining the peak values ​​of the autocorrelation and cross-correlation within the decision time. (2) Perform a dual judgment. The first judgment is whether the peak ratio meets the above-mentioned first judgment condition: The second judgment is the second decision condition: whether the interval between the current successful detection and the previous successful detection is greater than or equal to one pulse period and less than three pulse periods. This step is to reduce the false alarm probability. If the above two conditions are met, it is considered that a synchronization pulse has been captured, and the counter cnt is incremented by 1; where the pulse period is the pulse period of the FH-TDMA signal.

[0028] (3) Determine whether the frame start time has been located. If cnt=Cnt_Pulse, the frame start time is considered to have been successfully located and the entire synchronization detection is completed. Set cnt=0 and perform the envelope initial detection in S105.

[0029] It is worth noting that the setting of Cnt_Pulse is related to the number of frame synchronization pulses of the FH-TDMA signal. When setting the detection threshold of Cnt_Pulse, some margin is usually left to prevent some fine synchronization pulses from being affected by interference and not being detected under extremely low signal-to-noise ratio.

[0030] In some implementations, the value of the pulse count detection threshold cnt_pulse is determined based on the total number of frame synchronization pulses, specifically set as follows: ; in, This represents the total number of frame synchronization pulses. This indicates the floor function.

[0031] It should be noted that when the total number of frame synchronization pulses When it is less than 3, the above proportionality coefficient No longer applicable, i.e., multiplication is no longer used. Instead of using the standard calculation method, the value of the pulse counting detection threshold cnt_pulse is set separately according to the actual needs of the system. For example, cnt_pulse can be directly set to... Or other preset fixed values. The above settings are intended to avoid the threshold being too low due to proportional calculation when the number of pulses is too small (e.g., The calculation is done proportionally and rounded down to 0, which affects the effectiveness of the test.

[0032] In summary, the rules for setting the pulse counting detection threshold include two constraints: firstly, when At that time, according to The proportion is calculated and then rounded down; secondly, when In this case, disable the proportional rule and determine the cnt_pulse value in another way.

[0033] S105, using the envelope initial detection module, based on the frame start time of the current time slot, the first envelope signal corresponding to each pulse is initially extracted, and the first amplitude value corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the first envelope signal. Specifically, in step S105, the envelope detection module is used to initially extract the first envelope signal corresponding to each pulse based on the frame start time of the current time slot, and capture the first amplitude corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point at the falling edge of the first envelope signal, including the following steps S1051 to S1055.

[0034] S1051, calculate the magnitude of the baseband complex signal and smooth the magnitude to obtain a smoothed magnitude sequence; Here, the formula for calculating the magnitude of the baseband complex signal is: (7); in, This indicates the operation of taking the real part; This indicates the operation of taking the imaginary part; Indicates a baseband complex signal; This represents the magnitude of the baseband complex signal.

[0035] S1052, Based on the modulus sequence and the configuration of the four sliding windows, calculate the first mean sequence, second mean sequence, third mean sequence and fourth mean sequence of each sliding window respectively; Here, the formula for calculating the first mean sequence is specifically expressed as: (8); The formula for calculating the second mean sequence is as follows: (9); The formula for calculating the third mean sequence is as follows: (10); The formula for calculating the fourth mean sequence is as follows: (11); in, Indicates the length of the sliding window; This represents the first, second, third, and fourth mean sequences. One point; Indicates the first in the sliding window One sampling point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the first mean sequence of the first mean sequence. The first mean corresponding to each point; Indicates the second mean sequence of the second mean sequence. The second mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the third mean sequence The third mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the fourth mean sequence. The fourth mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point.

[0036] S1053, determine the rising edge judgment condition and the falling edge judgment condition based on the first mean sequence, the second mean sequence, the third mean sequence and the fourth mean sequence, and convert the modulus into the first envelope signal based on the rising edge judgment condition and the falling edge judgment condition; S1054, Count the duration of the high level in the first envelope signal, and determine whether a valid FH-TDMA signal is detected in the first envelope signal based on the pulse width and detection accuracy of the FH-TDMA signal; S1055 If a valid FH-TDMA signal is detected, the first amplitude corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the envelope signal; otherwise, the first amplitude corresponding to the previous pulse and the first frequency value corresponding to the actual frequency hopping point are not captured.

