CN102508264B - Method for inhibiting pulse interference of satellite navigation system - Google Patents

Abstract

The invention discloses a method for inhibiting pulse interference of a satellite navigation system. The method comprises the following steps: primarily inhibiting the interference by a time-domain and frequency-domain mixed filtering method to obtain an initial estimate value of a position signal of the pulse interference; replacing distorted data with the initial estimate value; and estimating a wave shape parameter of a satellite signal and the distorted data by using all data which is subjected to interference inhibition so as to obtain accurate estimate values of the parameter of the satellite signal and the distorted data. In the method, properties of an interference source of civil airborne navigation equipment and actual application problems are taken full account, the interference is inhibited by a signal processing method, the distorted data in the satellite signal is restored, and the availability of the satellite signal is ensured on the premise that the interference is inhibited, so that the flying safety coefficient can be increased and the system has a positive meaning for implementation of a future GNSS (global navigation satellite system)-based civil aviation navigation system.

Description

A kind of pulse interference suppression method for satellite navigation system

Technical field

The invention belongs to the Satellite Navigation Technique field, but particularly relate to a kind of pulse interference suppression method that is used for satellite navigation system of enhanced navigation satellite receiver availability of data.

Background technology

The prime navaids facilities of following civil aviation are GLONASS (Global Navigation Satellite System) (GNSS, Global Navigation Satellite System), and wherein gps system is the important component part of GNSS system.Gps system is constantly being accelerated modernization, and one of them important component part is exactly the setting of L5 signal.The L5 signal aims at civil navigation and arranges, and takies special-purpose frequency range.The L5 signal has higher received energy with respect to L1, L2 signal, higher bearing accuracy, and stronger antijamming capability, and be easier to realize that it will consist of the important component part of following GNSS together with the Beidou satellite navigation system of China.

But many systems have been arranged in work because L5 signal place centre frequency is the frequency range of 1176.54MHz, so GPS L5 receiver faces complicated electromagnetic environment.Near the main existing system of working the frequency range of L5 signal place has aeronautical radio system and various radar system, in these systems, DME (Distance Measurement Equipment stadimeter) equipment is the principal element that affects GPS L5 signal.The impulse disturbances of DME equipment can reduce the signal to noise ratio (S/N ratio) of gps signal, and the receiver tracking ring can not be restrained, and causes the decode procedure difficulty.The present U.S.'s 1000 DME earth stations of having an appointment, by in June, 2009, China's DME earth station has 337 covers, mainly is distributed in region of Southeast.DME equipment is except causing the interference GPS L5 signal, also can cause interference to the B2a signal with the Big Dipper two generations navigational system of L5 common center frequency.

The researcher is studied for the DME interference problem both at home and abroad, and has developed various Anti-Jamming Techniques, and wherein time domain pulse blanking method is the most frequently used method.This method is will be above the direct zero setting of the pulse interference signal of predetermined threshold, to reach the purpose of eliminating impulse disturbances in time domain.Also there is the researchist to utilize time domain impulse disturbances technology for eliminating to suppress DME and disturbs, and carried out experimental verification at GPS L5 hardware receiver.The advantage of this method is that treatment effeciency is high, system architecture is simple, shortcoming is to cause losing or can not using of certain section satellite-signal data, forms gap data, when breach is more, data can produce serious distortion, and can have influence on catching, follow the tracks of and locating of receiver.

The mixing inhibition method of time-frequency domain is at first by a sliding window with signal, then carry out the time domain pulse detection, if the amplitude of signal surpasses certain particular value on the noise level, data slot in the window is transformed into frequency domain disturbs inhibition, and then convert back time domain, replace raw data as output with this segment data.Although the method can keep more signal energy than time domain pulse blanking method, the method also can be with the useful signal filtering of a part, thereby also can cause data distortion, thereby affects integrity and the availability of satellite-signal data.

Summary of the invention

In order to address the above problem, the object of the present invention is to provide a kind of impulse disturbances that can effectively suppress gps satellite signal, can recover satellite-signal distortion data again simultaneously, thus the pulse interference suppression method that is used for satellite navigation system of assurance gps signal availability.

In order to achieve the above object, the pulse interference suppression method for satellite navigation system provided by the invention comprises the following step that carries out in order:

(1) with time-frequency domain mixed filtering method preliminary Inhibitory signal impulse disturbances, and with the data after suppressing to disturb as the alternative distortion data of the initial value of distorted portion data, the satellite-signal data of tentatively being revised;

(2) utilize above-mentioned revised satellite-signal data estimation satellite-signal waveform parameter;

(3) the satellite-signal waveform parameter that obtains with above-mentioned estimation reappraises the data of distorted portion, obtains the valuation of distorted portion data, and substitutes original distortion data with this valuation, obtains again revised whole satellite data;

(4) repeating step (2) and (3) are until algorithm convergence.

