A target detection method for OFDM sounding integrated signal based on filter bank

Abstract

The present invention belongs to the technical field of integrated detection and communication, and specifically relates to an OFDM sounding integrated signal target detection method based on a polyphase filter group. The present invention designs a polyphase filter group to divide and shape the echo signal into subcarriers, thereby accelerating the out-of-band attenuation rate of the OFDM signal, reducing the interference between the subband sidelobes and the carriers, and by correcting the filter group, the dependence on the orthogonality of the OFDM signal is reduced to the greatest extent, and still has good detection performance under the influence of multipath and frequency deviation. In addition, the present invention greatly reduces the amount of calculation by using a polyphase filtering structure, and can process high-speed transmitted data in real time.

Description

Technical Field

[0001] The invention belongs to the technical field of detection and communication integration, and in particular relates to an OFDM sounding integrated signal target detection method based on a polyphase filter bank. Background Art

[0002] Orthogonal Frequency Division Multiplexing (OFDM) signals have become an integrated waveform widely studied in detection and communication integrated systems due to their high spectrum efficiency, flexible modulation methods, thumbtack-shaped fuzzy functions, and ease of implementation. However, OFDM integrated signals also have some unavoidable disadvantages: the cyclic prefix used in OFDM signals to deal with multipath interference will form false targets in the range image of the detection echo; in addition, OFDM sub-band sidelobes are high and inter-carrier interference is serious, so it is very sensitive to frequency offset and has strict requirements on orthogonality.

[0003] In the integrated sounding system, complex application scenarios will lead to serious multipath effects. In order to ensure the multipath suppression capability of OFDM signals, its detection capability will be severely limited; the integrated OFDM sounding system will inevitably produce Doppler frequency deviation during operation, and multi-carrier signals are more sensitive to frequency deviation than single-carrier signals, and the Doppler frequency deviation generated in the terahertz band is also much larger than that in the ordinary frequency band. Therefore, for OFDM signals with strict orthogonality requirements, the frequency deviation will destroy the orthogonality between subcarriers, causing mutual interference between carrier frequencies, resulting in a large performance loss. Traditional OFDM detection echo processing usually uses pulse compression and two-dimensional FFT transform. Among them, pulse compression is to match the echo signal with the reference signal to obtain HRRP, and then use the coherent accumulation between pulses to obtain the target speed image; the two-dimensional fast Fourier transform uses the orthogonality between OFDM signal subcarriers to quickly solve the target parameters. However, the detection performance of these two methods will drop sharply when the orthogonality of OFDM in the terahertz frequency band is severely damaged. Therefore, it is of great significance to find a target detection method that can maintain high detection accuracy when the orthogonality of OFDM signals is destroyed. Summary of the invention

[0004] The purpose of the present invention is to provide an integrated signal target detection method based on a filter group in response to the above-mentioned problems and shortcomings. This method performs subcarrier division and shaping through multi-phase filtering, accelerates the out-of-band attenuation rate of OFDM signals, reduces sub-band side lobes, and thus reduces interference between carriers. Compared with traditional detection methods, this method obtains sub-band signals through a filter group method, so there is no need to ensure strict orthogonality between the carriers, and it can still have good detection performance under the influence of multipath and frequency deviation. It is an efficient integrated signal target detection and processing method for OFDM sounding in the terahertz frequency band.

[0005] The technical solution of the present invention is: an OFDM sounding integrated signal target detection method based on a filter bank, comprising the following steps:

[0006] Step 1: Preprocess the original signal s(τ, t) at the receiving end by mixing it with the transmitting carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(τ, t) after down-conversion processing.

[0007] Step 2: According to the transmitted baseband signal s m,n The number of subcarriers and the subcarrier spacing in (t) set the order K of the polyphase filter, and let its decimation rate be the number of subcarriers N to determine the polyphase filter bank; the baseband echo signal r(τ,t) is first extracted and then filtered by the polyphase filter bank h n (m) Divide the signal into sub-band echo signals r with the same number of transmitted carriers N m,n (t).

