CN103267963A - Multi-criterion radar emission signal design method based on phase modulation - Google Patents

Abstract

The invention discloses a multi-criterion radar emission signal design method based on phase modulation. The multi-criterion radar emission signal design method based on the phase modulation mainly solves the problems that an existing technical signal is single in optimization criterion and ideal and simple in environmental model. The realization process of the multi-criterion radar emission signal design method based on the phase modulation includes the following steps: (1) respectively conducting frequency domain sampling on power spectra of a radar target, clutter and noise, (2) obtaining a power spectral density vector of an optimal signal with the maximized output signal-to-noise ratio as a criterion, (3) obtaining a power spectral density vector of the optimal signal with the maximized mutual information between the radar received echo and the target as a criterion, (4) building a multi-criterion signal optimization function in a weighting mode after conducting dimension unifying on the criterion of the maximized output signal-to-noise ratio and the criterion of the maximized mutual information between the radar received echo and the target, and (5) solving a radar phase coding signal through iteration to make the multi-criterion signal optimization function smallest. The multi-criterion radar emission signal design method based on the phase modulation takes effects of the clutter into consideration, and is more flexible in signal design, capable of meeting working requirements of a multitask radar system and favorable for promoting working performance of the whole system of a radar.

Description

Multiple criteria radar emission Design of Signal method based on the phase place modulation

Technical field

The invention belongs to the Radar Technology field, be specifically related to a kind of multiple criteria radar emission Design of Signal method based on the phase place modulation, can be used for the design and optimization of radar emission signal in the clutter environment, satisfy the job requirement of multitask radar system.

Background technology

The tradition radar is usually operated at fixing frequency range, uses fixing transmitting, and perhaps periodically transmits according to certain fixed mode, and this can not change along with the variation of radar external environment condition with regard to having limited traditional radar.Because the working environment that faces of radar is complicated day by day, traditional radar underaction that seems at aspects such as target detection, tracking and identifications.Along with the continuous development of technology such as modern digital and random waveform generation, and environment is cognitive and the support of waveform diversity scheduling theory achievement, and the signal that online design and emission and radar target are complementary has become possibility.The radar emission Design of Signal is mainly considered target type, the surrounding environment of target and the target information that needs extraction that radar need be surveyed, and the radar emission signal is designed and optimize according to certain criterion, thereby make radar system obtain good working performance.At different system task requirements, the radar emission Design of Signal has different optimization criterions.But for the radar system of multitask, single optimization criterion is no longer suitable.

At above problem, people such as J.D.Zhang have proposed a kind of multiple criteria radar emission Design of Signal method in " 5Multi-objectiveWaveformDesignforCognitive Radar; IEEECIEInternationalConference; 2011; pp:580-583 " article, multinomial criterion is unified the dimension rear weight determine the signal optimizing function, and try to achieve the radar emission signal by iteration, increase output signal-to-noise ratio and radar simultaneously and received mutual information between echo and target.But this method has been ignored the influence of clutter, and the environmental model of foundation is crossed to ideal is simple, can produce considerable influence to the matching degree of optimizing signal and target in actual environment, causes than mistake, influences the raising of radar total system serviceability.

Summary of the invention

The objective of the invention is to the deficiency at above-mentioned prior art, under the condition of considering the clutter influence, a kind of multiple criteria radar emission Design of Signal method based on the phase place modulation has been proposed, to satisfy the job requirement of multitask radar system, increase the mutual information between output letter miscellaneous noise ratio and radar reception echo and target simultaneously.

The technical thought that realizes the object of the invention is: will maximize mutual information criterion between output letter miscellaneous noise ratio criterion and maximization radar reception echo and target and unify the dimension rear weight and determine multiple criteria signal optimizing function, try to achieve the radar phase modulated signal by iteration, implementation step comprises as follows:

1) respectively the power spectrum of radar target, clutter and noise is carried out frequency domain sample with the frequency domain sample frequency f, obtain the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, the length of these three vectors is M, M=B/f, wherein, B is the radar emission signal bandwidth;

2) according to the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, SCNR designs the radar emission signal for criterion with maximization output letter miscellaneous noise ratio, try to achieve the power spectrum density vector d of Optimal Signals, the length of vectorial d is M;