Described satellite-signal waveform parameter comprises amplitude and phase place.

Pulse interference suppression method for satellite navigation system provided by the invention fully takes into account interference source characteristic and the actual application problem of civil aviaton's Airplane Navigation Equipment, adopt signal processing method to suppress to disturb and the distortion data in the satellite-signal is recovered, when suppressing interference, guaranteed the availability of satellite-signal, therefore can strengthen the flight safety coefficient, the enforcement based on the civil aviation navigational system of GNSS in future is had positive meaning.

Description of drawings

Fig. 1 is the pulse interference suppression method pulse interference suppression for satellite navigation system provided by the invention and distortion data Recovery processing process flow diagram.

Fig. 2 is the pulse interference suppression method distortion data recovery algorithms process flow diagram for satellite navigation system provided by the invention.

Fig. 3 is the satellite-signal oscillogram that disturbed by the DME device pulse.

Fig. 4 is signal and undisturbed raw satellite signal and the mixed filtering method result comparison of wave shape figure after the inventive method is processed.

Fig. 5 is the inventive method and time-frequency domain mixed filtering method result performance comparison figure.

Embodiment

With specific embodiment the pulse interference suppression method for satellite navigation system provided by the invention is elaborated with reference to the accompanying drawings.

As shown in Figure 1 and Figure 2, the pulse interference suppression method for satellite navigation system provided by the invention comprises the following step that carries out in order:

(1) with time-frequency domain mixed filtering method preliminary Inhibitory signal impulse disturbances, and with the data after suppressing to disturb as the alternative distortion data of the initial value of distorted portion data, the satellite-signal data of tentatively being revised.

Consider the satellite-signal that comprises interference and noise, its model can be expressed as:

y(t)＝s(t)+J(t)+e(t) (1)

Wherein, s (t) is undisturbed satellite-signal, and J (t) is pulse interference signal, and e (t) is noise.GPS L5 signal place frequency is being caused in the equipment of interference, and DME equipment is main interference source.In the DME system, airborne inquisitor per second sends 5-150 pulse pair, and pulse peak power is from 50W to 2kW.DME equipment will make signal to noise ratio (S/N ratio) reduce to the interference of L5 receiver, and the receiver tracking ring can not be restrained, thereby cause the decode procedure difficulty.Therefore need to disturb DME and suppress, and distortion data is recovered.

Signal sampling is processed to intermediate frequency, and signal can be expressed as:

y(n)＝s(n)+J(n)+e(n)，n＝1，…，N (2)

Wherein, s (n) is undisturbed satellite-signal, and J (n) is disturbed by satellite-signal, and e (n) is noise.

Mixed filtering pulse interference suppression method is that input signal is at first passed through a time domain moving window, then carries out the time domain pulse detection.When detecting the DME impulse disturbances, the data slot in the window of impulse disturbances position is transformed into frequency domain, interference is carried out being transformed into time domain behind the frequency domain filtering again, substitute with this segment data and disturb distortion data as output.Data before suppressing to disturb can be expressed as:

y＝[y ₀ … y _m-1 y _m y _n+1 … y _m+P-1 y _m+P … y _N-1] (3)

Wherein being subjected to impulse disturbances to cause the data of distorted portion is μ=[y _mY _M+P-1] _{1 * P}, the data of this part need to be estimated with restoration and reconstruction.The data of undisturbed part are:

γ＝[y ₀ y ₁ … y _m-1 y _m+P … y _N-1] _1×(N-P) (4)

After disturbing, preliminary inhibition can obtain the value of distorted portion data

Substitute distortion data, the satellite-signal of tentatively being revised with this initial value that is worth as distortion data

(2) utilize above-mentioned revised satellite-signal data estimation satellite-signal waveform parameter.

For the i time iteration (i 〉=1), order

Represent whole satellite-signals (for the iteration first time,

Be the satellite-signal of the preliminary correction that obtains in the step (1)).Theoretical according to discrete Fourier transformation, but the discrete-time signal approximate representation is the stack of the sinusoidal signal of a series of different frequencies:

$y_n=\underset{k=1}{\overset{K}{\mathrm{\Σ}}}\left[\mathrm{\α}\left({\mathrm{\ω}}_k\right)e^{j{\mathrm{\ω}}_{k}n}\right]+e\left(n\right),n=0,...,N-1,{\mathrm{\ω}}_k\∈\left[\mathrm{0,2}\mathrm{\π}\right)---\left(5\right)$

Wherein e (n) is the discrepance except satellite-signal, k=1 ..., K.In the present embodiment, the K value is 2N.

