A sliding correlation channel detection system and PN sequence pulse waveform optimization method

Abstract

The present invention provides a sliding correlation channel detection system and a PN sequence pulse waveform optimization method, which belongs to the field of wireless channel detection technology. The sliding correlation channel detection system includes a transmitting end and a receiving end. The PN sequences generated by the transmitting end and the receiving end are multiplied and filtered. The dynamic range formula of the time expansion autocorrelation of any PN sequence pulse waveform is used to measure the relationship between the main lobe and the side lobe. The 3dB width of the time expansion autocorrelation is established to measure whether the energy is concentrated on the main lobe. The dynamic range and the 3dB width are combined to measure the similarity between the time expansion autocorrelation and the Dirac function to obtain a more accurate channel pulse response. The PN sequence pulse waveform is constructed by using PSWF, and then converted into a nonlinear constraint optimization problem and solved. The present invention solves the generation of the PN sequence pulse shape used in the sliding correlation channel detection system, so that the time expansion autocorrelation is closer to the Dirac function, and the detected wireless channel pulse response can be made more accurate.

Description

Technical field:

[0001] The invention relates to the technical field of wireless channel detection, in particular to a sliding correlation channel detection system and a PN sequence pulse waveform optimization method. Background technology:

[0002] Channel detection is a method of sensing the characteristics of a wireless channel by transmitting a known sequence and observing how the signal is affected from transmission to reception. It can observe and analyze the impact and changes on the amplitude and phase of signals from different paths.

[0003] In the prior art, a wireless channel parameter measurement system is usually required for wireless channel parameter measurement. The wireless channel parameter measurement system includes a measurement signal transmitter and a receiving processor. The core of the wireless channel parameter measurement system design is the waveform generation method of the measurement signal. Different measurement signal waveforms not only determine the system structure and signal processing method of the transmitter and receiving processor of the wireless channel parameter measurement system, but also determine the measurement capabilities of the system such as the types of measurable parameters, measurement accuracy, real-time performance, bandwidth, and multi-antenna measurement.

[0004] As the measurement bandwidth increases, the sampling rate requirements of the analog-to-digital converter (ADC) also increase accordingly, and the cost also increases accordingly. Therefore, in actual application scenarios, it is necessary to consider how to reduce the requirements for hardware devices.

[0005] The emergence of sliding correlator channel sounder provides a new method for detecting wireless channels. Its most notable feature is that the time dilation effect leads to bandwidth compression, which makes the sampling rate of the ADC converter not as high as required by the cross-correlation reception technology, allowing measurements in ultra-wideband channels. In sliding correlation, a signal modulated by a PN sequence is transmitted through a wireless channel, and at the receiving end, a cross-correlation is performed between the received signal and a copy of the PN sequence delayed from each other. In most studies, only rectangular pulses are used as the chip pulse shape of the PN sequence. However, the chip pulse shape of the PN sequence has an unignorable impact on the performance of the system, and there is little discussion on the chip pulse shape. In addition, the signal design should be able to achieve robust code tracking and acquisition in the receiving processor, and whether the receiving processor can capture the CIR well is worth considering. Finally, dynamic range is an important indicator to measure the performance of the sliding correlator. The existing calculation formula for the dynamic range of the sliding correlator is only for rectangular pulses, and no general expression for any chip pulse shape is given. Therefore, it is particularly important to find a waveform optimization method to solve the above problems. Summary of the invention:

[0006] The technical problem solved by the present invention is: in view of the shortcomings of the prior art, the present invention designs a sliding correlation channel detection system and a PN sequence pulse waveform optimization method, which has simple steps, good detection effect, can quickly detect wireless channels, and has low hardware requirements.

[0007] To achieve the above technical objectives, the present invention adopts the following technical solutions: a sliding correlation channel detection system, including a transmitting end and a receiving end, wherein the transmitting end sends a PN sequence to the receiving end through a channel, and then the receiving end generates a PN sequence that is only slightly delayed and has the same other parameters, and multiplies the PN sequence with the transmitting end, and performs low-pass filtering on the multiplied signal to filter out the self-interference generated by the sliding correlation channel detection, thereby obtaining a time expansion autocorrelation, and then obtaining a channel impulse response. The closer the obtained time expansion autocorrelation is to the Dirac function, the more accurate the obtained channel impulse response is, and the better the detection performance of the channel detection system is; the time expansion autocorrelation is affected by the pulse shape of the PN sequence, so by selecting a suitable PN sequence pulse shape, the time expansion autocorrelation is closer to the Dirac function, thereby obtaining a more accurate channel impulse response.