3) according to the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, for criterion designs the radar emission signal, try to achieve the power spectrum density vector of Optimal Signals with the mutual information MI between maximization radar reception echo and target

Vector

Length be M;

4) according to the Optimal Signals power spectrum density vector d and the vector that obtain under the different criterions Set up multiple criteria signal optimizing function J:

$J=\mathrm{\λ}{\left|\right|F^{H}s-\mathrm{Ra}\left|\right|}_2^2+(1-\mathrm{\λ}){\left|\right|F^{H}s-\tilde{R}\mathrm{\β}\left|\right|}_2^2,$

Wherein,

Represent vectorial 2-norm square, [] ^HThe operation of expression conjugate transpose, λ represents to connect the weights of two criterions, 0≤λ≤1, R is that dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

d _mBe m element among the Optimal Signals power spectrum density vector d,

For dimension is the diagonal matrix of M * M, m diagonal element is in the matrix Wherein,

Be Optimal Signals power spectrum density vector

In m element, m=1,2 ... M, α are that length is the auxiliary vector of letter miscellaneous noise ratio of M, and β is that length is the auxiliary vector of mutual information of M, the radar phase modulated signal vector of s for finding the solution, F is that dimension is the discrete fourier transform matrix of L * M, wherein the element representation of the capable m row of l is as follows:

${\left[F\right]}_{l,m}=\frac{1}{\sqrt{m}}{e}^{\left[\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\frac{(l-1)(m-1)}{M}\right]},$

Wherein, j represents imaginary number,

L=1,2 ..., L, m=1,2 ..., M, L≤M;

5) find the solution radar phase modulated signal vector s, make multiple criteria signal optimizing function J minimum:

(5a) iteration initialize signal vector s is set at random ⁽⁰⁾

(5b) establish iterations q=1, in the q time iteration, utilize the signal vector s in the iteration q-1 time ^(q-1), try to achieve the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q)

(5c) utilize the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q), calculate the signal vector s in the iteration the q time ^(q)

(5d) error delta of the q-1 time iteration of calculating and the q time iteration, if error delta less than preset threshold ρ=0.8, then stops iteration, obtaining length is the radar phase modulated signal vector s=s of L ^(q)Otherwise, make q=q+1, repeating step (5b)-(5d) continues iteration, up to error delta less than preset threshold.

The present invention compared with prior art has the following advantages:

1) existing radar emission Design of Signal method optimization criterion is single, can only satisfy radar system mission requirements in a certain respect, in the present invention, two optimization criterions are unified the dimension rear weight obtain a multiple criteria signal optimizing function, increased the mutual information between output letter miscellaneous noise ratio and radar reception echo and target, satisfy the job requirement of multitask radar system, help to improve the integral working of radar system.

2) existing radar emission Design of Signal method is based upon comparatively under the ideal environment model mostly, because in actual radar operating environment, clutter be can not ignore, the inventive method has been considered the influence of clutter power spectrum, more realistic, strengthened the matching degree of radar emission signal and radar actual observation environment.

Description of drawings

Fig. 1 is realization flow figure of the present invention;

Fig. 2 is the sub-process figure that the present invention finds the solution the radar phase modulated signal;

Fig. 3 is in weights λ=1 o'clock, and output letter miscellaneous noise ratio SCNR is with the situation of change figure of iterations;

Fig. 4 is in weights λ=0 o'clock, receives mutual information MI between echo and target with the situation of change figure of iterations;

Fig. 5 is respectively λ=1 at weights, λ=0.3, and λ=0 o'clock utilizes the comparison diagram of the radar phase modulated signal that the inventive method obtains.

Embodiment

With reference to Fig. 1, performing step of the present invention is as follows:

Step 1, the power spectrum to radar target, clutter and noise carries out frequency domain sample respectively.

In this example, the power spectrum of supposing radar target, clutter and noise is all known, carry out frequency domain sample with the frequency domain sample frequency f, obtain the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, the length of these three vectors is M, M=B/f, wherein, B is the radar emission signal bandwidth.

Step 2, SCNR is criterion with maximization output letter miscellaneous noise ratio, tries to achieve the power spectrum density vector of Optimal Signals.