[0037] For example, see Figure 4 This is illustrated using a sliding window length of 32 as an example. There are 16 points between two adjacent sliding windows, and the rising edge judgment condition is: (1) (2) ,in, Indicates the average noise floor value; Indicates the first The baseband complex signal magnitude at each sampling point; the falling edge judgment condition is: It is important to note that the two conditions for the rising edge condition must be met simultaneously at the same time.

[0038] according to Figure 4 The rising edge and falling edge conditions shown will determine the magnitude of the baseband complex signal. The signal is converted into the first envelope signal. The average noise floor value in the rising edge judgment condition is calibrated according to the actual environment. The average noise floor value in this invention is selected based on the measured system sensitivity, that is, to ensure that the average noise floor value can be used to detect correctly under the sensitivity condition.

[0039] Statistical analysis of the high-level duration of the first envelope signal Assuming the actual width of the FH-TDMA signal If the high level duration satisfy: When a valid FH-TDMA signal is detected, the frequency and amplitude of the current pulse are simultaneously acquired. This indicates the required pulse detection accuracy.

[0040] S106, based on the first amplitude and the first frequency, perform envelope re-examination on the preprocessed intermediate frequency signal, extract the second envelope signal corresponding to each pulse for the second time, capture the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point at the falling edge of the second envelope signal, and output the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point. Specifically, in step S106, based on the first amplitude and the first frequency value, the preprocessed intermediate frequency signal is subjected to envelope re-examination, and the second envelope signal of each pulse is extracted a second time. The second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the second envelope signal, and the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are output. This includes the following steps: (1) Using the envelope re-detection method, based on the first amplitude and the first frequency value, output the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point; wherein, the envelope re-detection method includes: (1.1) Determine whether the time during which no valid FH-TDMA signal pulse is detected during the current pulse envelope re-examination is greater than the stop threshold. If so, stop the envelope re-examination and wait for the next synchronization field to appear. If not, envelope re-examination is performed on the preprocessed intermediate frequency signal to extract the re-examination envelope signal of each pulse of the FH-TDMA signal; the re-examination amplitude and re-examination frequency value corresponding to the current pulse are captured at the falling edge of the re-examination envelope signal; the re-examination amplitude and re-examination frequency value are output as the second amplitude value corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point, respectively.

[0041] For example, (1) the preprocessed intermediate frequency signal after S101 is delayed, and the total delay length is equal to the time taken by S102 to S105. Based on the first frequency value of each pulse collected in S105, the operation in S103 is repeated to generate a complex baseband signal.

[0042] (2) Perform envelope re-inspection. The specific operation method is as shown in S105. The difference is that in this envelope detection, the average noise floor value used in the rising edge judgment condition needs to be set according to the first amplitude value of each pulse obtained in S105, and set to 1 / 4 of the amplitude value.

[0043] Set the stop threshold to TH_STOP. Determine whether a valid FH-TDMA signal has not been detected within the time limit of TH_STOP in S106. If so, return to S104 and prepare to capture the start time of the next time slot data; otherwise, continue to execute S105 to S106.

[0044] This invention provides a high-sensitivity adaptive detection algorithm based on FH-TDMA signals. It utilizes an ordered statistical constant false alarm rate (OS-CFAR) frequency detection algorithm to detect the frequency hopping points carried by each pulse of the FH-TDMA signal. Based on the detected frequency hopping pattern, the FH-TDMA signals are down-converted in parallel to obtain multiple baseband signals, which are then detected in parallel. Each baseband signal is sent to a synchronization detection module, and the data start time of the corresponding time slot is located based on the frame synchronization header field of the FH-TDMA signal. The data start time of the corresponding time slot is then determined. After a certain time, the frame signal is sent from the beginning to the envelope detection module to initially extract the envelope of each pulse of the FH-TDMA signal. The amplitude and actual frequency hopping point of the current pulse are captured at the falling edge of the envelope. Based on the captured frequency and amplitude values, the buffered intermediate frequency data is down-converted and the envelope is re-detected to further improve the pulse width detection accuracy of the FH-TDMA signal. It is determined whether the detection gap period of the envelope re-detection module is greater than the set threshold. If so, the synchronization detection state is returned to prepare to capture the start time of the next time slot data; otherwise, the envelope detection continues.

[0045] This invention employs an ordered statistical constant false alarm rate (OS-CFAR) frequency detection algorithm, which significantly reduces the probability of frequency misdetection in cluttered environments. Furthermore, it utilizes multi-path parallel detection to minimize the negative impact of fixed-frequency interference signals on subsequent detection. In addition, this invention combines the high precision of synchronous detection with the low complexity of envelope detection. During down-conversion, channelization is used to partition the wideband spectrum of FH-TDMA, reducing the rate of subsequent processing. The algorithm features high sensitivity and low hardware overhead, making it easier to implement on FPGA hardware.