[0008] A PN sequence pulse waveform optimization method applied to a sliding correlation channel detection system, by changing the PN sequence pulse waveform to make the time dilation autocorrelation closer to the Dirac function, so as to improve the accuracy of the channel impulse response obtained by detection, the method comprising:

[0009] Step 1: Design a sliding correlation channel detection system. The transmitter sends a PN sequence x(t) to the receiver through the channel. The receiver generates a PN sequence x'(t) with a slight delay and the same other parameters. Compared with the transmitter's PN sequence x(t), the PN sequence x'(t) only has a shift register clock frequency f c 'Slightly less than f c , other parameters are the same;

[0010] The expression of the PN sequence x(t) sent by the transmitter is:

[0011]

[0012] In the formula, ai is the length L = 2 m -1 m sequence, a i ∈{0,1}; Bi-phase output with amplitude ±V0, when ai=0 corresponds to -V0, when a i =1 corresponds to +V0; g(t) is the PN sequence pulse shape; t0 = 1 / f c , f c is the clock frequency of the shift register, that is, after every t0 seconds, the next item of the m sequence is generated; it0 represents the translation of g(t);

[0013] Step 2: construct a time dilation autocorrelation suitable for calculating any PN sequence pulse waveform to measure the relationship between the main lobe peak and the side lobe peak of the time dilation autocorrelation; this step is the first indicator to measure the similarity between the time dilation autocorrelation and the Dirac function, and the expression is:

[0014]

[0015] Where, the numerator is the time dilation autocorrelation function q c (t) Peak amplitude after filtering, the denominator is the self-interference q d (t) The amplitude after filtering and q c (t) side lobe q c,sl (t) The peak value of the filtered amplitude sum; H LP (t) represents a low-pass filter;

[0016] Step 3: Use the 3dB width of the time dilation autocorrelation to measure whether the time dilation autocorrelation energy is concentrated on the main lobe. This step is the second indicator to measure the similarity between the time dilation autocorrelation and the Dirac function. The expression is:

[0017] W 3dB =20log 10 (AB)(dB),

[0018] In the formula, A represents q c (t) The position corresponding to half the peak amplitude after filtering is 1, and B represents q c (t) Position 2 corresponding to half the peak amplitude after filtering;

[0019] Step 4: Use the ellipsoidal spherical wave function PSWF to design the pulse shape of the PN sequence. The expression is:

[0020]

[0021] In the formula, ψ n (c, t) is the ellipsoidal wave function, N e is the order of the ellipsoidal wave function, and its value is determined by N e =2c / π, μ n is the coefficient of each order expansion of the ellipsoidal wave function, c=2πt0f0 is the time-bandwidth product, t0 is the duration of the signal pulse, and f0 is the upper frequency limit;

[0022] Step 5: Construct the optimization problem with nonlinear constraints: Optimize the PN sequence pulse waveform by constructing the coefficients of each order expansion of the elliptical spherical wave function. The expression is:

[0023]

[0024] st

[0025] DR PSWF >DR RECT

[0026] and

[0027] W 3dB,PSWF <W 3dB,RECT ,

[0028] In the formula, represents the coefficient of each order expansion of the ellipsoidal wave function obtained after optimization, Indicates that when {·} reaches its maximum value, the variable μ n The value of β DR and β 3dB Both are non-negative numbers, representing the dynamic range of time dilation autocorrelation and the weight of 3dB width, respectively. They are selected according to actual needs. DR PSWF and DR RECT They represent the dynamic range of the time dilation autocorrelation of the optimized PN sequence pulse waveform and the dynamic range of the PN sequence pulse shape as a rectangular pulse, respectively. 3dB,PSWF and W 3dB,RECT They represent the 3dB bandwidth of the time dilation autocorrelation of the optimized PN sequence pulse waveform and the 3dB bandwidth of the PN sequence pulse shape as a rectangular pulse;

[0029] Step 6: Use genetic algorithm to solve the optimization problem of nonlinear constraints in step 5. During the genetic algorithm calculation process, the algorithm stops when the average relative change of the best fitness function value is less than or equal to the threshold; that is, when β DR ×DR PSWF -β 3dB ×W 3dB,PSWF The average relative change is less than or equal to 1e -6 Get the optimized result The coefficients of each order expansion of the ellipsoidal wave function obtained through optimization are Substitute the pulse shape formula of the PN sequence in step 4 and solve it, and finally obtain the optimized waveform of the PN sequence pulse shape.