[0008] Step 3: Return different subbands to wave r m,n (t) corresponds to the baseband transmission subcarrier s m,n (t) is divided to obtain the channel matrix H formed under the influence of the detected target m*n , the obtained channel matrix H m*n Perform Fourier transform along the direction of the M pulses to the frequency domain to obtain the Doppler velocity image D(p), and perform peak search on the result to achieve a preliminary estimate v of the target velocity.

[0009] Step 4: Use the Doppler frequency deviation f generated by the initial estimated velocity d '=2v / λ to correct the filter bank and get h n '(m), with a new filter bank h n '(m) Re-filter the baseband echo r(τ,t) to obtain the sub-band echo signal r with the same number of transmitted carriers N m,n '(t), execute step 3 again to obtain the target final velocity estimate v.

[0010] Step 5: Correct the channel matrix H m*n'Perform inverse Fourier transform along the direction of N sub-bands to obtain the one-dimensional range image R(k) of the target, and perform peak search on the result to realize the distance estimation R of the target.

[0011] The present invention determines the echo signal of the signal in each sub-band by means of a polyphase filter group, makes a preliminary estimate of the target motion speed, corrects the filter group by using the Doppler frequency deviation generated, and processes the new filtering result to reduce the influence of the frequency deviation on the detection accuracy.

[0012] The beneficial effects of the present invention are to improve the detection performance of OFDM integrated signals in the terahertz frequency band or when the signal orthogonality is severely destroyed, and to accelerate the out-of-band attenuation rate of OFDM subcarriers by using a multi-phase filter group, reduce interference between subcarriers, and improve the operation speed. BRIEF DESCRIPTION OF THE DRAWINGS

[0013] Figure 1 is a flow chart of the present invention;

[0014] Figure 2 This is the distance estimation result diagram of point targets with the same scattering intensity;

[0015] Figure 3 This is a graph of distance estimation results for point targets with different scattering intensities;

[0016] Figure 4 This is the result of the target distance estimation under low signal-to-noise ratio;

[0017] Figure 5 The results of low-speed target detection. (a) Doppler velocity image of low-speed target; (b) range-velocity joint detection result of classic FFT algorithm; (c) range-velocity joint detection result of the algorithm proposed in this paper

[0018] Figure 6 The results of high-speed target detection. (a) Doppler velocity image of high-speed target; (b) range-velocity joint detection result of classic FFT algorithm; (c) range-velocity joint detection result of the algorithm proposed in this paper DETAILED DESCRIPTION

[0019] The present invention is described in detail below with reference to the accompanying drawings and simulation examples to demonstrate the practicability of the present invention.

[0020] As attached Figure 1 As shown, through an integrated signal target detection method based on a filter bank of the present invention, the signal echo can be subcarrier divided and shaped, the out-of-band attenuation speed of the OFDM signal can be accelerated, the interference between the sub-band side lobes and the carriers can be reduced, and good detection performance can still be achieved under the influence of multipath and frequency offset. The specific implementation steps are as follows:

[0021] Step 1: Consider that there is a moving point target at a distance R0 from the integrated system with a relative speed v to the integrated system. At time t, the radial relative distance between its scattering center and the integrated system is R(t) = R0-vt, and the resulting delay is τ = 2R(t) / c, where c is the speed of light. The original signal s(τ, t) at the receiving end is preprocessed, that is, it is mixed with the transmission carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(τ, t) after down-conversion processing, which is expressed as:

[0022]

[0023] Where r represents the backscattering coefficient of the target scattering center; B m,n is the complex input data carried by the nth subcarrier of the mth pulse; N is the number of subcarriers; △f is the interval between adjacent carriers; T p is the pulse width; T r is the pulse repetition period; f0 is the transmitting center carrier frequency, f d,n =2v(f0+n△f) / c is the Doppler information carried by the nth subcarrier.