Radar transmits by optimization, and miscellaneous noise ratio SCNR is believed in maximization output, can improve the target detection performance of radar system.This step is utilized method of Lagrange multipliers, tries to achieve the power spectrum density vector d of Optimal Signals as follows:

2a) set up maximization output letter miscellaneous noise ratio SCNR criterion function:

$\max \mathrm{SCNR}=\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\frac{h_m{d}_m}{n_m+c_m{d}_m},$

Wherein, n _mBe m element in the power spectrum density vector n of radar noise, h _mBe m element among the power spectrum density vector h of radar target, c _mBe m element among the power spectrum density vector c of clutter, d _mM element among the Optimal Signals power spectrum density vector d that finds the solution for needs, m=1,2 ..., M;

2b) according to maximization output letter miscellaneous noise ratio SCNR criterion function, the value d of m element among the rated output spectral density vector d _m:

$d_m=\max [\frac{\sqrt{n_m{h}_m/\mathrm{\&gamma;}}-n_m}{c_m},0],$

Wherein, max[x, y] represent to get x, the big value between the y, γ is Lagrange multiplier, its value satisfies following formula: Wherein, E represents the radar emission signal energy;

2c) value of usefulness all elements is formed the power spectrum density vector d=[d of Optimal Signals ₁..., d _m..., d _M] ^T, wherein, [] ^TThe expression matrix transpose operation.

Step 3, the mutual information MI that receives between echo and target with the maximization radar is criterion, tries to achieve the power spectrum density vector of Optimal Signals.

The information theory technology is an important method of design radar emission signal optimizing, by the mutual information MI between maximization radar reception echo and target, can reduce the uncertainty of target response, help to improve performances such as the estimation of radar system target, identification and tracking.

This step is utilized method of Lagrange multipliers, tries to achieve the power spectrum density vector of Optimal Signals

Its step is as follows:

3a) set up the mutual information MI criterion function that maximizes between radar reception echo and target:

$\max \mathrm{MI}\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\ln (1+\frac{h_m\widetilde{d_m}}{n_m+c_m\widetilde{d_m}}),$

Wherein, ln (~) expression is the logarithm at the end with e, n _mBe m element in the power spectrum density vector n of radar noise, h _mBe m element among the power spectrum density vector h of radar target, c _mBe m element among the power spectrum density vector c of clutter,

The Optimal Signals power spectrum density vector of finding the solution for needs

In m element, m=1,2 ..., M;

3b) according to the mutual information MI criterion function between maximization radar reception echo and target, rated output spectral density vector

In the value of m element

${\tilde{d}}_m=\max [\sqrt{{\left(\frac{2c_m{n}_m+h_m{n}_m}{2(c_{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">m</figure-callout>}}^2+h_m{c}_m)}\right)}^2+\frac{n_m{h}_m/\widetilde{\mathrm{\&gamma;}}-n_m^2}{{\mathrm{<figure-callout\; id="2"\; label="c\; m"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">c</figure-callout>}}_m^{}+h_m{c}_m}}-\frac{2c_m{n}_m+h_m{n}_m}{2(c_{\mathrm{<figure-callout\; id="2"\; label="m"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">m</figure-callout>}}^2+h_m{c}_m)},0],$

Wherein,

Be Lagrange multiplier, its value satisfies following formula:

Wherein, E represents the radar emission signal energy;

3c) value of usefulness all elements is formed the power spectrum density vector of Optimal Signals

Step 4 is set up multiple criteria signal optimizing function.

This step will maximize output letter miscellaneous noise ratio SCNR criterion and maximization and receive mutual information MI criterion between echo and target and unify the dimension rear weight and obtain multiple criteria signal optimizing function, if λ is for connecting the weights of two criterions, 0≤λ≤1, multiple criteria signal optimizing function J then is expressed as follows:

$J=\mathrm{\&lambda;}{\left|\right|F^{H}s-\mathrm{Ra}\left|\right|}_2^2+(1-\mathrm{\&lambda;}){\left|\right|F^{H}s-\tilde{R}\mathrm{\&beta;}\left|\right|}_2^2,$