[0030] Furthermore, in step 1, the signal obtained by multiplying the PN sequence sent by the transmitter and the delayed PN sequence generated by the receiver is used as the output signal of the sliding correlation channel detection system, which can be divided into two parts: time expansion autocorrelation q c (t) and self-interference q d (t), self-interference q needs to be filtered out d (t), retaining the time-dilated autocorrelation q c (t); where the time dilation autocorrelation q c The autocorrelation function R of (t) and x(t) x(t) satisfies the following formula:

[0031]

[0032] Time dilated autocorrelation q c (t) is equivalent to the autocorrelation function R of x(t) x (t) dilation in the time domain. According to the relationship between the Fourier transform time domain and the frequency domain, the time dilation autocorrelation q c (t) has a compression effect in the frequency domain.

[0033] Furthermore, in step 2, the dynamic range DR of the time dilation autocorrelation is an indicator to measure the similarity between the time dilation autocorrelation and the Dirac function, which is the dynamic range of the time dilation autocorrelation after filtering. c (t) Main lobe peak and q c (t) sidelobes and self-interference q d (t) is the ratio of the sum of the peaks.

[0034] Furthermore, in step 2 and step 3, the dynamic range of the time dilation autocorrelation and the 3 dB width of the time dilation autocorrelation are considered together to measure the similarity between the time dilation autocorrelation and the Dirac function, and the two performances can be given weights β according to the actual measurement scenario. DR and β 3dB To meet the scene requirements and obtain a more accurate channel impulse response.

[0035] Furthermore, in step 4, the pulse waveform of the PN sequence at the transmitting end can be constructed by linear superposition of PSWF, and the energy of the optimized pulse waveform is concentrated at [-t0, t0] in the time domain and is limited to [-f0, f0] in the frequency domain.

[0036] Further, in step 5, by formula: β DR ×DR PSWF -β 3dB ×W 3dB,PSWF It is used to compare the similarity between the time dilation autocorrelation of the optimized PN sequence pulse shape and the Dirac function. The larger the value, the higher the similarity, that is, the better the detection performance of the system.

[0037] Beneficial effects:

[0038] The present invention proposes a dynamic range formula suitable for calculating the time expansion autocorrelation of any PN sequence pulse waveform to measure the relationship between the main lobe and the side lobe, establishes the 3dB width of the time expansion autocorrelation to measure whether the energy is concentrated on the main lobe, combines the dynamic range and the 3dB width to measure the similarity between the time expansion autocorrelation and the Dirac function, and is used to obtain a more accurate channel impulse response. Finally, a method for designing and optimizing the PN sequence pulse waveform is provided, and the PN sequence pulse waveform is constructed by using a PSWF, and then converted into an optimization problem with nonlinear constraints, and solved by a genetic algorithm. The time expansion autocorrelation is made closer to the Dirac function, so that the detected wireless channel impulse response is more accurate, so that the currently detected wireless channel is better evaluated.

[0039] The present invention generates a pulse shape of a PN sequence based on a PSWF, solves the generation of a PN sequence pulse shape for a sliding correlation channel detection system, and makes the time dilation autocorrelation closer to a Dirac function to obtain a more accurate detection value, thereby making a more accurate evaluation of a wireless channel. Description of the drawings:

[0040] Figure 1 It is a schematic diagram of the sliding correlation channel detection system of the present invention;

[0041] Figure 2 : is a schematic diagram of 3dB bandwidth in an embodiment of the present invention. Specific implementation method:

[0042] The embodiments of the present invention are further described below in conjunction with the accompanying drawings:

[0043] like Figure 1 As shown, a sliding correlation channel detection system includes a transmitting end and a receiving end, wherein the transmitting end sends a PN sequence to the receiving end through a channel, and then the receiving end generates a PN sequence with only a slight delay and the same other parameters, which is multiplied with the PN sequence of the transmitting end, and the multiplied signal is low-pass filtered to filter out the self-interference generated by the sliding correlation channel detection, so as to obtain a time expansion autocorrelation, and then obtain a channel impulse response. The closer the obtained time expansion autocorrelation is to the Dirac function, the more accurate the obtained channel impulse response is, and the better the detection performance of the channel detection system is; the time expansion autocorrelation is affected by the pulse shape of the PN sequence, so by selecting a suitable PN sequence pulse shape, the time expansion autocorrelation is closer to the Dirac function, so as to obtain a more accurate channel impulse response.