Order

(l=0 wherein ..., L-1, L=N-M+1) and the data vector of expression M * 1 dimension forward direction sliding overlapping.

For the waveform parameter of estimated signal, for each frequency-of-interest, the wave filter of a M tap of design

(wherein i represents iteration the i time), make the output after signal passes through wave filter

On the least square meaning as far as possible near the sinusoidal signal of frequency-of-interest and constant amplitude.Estimated signal range parameter and bank of filters coefficient

Method be exactly to find the solution following optimization problem:

$\underset{{\mathrm{\α}}^i\left({\mathrm{\ω}}_k\right),h^i\left({\mathrm{\ω}}_k\right)}{\min }\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{|{\left[h^i\left({\mathrm{\ω}}_k\right)\right]}^H{\stackrel{\‾}{y}}_l^i-{\mathrm{\α}}^i\left({\mathrm{\ω}}_k\right)e^{j{\mathrm{\ω}}_{k}l}|}^2,s.t.{\left[h^i\left({\mathrm{\ω}}_k\right)\right]}^{H}a\left({\mathrm{\ω}}_k\right)=1---\left(6\right)$

Wherein $a\left({\mathrm{\ω}}_k\right)={\left[\begin{array}{cccc}1& e^{j{\mathrm{\ω}}_k}& ...& e^{j{\mathrm{\ω}}_k(M-1)}\end{array}\right]}^T,k=1,...,K.$

The below finds the solution this optimization problem.Definition

Expression The normalization Fourier transform:

${\stackrel{\‾}{g}}^i\left({\mathrm{\ω}}_k\right)=\frac{1}{L}\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{\stackrel{\‾}{y}}_l^i{e}^{-j{\mathrm{\ω}}_{k}l}---\left(7\right)$

And definition:

${\hat{R}}^i=\frac{1}{L}\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{\stackrel{\‾}{y}}_l^i{\left({\stackrel{\‾}{y}}_l^i\right)}^H---\left(8\right)$

Order

Find the solution this optimization problem, can obtain the coefficient of bank of filters and the estimation of signal waveform parameter:

${\hat{h}}^i\left({\mathrm{\ω}}_k\right)=\frac{{\left({\hat{S}}^i\right)}^{-1}\left({\mathrm{\ω}}_k\right)a\left({\mathrm{\ω}}_k\right)}{a^H\left({\mathrm{\ω}}_k\right){\left({\hat{S}}^i\right)}^{-1}\left({\mathrm{\ω}}_k\right)a\left({\mathrm{\ω}}_k\right)},$ ${\widehat{\mathrm{\α}}}^i\left({\mathrm{\ω}}_k\right)=\frac{a^H\left({\mathrm{\ω}}_k\right){\left({\hat{S}}^i\right)}^{-1}\left({\mathrm{\ω}}_k\right){\hat{g}}^i\left({\mathrm{\ω}}_k\right)}{a^H\left({\mathrm{\ω}}_k\right){\left({\hat{S}}^i\right)}^{-1}\left({\mathrm{\ω}}_k\right)a\left({\mathrm{\ω}}_k\right)}---\left(9\right)$

Wherein

Be plural number, estimated signal amplitude is

Phase place is

(3) the satellite-signal waveform parameter that obtains with above-mentioned estimation reappraises the data of distorted portion, obtains the valuation of distorted portion data, and substitutes original distortion data with this valuation, obtains again revised whole satellite data.

The satellite-signal waveform parameter valuation that utilizes the i time iteration obtains Come the distortion estimator data vector

To the estimation of distortion data based on following thought: the spectrum signature that is disturbed the spectrum signature of the satellite-signal data that cause distortion and undisturbed data is basically identical.Estimation to distortion data is namely asked least square solution to following cost function:

$\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}\underset{l=0}{\overset{L-1}{\mathrm{\Σ}}}{|{\left[{\hat{h}}^i\left({\mathrm{\ω}}_k\right)\right]}^H{\stackrel{\‾}{y}}_l^i-{\widehat{\mathrm{\α}}}^i\left({\mathrm{\ω}}_k\right)e^{j{\mathrm{\ω}}_{k}l}|}^2---\left(10\right)$

Order:

${\mathrm{\η}}^i\left({\mathrm{\ω}}_k\right)={\widehat{\mathrm{\α}}}^i\left({\mathrm{\ω}}_k\right){\left[\begin{array}{c}1\\ e^{j{\mathrm{\ω}}_k}\\ .\\ .\\ .\\ e^{j{\mathrm{\ω}}_k(L-1)}\end{array}\right]}_{L\×1}---\left(12\right)$