[0024] Step 2: According to the transmission baseband signal The number of subcarriers and the subcarrier spacing in the multiphase filter are used to set the order K of the multiphase filter. The decimation rate of the multiphase filter bank is equal to the number of subcarriers N. After selecting the window function, the expression h is used. n (k) = h(kN+n) determines a polyphase filter bank, where k = 1, 2, ..., K / N, n = 1, 2, ..., N;

[0025] The baseband echo signal r(τ,t) is passed through a polyphase filter bank h that first extracts and then filters. n (m) Divide the signal into sub-band echo signals r with the same number of transmitted carriers N m,n (t), its expression is:

[0026]

[0027] Among them, r m,n (t) is the signal of the nth subband in the mth pulse after passing through the filter bank; A m,n is the amplitude attenuation and phase shift caused by the echo signal gain of the nth subcarrier in the mth pulse; n m,n (t) is the noise impact on the nth subcarrier of the integrated signal at the mth pulse.

[0028] Step 3: Bring the different sub-bands back to wave r m,n (t) corresponds to the baseband transmission subcarrier s m,n By dividing (t), we can get the expression:

[0029]

[0030] Rewriting it into matrix form gives:

[0031]

[0032] Among them, η is equal to For the same target echo, this value is a constant. α represents the influence of the relative position of the target and the integrated system on the modulation symbol, α=[0 α1 α2 ... α N-1 ], β represents the influence of Doppler information on the modulation symbol, β=[0 β1 β2 ... β M-1 ], α n and β m The specific parsing form is:

[0033] k R (n) = exp(-j2πn△fτ)

[0034] k D (m) = exp(j2πmT r f d )

[0035] The obtained channel matrix H m*n Performing Fourier transform along the direction of M pulses, that is, performing Fourier transform on β, obtains the Doppler velocity image D(p), which is expressed as:

[0036]

[0037] Performing a peak search on the result yields a preliminary estimate of the target speed v:

[0038]

[0039] Where p is the index corresponding to the peak value of the Doppler velocity image.

[0040] Step 4: Use the Doppler frequency shift f′ generated by the initial velocity estimate d =2vf0 / c Correct the filter bank to get h n '(m), with a new filter bank h n '(m) Re-filter the baseband echo r(τ,t) to obtain the sub-band echo signal r with the same number of transmitted carriers N m,n '(t), execute step 3 again to obtain the target final velocity estimate v.

[0041] Step 5: Correct the channel matrix H m*n'Perform inverse Fourier transform along the direction of N sub-bands, that is, perform inverse Fourier transform on α to obtain the one-dimensional range image R(k) of the target, which is expressed as:

[0042]

[0043] The peak value search is performed on the result to estimate the distance R of the target:

[0044]

[0045] Where k is the index corresponding to the peak value of the one-dimensional range image.

[0046] Simulation Example

[0047] According to the above method, Figure 1 The specific settings of simulation parameters for the process operation shown are as follows: the center carrier frequency of the transmitted signal is 220 GHz, the bandwidth is 1.6 GHz, the number of subcarriers is 128, and the initial radial distances between the target and the integrated system are 2 m, 2.1 m, and 8 m respectively;

[0048] Figure 2 The relative distances between the three point targets with the same scattering intensity and the integrated system are detected and estimated when the signal-to-noise ratio is 20 dB. The simulation results of the comparison between the present invention and the traditional method show that the target sidelobe of the present invention is reduced by about 5 dB compared with the traditional FFT algorithm, and the signal-to-noise ratio is improved.

[0049] Figure 3 The simulation results after changing the scattering intensity of the above three point targets are given; the scattering coefficient of target one at 2m is set to 0.2, the scattering coefficient of target two at 2.1m is set to 0.3, and the scattering coefficient of target three at 8m is set to 1. It can be seen that the present invention effectively reduces the target side lobe and solves the problem that weak targets are easily submerged by the side lobes of strong targets;

[0050] Figure 4 The simulation results of detecting the target again after reducing the signal-to-noise ratio to 5 dB are given, and it can be seen that the present invention has good anti-noise performance;

[0051] Figure 5 The simulation results of target distance-velocity joint estimation for moving point targets: target 1 (2m, 2m / s), target 2 (2.1m, 4m / s), target 3 (8m, 6m / s) are given. (a) is the Doppler velocity image of the integrated signal echo, (b) and (c) are the target distance-velocity joint detection results of the classic FFT algorithm and the present invention, respectively. It can be seen that the present invention has almost no side lobes of the distance image and can achieve more accurate distance estimation of the target.