[0044] A PN sequence pulse waveform optimization method applied to a sliding correlation channel detection system is used to detect the quality of wireless channels in wireless communication technology. Through the dynamic range formula applicable to calculating any pulse waveform proposed by the present invention, the 3dB width of the channel pulse response is established to measure whether the energy is concentrated on the main lobe. Finally, a method for designing and optimizing the PN sequence pulse waveform is given. Compared with the system in which the PN sequence is a rectangular pulse, the waveform designed by the present invention has a good Dirac function similarity.

[0045] The details are as follows:

[0046] Step 1: At the transmitting end, the transmitter sends a PN sequence, which is expressed as:

[0047]

[0048] In the formula, ai is the length L = 2 m -1 m sequence, a i ∈{0,1}. Bi-phase output with amplitude ±V0, when ai=0 corresponds to -V0, when a i =1 corresponds to +V0. g(t) is the PN sequence pulse shape. t0 = 1 / f c , f c is the clock frequency of the shift register, that is, after every t0 seconds, the next item of the m sequence is generated.

[0049] Step 2: At the receiving end, generate a clock frequency of f c 'Slightly slower than the sender, but with the same other parameters as the sender. Multiply these two similar signals, and the signal expression after multiplication is:

[0050] p(t)=x(t)x′(t),

[0051] In the formula, x(t) and x'(t) are two PN sequences, and the chip rate of x(t) is f c , the chip rate of x'(t) is f c ', these two PN sequences are m sequences of the same length L and amplitude V0.

[0052] Step 3: Low-pass filter the multiplied signal p(t). The signal generated after filtering will produce a time-expanded autocorrelation. The ideal low-pass filter for filtering the multiplied information is:

[0053] h LP (f) = u(f + |f c / γ)-u(f-|f c / γ|),

[0054] Where u(·) is a step function, γ = f c / (fc -f c ') is called the sliding factor.

[0055] Step 4: Convert the multiplied signal to the frequency domain for analysis, and obtain the Fourier transform expression of p(t), and achieve compression in the frequency domain by sliding correlation:

[0056]

[0057] Where P(f), X(f) and X'(f) are the Fourier transforms of p(t), x(t) and x'(t), respectively, and the expression of X(f) is:

[0058]

[0059] Wherein, L is the length of the m-sequence, i represents the chip number of the m-sequence, T represents one period of the PN sequence, T=Lt0, G(·) is the Fourier transform of g(t), δ(·) represents the Dirac function, k represents the harmonic number of the Fourier series of X(f), which is an integer from negative infinity to positive infinity, and j represents the imaginary unit.

[0060] X'(f) can also be calculated by a similar formula:

[0061]

[0062] Wherein, L is the length of the m-sequence, i' represents the chip number of the m-sequence, T' represents one period of the PN sequence, T'=Lt0', G(·) is the Fourier transform of g(t), δ(·) represents the Dirac function, k' represents the harmonic number of the Fourier series of X'(f), which is an integer from negative infinity to positive infinity, and j represents the imaginary unit.

[0063] Therefore, P(f) can be expressed as:

[0064]

[0065] Step 5: Decompose P(f) into two parts: k=-k' and k≠-k', and use Q c (f) and Q d (f) means:

[0066]

[0067]

[0068] Part IQ c (f) is called the time-dilated autocorrelation spectrum, which is compressed in frequency relative to the actual spectrum of x(t); the second part Q d(f) is the noise-like broadband distortion signal generated by the sliding correlator itself, which is called self-interference or self-noise. This distortion signal will reduce the dynamic range of the system.