[0046] The various embodiments described in this specification are presented in a progressive manner. Similar or identical parts between embodiments can be referred to interchangeably. Each embodiment focuses on its differences from other embodiments. All or part of this invention can be used in numerous general-purpose or special-purpose computer system environments or configurations. Examples include: personal computers, server computers, handheld or portable devices, tablet devices, mobile communication terminals, multiprocessor systems, microprocessor-based systems, programmable electronic devices, network PCs, minicomputers, mainframe computers, and distributed computing environments including any of the above systems or devices, etc.

[0047] The above embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some or all of the technical features therein. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the present invention.

Claims

1. A high-sensitivity adaptive detection method based on FH-TDMA signals, characterized in that, include: The input intermediate frequency signal to be detected is acquired, and the intermediate frequency signal to be detected is initially frequency-converted and filtered to obtain a preprocessed intermediate frequency signal; A frequency hopping frequency point carried by each pulse in the preprocessed intermediate frequency signal is detected in real time using an ordered statistical constant false alarm rate (CFAR) frequency detection algorithm to obtain the frequency hopping pattern corresponding to the current time slot. The preprocessed intermediate frequency signal is down-converted twice according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point of the current time slot; The frame start time of the current time slot is obtained by performing cross-correlation-autocorrelation peak ratio detection based on the frame synchronization header field of the baseband complex signal; Using the envelope detection module, the first envelope signal corresponding to each pulse is initially extracted based on the frame start time of the current time slot, and the first amplitude value corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the first envelope signal. Based on the first amplitude and the first frequency value, the preprocessed intermediate frequency signal is subjected to envelope re-examination, and the second envelope signal corresponding to each pulse is extracted for the second time. At the falling edge of the second envelope signal, the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are captured, and the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point are output.

2. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The initial frequency conversion filtering of the intermediate frequency signal to be detected to obtain a preprocessed intermediate frequency signal includes: The intermediate frequency signal to be detected is subjected to preliminary down-conversion processing to obtain a preliminary down-converted signal; The initial down-conversion signal is input to a finite impulse response filter to obtain a pre-processed intermediate frequency signal; wherein, the passband and stopband frequencies of the finite impulse response filter are designed based on the frequency hopping frequency range of the FH-TDMA signal.

3. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The frequency hopping frequency points carried by each pulse in the preprocessed intermediate frequency signal are detected in real time using an ordered statistical constant false alarm rate (CFAR) frequency detection algorithm to obtain the frequency hopping pattern corresponding to the current time slot, including: The preprocessed intermediate frequency signal is windowed to split the continuous preprocessed intermediate frequency signal into multiple short-time segment signals; The complex spectrum of the current time frame is obtained by performing Fast Fourier Transform on multiple short-time segment signals respectively. The complex spectrum is mapped onto D channels, and the power integral value of each channel is calculated; wherein, the D channels are defined by dividing the working bandwidth of the FH-TDMA signal into D channels with a first bandwidth of B, centered on each frequency hopping point of the FH-TDMA signal. Using a dual-threshold channel discrimination method, the frequency hopping points carried by each pulse of the FH-TDMA signal in the preprocessed intermediate frequency signal are determined based on the power integral value of each channel, thus obtaining the frequency hopping pattern corresponding to the current time slot.

4. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The step of performing a second down-conversion on the preprocessed intermediate frequency signal according to the frequency hopping pattern to obtain the baseband complex signal corresponding to each frequency point in the current time slot includes: The preprocessed intermediate frequency signal is divided into X sub-channels using a channelized receiver, and the second bandwidth and sampling rate of the X sub-channels are determined. Based on the second bandwidth, the frequency points in the frequency hopping pattern are mapped to X sub-channels, and the narrowband low-speed signal corresponding to each frequency point is output according to the sampling rate; The narrowband low-speed signals corresponding to each frequency point are subjected to secondary down-conversion processing to obtain the baseband complex signals corresponding to each frequency point in the current time slot.

5. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The step of performing cross-correlation-autocorrelation peak ratio detection based on the frame synchronization header field of the baseband complex signal to obtain the frame start time of the current time slot includes: Based on the frame synchronization header field of the baseband complex signal, calculate the autocorrelation sequence and cross-correlation sequence of the baseband complex signal; The peak ratio is calculated based on the autocorrelation sequence and the cross-correlation sequence, and a first decision condition is determined based on the peak ratio; Obtain the pulse period of the FH-TDMA signal, and determine the second decision condition based on the pulse period; If the baseband complex signal satisfies the first decision condition and the second decision condition, then the synchronization pulse in the baseband complex signal is captured, and the counter value is incremented. Based on the state transitions of the counter and state flag, as well as the frame positioning operation, the frame start time of the current time slot is obtained.

6. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 5, characterized in that, The formula for calculating the autocorrelation sequence is as follows: ； The formula for calculating the cross-correlation sequence is expressed as follows: ； in, Indicates the length of the relevant window; Indicates the first in the relevant window One sampling point; This represents the first autocorrelation sequence or cross-correlation sequence. One point; Indicates the first The value of the baseband complex signal at each sampling point; Indicates the first The conjugate value of the baseband complex signal at each sampling point; Indicates the first cross-correlation sequence. The value at each point; Indicates reference signal The conjugate signal; Indicates the th in the autocorrelation sequence The value at each point.

7. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The envelope detection module initially extracts the first envelope signal corresponding to each pulse based on the frame start time of the current time slot, and captures the first amplitude value corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point at the falling edge of the first envelope signal, including: Calculate the magnitude of the baseband complex signal and smooth the magnitude to obtain a smoothed magnitude sequence; Based on the modulus sequence and the configuration of the four sliding windows, calculate the first mean sequence, second mean sequence, third mean sequence and fourth mean sequence of each sliding window respectively; The rising edge judgment condition and the falling edge judgment condition are determined based on the first mean sequence, the second mean sequence, the third mean sequence and the fourth mean sequence, and the modulus value is converted into a first envelope signal based on the rising edge judgment condition and the falling edge judgment condition; The duration of the high level in the first envelope signal is counted, and based on the pulse width of the FH-TDMA signal and the detection accuracy, it is determined whether a valid FH-TDMA signal is detected in the first envelope signal. If a valid FH-TDMA signal is detected, the first amplitude corresponding to the current pulse and the first frequency value corresponding to the actual frequency hopping point are captured at the falling edge of the envelope signal; otherwise, the first amplitude corresponding to the previous pulse and the first frequency value corresponding to the actual frequency hopping point are not captured.

8. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 7, characterized in that, The formula for calculating the first mean sequence is as follows: ； The formula for calculating the second mean sequence is as follows: ； The formula for calculating the third mean sequence is as follows: ； The formula for calculating the fourth mean sequence is as follows: ； in, Indicates the length of the sliding window; This represents the first, second, third, and fourth mean sequences. One point; Indicates the first in the sliding window One sampling point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the first mean sequence of the first mean sequence. The first mean corresponding to each point; Indicates the second mean sequence of the second mean sequence. The second mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the third mean sequence The third mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point; Indicates the fourth mean sequence. The fourth mean corresponding to each point; Indicates the first The baseband complex signal modulus at each sampling point.

9. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 7, characterized in that, The formula for calculating the magnitude of the baseband complex signal is as follows: ； in, This indicates the operation of taking the real part; This indicates the operation of taking the imaginary part; Indicates a baseband complex signal; This represents the magnitude of the baseband complex signal.

10. The high-sensitivity adaptive detection method based on FH-TDMA signals according to claim 1, characterized in that, The step of performing envelope re-examination on the preprocessed intermediate frequency signal based on the first amplitude and the first frequency value, extracting the second envelope signal of each pulse for the second time, capturing the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point at the falling edge of the second envelope signal, and outputting the second amplitude corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point, includes: Using an envelope re-detection method, based on the first amplitude and the first frequency value, a second amplitude corresponding to the current pulse and a second frequency value corresponding to the actual frequency hopping point are output; wherein, the envelope re-detection method includes: If the time during which no valid FH-TDMA signal pulse is detected during the current pulse envelope re-examination is greater than the stop threshold, then the envelope re-examination is stopped and the system waits for the next frame synchronization field to appear. If not, envelope re-examination is performed on the preprocessed intermediate frequency signal to extract the re-examination envelope signal of each pulse of the FH-TDMA signal; the re-examination amplitude and re-examination frequency value corresponding to the current pulse are captured at the falling edge of the re-examination envelope signal; the re-examination amplitude and the re-examination frequency value are output as the second amplitude value corresponding to the current pulse and the second frequency value corresponding to the actual frequency hopping point, respectively.

Images