[0052] Figure 6The simulation results of target distance-speed joint estimation are given for moving target one (2m, 22m / s), target two (2.1m, 24m / s) and target three (8m, 26m / s) after increasing the target speed. It can be seen that the present invention solves the problem of wide main lobe and blurred peak value in traditional algorithms for speed detection of high-speed moving targets, and the ability to suppress side lobes is more obvious.

[0053] In summary, judging from the processing results, the method provided by the present invention has practical value.

Claims

1. A method for OFDM sounding integrated signal target detection based on filter bank, characterized in that: The following steps are involved: S1. Consider that there is a moving point target at a distance R0 from the integrated system with a relative speed v to the integrated system. At time t, the radial relative distance between its scattering center and the integrated system is R(t)=R0-vt, and the resulting delay is t=2R(t) / c, where c is the speed of light. The original signal s(t,t) at the receiving end is preprocessed, that is, it is mixed with the transmission carrier frequency exp{j2πf0t} to obtain the integrated signal baseband echo r(t,t) after down-conversion processing, which is expressed as: Where r represents the backscattering coefficient of the target scattering center; B m,n is the complex input data carried by the nth subcarrier of the mth pulse; N is the number of subcarriers; △f is the interval between adjacent carriers; T p is the pulse width; T r is the pulse repetition period; f0 is the transmitting center carrier frequency, f d,n =2v(f0+n△f) / c is the Doppler information carried by the nth subcarrier; S2, according to the transmitted baseband signal s m,n (t) = B m,n e j2πn△f(t-mTr) The number of subcarriers and the subcarrier spacing in the multiphase filter are used to set the order K of the multiphase filter. The decimation rate of the multiphase filter bank is equal to the number of subcarriers N. After selecting the window function, the expression h is used. n (k) = h(kN+n) determines a polyphase filter bank, where k = 1, 2, ..., K / N, n = 1, 2, ..., N; The baseband echo signal r(t,t) is passed through a polyphase filter bank h that first extracts and then filters. n (m) Divide the signal into sub-band echo signals r with the same number of transmitted carriers N m,n (t), its expression is: Among them, r m,n (t) is the signal of the nth subband in the mth pulse after passing through the filter bank; A m,n is the amplitude attenuation and phase shift caused by the echo signal gain of the nth subcarrier in the mth pulse; n m,n (t) is the noise impact on the nth subcarrier of the integrated signal at the mth pulse; S3, bring different sub-bands back to wave r m,n (t) corresponds to the baseband transmission subcarrier s m,n (t) and we get the expression: Rewriting it into matrix form: Among them, η is equal to For the same target echo, this value is a constant. α represents the influence of the relative position of the target and the integrated system on the modulation symbol. α=[0α1α2…α N-1 ], β represents the influence of Doppler information on the modulation symbol, β=[0β1β2…β M-1 ], α n and β m The specific parsing form is: k R (n)=exp(-j2πn△ft) k D (m)=exp(j2πmT r f d ) The obtained channel matrix H m*n Perform Fourier transform along the direction of M pulses, that is, perform Fourier transform on β to obtain the Doppler velocity image D(p), which is expressed as: Performing a peak search on the result yields a preliminary estimate of the target speed v: Where p is the index corresponding to the peak value of the Doppler velocity image; S4, using the Doppler frequency deviation f generated by the initial estimated velocity d '=2vf0 / c to correct the filter bank and get h n '(m), with a new filter bank h n '(m) Re-filter the baseband echo r(t,t) to obtain the sub-band echo signal r with the same number of transmitted carriers N m,n '(t), execute the operation of step S3 again to obtain the target final speed estimate v; S5. Correct the channel matrix H m*n 'Perform inverse Fourier transform along the direction of N sub-bands, that is, perform inverse Fourier transform on α to obtain the one-dimensional range image R(k) of the target, which is expressed as: The peak value search is performed on the result to estimate the distance R of the target: Where k is the index corresponding to the peak value of the one-dimensional range image.

Images