[0069] Step 6: According to the relationship between k and k', p(t) can be expressed in the time domain as:

[0070] p(t)=q c (t)+q d (t),

[0071] In the formula, q c (t) is the time-dilated autocorrelation function of x(t), which is what we want; q d (t) is the distortion signal generated by the sliding correlator itself;

[0072] Step 7: The channel impulse response h(t) of the tapped delay line model TDL is expressed as:

[0073]

[0074] In the formula, R xy (τ) is the cross-correlation function between the input signal x(t) and the output signal y(t), R xy (τ) can be viewed as the scaled channel impulse response, and has:

[0075]

[0076] Where δ(τ) is the Dirac function, N is the number of TDL channel taps, and τ s Respectively represent the complex amplitude and delay corresponding to each tap. Therefore, the channel impulse response can be obtained by detecting the autocorrelation of the PN sequence.

[0077] Furthermore, the time-dilated autocorrelation q c The autocorrelation function R of (t) and x(t) x (t) satisfies the following formula:

[0078]

[0079] That is, q c (t) has a dilation in the time domain, so Q c (f) There is a compression effect in the frequency domain relative to the actual spectrum of x(t), that is, through sliding correlation detection, frequency domain compression can be achieved to reduce the requirements for the analog-to-digital converter (ADC) of the hardware device in practical applications, and a wider bandwidth can be measured.

[0080] A dynamic range formula suitable for calculating the time dilation autocorrelation of arbitrary pulse waveforms. The dynamic range is used to measure the relationship between the main lobe and side lobes of the time dilation autocorrelation, which is an indicator of the similarity with the Dirac function:

[0081]

[0082] Where, the numerator is the time dilation autocorrelation function q c (t) Peak amplitude after filtering, the denominator is the self-interference q d (t) The amplitude after filtering and q c (t) side lobe q c,sl (t) The peak value of the filtered amplitude sum, H LP (t) is a low-pass filter.

[0083] The 3dB width of the time dilation autocorrelation is used to measure whether the energy is concentrated in the main lobe. The 3dB width diagram is shown in Figure 2 As shown:

[0084] W 3dB =20log 10 (|AB|)(dB),

[0085] In the formula, A represents q c (t) The position corresponding to half the peak amplitude after filtering is 1, and B represents q c (t) Position 2 corresponding to half the peak amplitude after filtering.

[0086] A method for designing and optimizing PN sequence pulse waveform:

[0087] The PN sequence pulse shape g(t) can be represented by the linear superposition of ellipsoidal spherical wave functions (PSWF):

[0088]

[0089] In the formula, ψ n (c, t) is the ellipsoidal wave function, N e is the order taken, the value is determined by N e =2c / π, μ n is the coefficient of each order expansion, c=2πt0f0 is the time-bandwidth product, t0 is the duration of the signal pulse, and f0 is the upper frequency limit. The Fourier transform of g(t) is:

[0090]

[0091] In the formula, Ψ n (c,ω) is ψ n The Fourier transform of (c, t) is expressed as:

[0092]

[0093] According to the requirements of the designed waveform, the time-bandwidth product of the pulse waveform is first determined, and then the order of the PSWF to be used is obtained. Then, the PN sequence pulse shape is constructed by changing the coefficients of the PSWF to improve the system performance. Finally, the problem is transformed into a parameter optimization problem, and the genetic algorithm is used to solve the optimal coefficient solution. The coefficients of the PSWF finally optimized are used to construct the pulse waveform of the PN sequence, so that the time dilation autocorrelation is closer to the Dirac function. The sliding correlation channel detection system using the PN sequence pulse shape has a stronger ability to detect low-intensity signals, and the detected channel pulse response is more accurate.

[0094] Embodiment 1

[0095] Step 1: Express the PN sequence pulse shape with PSWF to obtain the time-expanded autocorrelation spectrum Q c,PSWF (f) and self-interference spectrum Q d,PSWF (f), the expression is as follows:

[0096]

[0097]

[0098] Step 2: Q c,PSWF (f) and Q d,PSWF (f) Perform inverse Fourier transform to obtain the time dilation autocorrelation q in the time domain c,PSWF (t) and self-interference q d,PSWF (t), the expression is as follows:

[0099]

[0100]

[0101] Step 3: Calculate the time-bandwidth product c of the optimized waveform, and then calculate the time-bandwidth product c according to N e =2c / π to get the PSWF order to be optimized. The optimization goal is to obtain the maximum dynamic range and the minimum 3dB bandwidth of the channel impulse response. The constraint is that the dynamic range of the optimized pulse waveform system must be greater than the dynamic range of the rectangular pulse system, and the 3dB bandwidth of the channel impulse response of the optimized pulse waveform system must be smaller than the 3dB bandwidth of the channel impulse response of the rectangular pulse system. Therefore, we can get an optimization problem with nonlinear constraints, which is expressed as follows:

[0102]

[0103] st

[0104] DR PSWF >DRRECT

[0105] and

[0106] W 3dB,PSWF <W 3dB,RECT ,

[0107] In the formula, β DR and β 3dB Both are non-negative numbers, representing the weights of dynamic range and 3dB width, depending on which performance we value more. PSWF and DR RECT They represent the optimized dynamic range and the dynamic range of the rectangular pulse, respectively. 3dB,PSWF and W 3dB,RECT They represent the optimized 3dB bandwidth and the 3dB bandwidth of the rectangular pulse respectively.

[0108] Step 4: Use a genetic algorithm to obtain the optimal value of the above nonlinear constraint optimization problem. When the average relative change of the best fitness function value is less than or equal to the threshold, the algorithm stops and the optimal PSWF coefficients are obtained. The optimal pulse waveform is obtained by summing up the various coefficients.

Claims

1. A PN sequence pulse waveform optimization method for a sliding correlation channel detection system, characterized in that: The sliding correlation channel detection system used includes a transmitting end and a receiving end, wherein the transmitting end sends a PN sequence to the receiving end through a channel, and then the receiving end generates a PN sequence with only a slight delay and the same other parameters, which is multiplied with the PN sequence of the transmitting end, and the multiplied signal is low-pass filtered to filter out the self-interference generated by the sliding correlation channel detection, so as to obtain a time expansion autocorrelation, and then obtain a channel impulse response. The closer the obtained time expansion autocorrelation is to the Dirac function, the more accurate the obtained channel impulse response is, and the better the detection performance of the channel detection system is; the time expansion autocorrelation is affected by the PN sequence pulse waveform, so by selecting a suitable PN sequence pulse waveform, the time expansion autocorrelation is made closer to the Dirac function, so as to obtain a more accurate channel impulse response; By changing the PN sequence pulse waveform to make the time dilation autocorrelation closer to the Dirac function, the accuracy of the channel impulse response obtained by detection is improved. The method includes: Step 1: Design a sliding correlation channel detection system. The transmitter sends a PN sequence x(t) to the receiver through the channel. The receiver generates a PN sequence x'(t) with a slight delay and the same other parameters. Compared with the transmitter's PN sequence x(t), the PN sequence x'(t) only has a shift register clock frequency f c 'Slightly smaller than f c , other parameters are the same; The expression of the PN sequence x(t) sent by the transmitter is: In the formula, a i The length is L = 2 m -1 m sequence, a i ∈{0,1}; bi-phase output with amplitude ±V0, when a i =0 corresponds to -V0, when a i =1 corresponds to +V0; g(t) is the PN sequence pulse waveform; t0 = 1 / f c , f c is the clock frequency of the shift register, that is, after every t0 seconds, the next item of the m sequence is generated; it0 represents the translation of g(t); Step 2: construct a time dilation autocorrelation suitable for calculating any PN sequence pulse waveform to measure the relationship between the main lobe peak and the side lobe peak of the time dilation autocorrelation; this step is used to calculate the first indicator for measuring the similarity between the time dilation autocorrelation and the Dirac function, and the expression is: Where, the numerator is the time dilation autocorrelation function q c (t) Peak amplitude after filtering, the denominator is the self-interference q d (t) The amplitude after filtering and q c Side lobe q of (t) c,sl (t) The peak value of the filtered amplitude sum; H LP (t) represents a low-pass filter; Step 3: Use the 3 dB width of the time dilation autocorrelation to measure whether the time dilation autocorrelation energy is concentrated on the main lobe. This step is used to measure the second indicator of the similarity between the time dilation autocorrelation and the Dirac function. The expression is: W 3dB =20log 10 (A-B)(dB), In the formula, A represents q c (t) The position corresponding to half the peak amplitude after filtering is 1, and B represents q c (t) Position 2 corresponding to half the peak amplitude after filtering; Step 4: Use the ellipsoidal spherical wave function PSWF to design the pulse shape of the PN sequence. The expression is: In the formula, ψ n (c, t) is the ellipsoidal wave function, N e is the order of the ellipsoidal wave function, and its value is determined by N e =2c / π, μ n is the coefficient of each order expansion of the ellipsoidal wave function, c=2πt0f0 is the time-bandwidth product, t0 is the duration of the signal pulse, and f0 is the upper frequency limit; Step 5: Construct the optimization problem with nonlinear constraints: Optimize the PN sequence pulse waveform by constructing the coefficients of each order expansion of the elliptical spherical wave function. The expression is: st DR PSWF >DR RECT and IN 3dB,PSWF <In 3dB,RECT , In the formula, represents the coefficient of each order expansion of the ellipsoidal wave function obtained after optimization, Indicates that when {·} reaches its maximum value, the variable μ n The value of β DR and β 3dB Both are non-negative numbers, representing the dynamic range of time dilation autocorrelation and the weight of 3dB width, respectively. They are selected according to actual needs. DR PSWF and DR RECT They represent the dynamic range of the time dilation autocorrelation of the optimized PN sequence pulse waveform and the dynamic range of the PN sequence pulse waveform as a rectangular pulse, respectively. 3dB,PSWF and W 3dB,RECT They represent the 3dB bandwidth of the time dilation autocorrelation of the optimized PN sequence pulse waveform and the 3dB bandwidth of the PN sequence pulse waveform as a rectangular pulse; Step 6: Use genetic algorithm to solve the optimization problem of nonlinear constraints in step 5. During the genetic algorithm calculation process, the algorithm stops when the average relative change of the best fitness function value is less than or equal to the threshold; that is, when β DR ×DR PSWF -β 3dB ×W 3dB,PSWF The average relative change is less than or equal to 1e -6 Get the optimized result The coefficients of each order expansion of the ellipsoidal wave function obtained through optimization are Substitute the pulse shape formula of the PN sequence in step 4 and solve it, and finally obtain the optimized waveform of the PN sequence pulse waveform.