Then the objective function of above-mentioned least square problem can be written as matrix form:

$\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{\left|\right|H^i\left({\mathrm{\ω}}_k\right)\left[\begin{array}{c}{y}_0^i\\ .\\ .\\ .\\ y_{N-1}^i\end{array}\right]-{\mathrm{\η}}_i\left({\mathrm{\ω}}_k\right)\left|\right|}^2---\left(13\right)$

Make A ⁱ=H ⁱ(:, [1:m-1, m+P:N]), B ⁱ=H ⁱ(:, m:m+P-1) (press the expression way among the Matlab), then have following formula to set up:

$H^i\left({\mathrm{\ω}}_k\right)\left[\begin{array}{c}{y}_0^i\\ .\\ .\\ .\\ y_{N-1}^i\end{array}\right]=A^i\left({\mathrm{\ω}}_k\right){\mathrm{\γ}}^i+B^i\left({\mathrm{\ω}}_k\right){\mathrm{\μ}}^i---\left(14\right)$

With seasonal d ⁱ(ω _k)=η ⁱ(ω _k)-A ⁱ(ω _k) γ ⁱ, then least square problem can be written as:

$\underset{{\mathrm{\μ}}^i}{\min }\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{\left|\right|B^i\left({\mathrm{\ω}}_k\right){\mathrm{\μ}}^i-d^i\left({\mathrm{\ω}}_k\right)\left|\right|}^2---\left(15\right)$

Find the solution this least square problem, can obtain the valuation of distorted portion data vector:

${\widehat{\mathrm{\μ}}}^i={\left(\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{\left[B^i\left({\mathrm{\ω}}_k\right)\right]}^H{B}^i\left({\mathrm{\ω}}_k\right)\right)}^{-1}\left(\underset{k=0}{\overset{K-1}{\mathrm{\Σ}}}{\left[B^i\left({\mathrm{\ω}}_k\right)\right]}^H{d}^i\left({\mathrm{\ω}}_k\right)\right)---\left(16\right)$

The distortion data vector that obtains with the i time iterative estimate

Fill back former data, the estimated value of the satellite-signal of again being revised

(4) return step (2), carry out iteration the i+1 time, repeating step (2) and (3) are until algorithm convergence.The criterion of convergence is that the cost function error of formula (15) twice iteration is less than preset value.

Used signal is the intermediate frequency emulation satellite-signal of frequency 4.309MHz in the present embodiment, and sampling rate is 6MHz, and signal to noise ratio (S/N ratio) is-18dB.Fig. 3 is the satellite-signal oscillogram that disturbed by the DME device pulse, as shown in Figure 3, made an uproar than (JNR with disturbing the dried of signal after inhibition and data are recovered by the satellite-signal of DME impulse disturbances, Jammer-to-Noise Ratio) is 20dB, the high level cadre make an uproar than the DME impulse disturbances under, receiver can't be worked.After utilizing the inventive method that satellite-signal is processed, can recover the data of distorted portion.Fig. 4 is signal and undisturbed raw satellite signal and the mixed filtering method result comparison of wave shape figure after the inventive method is processed, and as can be seen from the figure the data of distorted portion are restored and rebuild.Fig. 5 is the inventive method and time-frequency domain mixed filtering method result performance comparison figure.Data after utilizing the inventive method to process and the root-mean-square error (RMSE) of raw data are as evaluation index, under different interference-to-noise ratio conditions, carry out many experiments, can find out, after the inventive method is processed distortion data, compare the mixed filtering method closer to raw data, so the inventive method is compared the mixed filtering method and is had better handling property.

Claims (2)

1. pulse interference suppression method that is used for satellite navigation system is characterized in that described pulse interference suppression method for satellite navigation system comprises the following step that carries out in order:

(1) with time-frequency domain mixed filtering method preliminary Inhibitory signal impulse disturbances, and with the data after suppressing to disturb as the alternative distortion data of the initial value of distorted portion data, the satellite-signal data of tentatively being revised;

(2) utilize revised satellite-signal data estimation satellite-signal waveform parameter;

(3) the satellite-signal waveform parameter that obtains with above-mentioned estimation reappraises the distorted portion data, obtains the valuation of distorted portion data, and substitutes original distortion data with this valuation, obtains again revised whole satellite data;

(4) repeating step (2) and step (3) are until algorithm convergence.

2. the pulse interference suppression method for satellite navigation system according to claim 1, it is characterized in that: described satellite-signal waveform parameter comprises amplitude and phase place.

Images