2. A PN sequence pulse waveform optimization method for a sliding correlation channel detection system according to claim 1, characterized in that: In step 1, the signal obtained by multiplying the PN sequence sent by the transmitter and the delayed PN sequence generated by the receiver is used as the output signal of the sliding correlation channel detection system, which can be divided into two parts: time expansion autocorrelation q c (t) and self-interference q d (t), self-interference q needs to be filtered out d (t), retaining the time-dilated autocorrelation q c (t); where the time dilation autocorrelation q c The autocorrelation function R of (t) and x(t) x (t) satisfies the following formula: Time dilated autocorrelation q c (t) is equivalent to the autocorrelation function R of x(t) x (t) dilation in the time domain. According to the relationship between the Fourier transform time domain and the frequency domain, the time dilation autocorrelation q c (t) has a compression effect in the frequency domain.

3. The PN sequence pulse waveform optimization method for a sliding correlation channel detection system according to claim 1, characterized in that: In step 2, the dynamic range DR of the time dilation autocorrelation is an indicator to measure the similarity between the time dilation autocorrelation and the Dirac function, which is the dynamic range of the time dilation autocorrelation after filtering. c (t) Main lobe peak and q c (t) sidelobes and self-interference q d (t) is the ratio of the sum of the peaks.

4. The PN sequence pulse waveform optimization method for a sliding correlation channel detection system according to claim 1, characterized in that: In step 2 and step 3, the dynamic range of the time dilation autocorrelation and the 3dB width of the time dilation autocorrelation are considered together to measure the similarity between the time dilation autocorrelation and the Dirac function, and the two performances can be given weights β according to the actual measurement scenario. DR and β 3dB To meet the scene requirements and obtain a more accurate channel impulse response.

5. The PN sequence pulse waveform optimization method for a sliding correlation channel detection system according to claim 1, characterized in that: In step 4, the pulse waveform of the PN sequence at the transmitting end can be constructed by linear superposition of PSWF. The energy of the optimized pulse waveform is concentrated at [-t0, t0] in the time domain and is limited to [-f0, f0] in the frequency domain.

6. The PN sequence pulse waveform optimization method for a sliding correlation channel detection system according to claim 1, characterized in that: In step 5, by formula: DR ×DR PSWF -β 3dB ×W 3dB,PSWF Compare the similarity between the time dilation autocorrelation of the optimized PN sequence pulse waveform and the Dirac function. The larger the value, the higher the similarity, that is, the better the detection performance of the system.

Images