Wherein,

Represent vectorial 2-norm square, [] ^HThe operation of expression conjugate transpose, R is that dimension is the diagonal matrix of M * M, m diagonal element is in the matrix Wherein, d _mBe m element among the Optimal Signals power spectrum density vector d,

For dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

Wherein,

Be Optimal Signals power spectrum density vector

In m element, m=1,2,, M, α are that length is the auxiliary vector of letter miscellaneous noise ratio of M, β is that length is the auxiliary vector of mutual information of M, the radar phase modulated signal vector of s for finding the solution, F is that dimension is the discrete fourier transform matrix of L * M, the element representation of the capable m row of l is as follows in this discrete fourier transform matrix:

${\left[F\right]}_{l,m}=\frac{1}{\sqrt{m}}{e}^{\left[\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\frac{(l-1)(m-1)}{M}\right]},$

Wherein, j represents imaginary number,

L=1,2 ..., L, m=1,2 ..., M, L≤M;

When weights λ=1, represent single maximization output letter miscellaneous noise ratio SCNR; When weights λ=0, represent that single maximization receives the mutual information MI between echo and target; When 0＜λ＜1, expression maximization output letter miscellaneous noise ratio SCNR and receive mutual information MI between echo and target, according to the different task requirement of radar system, can be to the flexible value of weights λ, the optimization signal of design trade-off.

Step 5 is found the solution the radar phase modulated signal.

Optimal Signals is the ideal signal that radar system realizes optimum performance, but amplitude and the phase place of Optimal Signals all do not suffer restraints, do not satisfy the requirement of radar actual transmit signal constant amplitude, this step is by the designed phase modulation signal, make multiple criteria signal optimizing function minimization, thereby make the radar system serviceability approach optimum performance.

With reference to Fig. 2, being implemented as follows of this step:

(5a) iteration initialize signal vector s is set at random ⁽⁰⁾:

$s^{\left(0\right)}={[s_1^{\left(0\right)},s_2^{\left(0\right)},...,s_L^{\left(0\right)}]}^T,$

Wherein, initialize signal vector s ⁽⁰⁾In l element be

Be random number, expression Phase place, a represents signal amplitude, establishes a=1 in this example, l=1,2 ..., L;

(5b) establish iterations q=1, in the q time iteration, the signal vector s that utilizes q-1 iteration to obtain ^(q-1), try to achieve the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q), undertaken by following formula:

${\mathrm{\&alpha;}}^{\left(q\right)}=e^{j\arg \left[{\left[{\left(\mathrm{FR}\right)}^H\left(\mathrm{FR}\right)\right]}^{-1}{\left(\mathrm{FR}\right)}^H{s}^{(q-1)}\right]},$

${\mathrm{\&beta;}}^{\left(q\right)}=e^{j\arg \left[{\left[{\left(F\tilde{R}\right)}^H\left(F\tilde{R}\right)\right]}^{-1}{\left(F\tilde{R}\right)}^H{s}^{(q-1)}\right],}$

Wherein, arg[] expression gets phase operation, [] ^-1Representing matrix is inverted, and R is that dimension is the diagonal matrix of M * M, and m diagonal element is in the matrix

Wherein, d _mBe m element among the Optimal Signals power spectrum density vector d,

For dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

Wherein,

Be Optimal Signals power spectrum density vector

In m element, m=1,2 ..., M, F are that dimension is the discrete fourier transform matrix of L * M, the element representation of the capable m row of l is as follows in this discrete fourier transform matrix:

${\left[F\right]}_{l,m}=\frac{1}{\sqrt{m}}{e}^{\left[\mathrm{<figure-callout\; id="2"\; label="j"\; filenames="HDA00003155192300041.png"\; state="\{\{state\}\}">j</figure-callout>}2\mathrm{\&pi;}\frac{(l-1)(m-1)}{M}\right]},$

Wherein, j represents imaginary number,

L=1,2 ..., L, m=1,2 ..., M, L≤M;

(5c) utilize the auxiliary vectorial α of letter miscellaneous noise ratio ^(q)With the auxiliary vectorial β of mutual information ^(q), calculate the signal vector s in the iteration the q time ^(q):

$s^{\left(q\right)}=a{e}^{j\arg [\mathrm{\&lambda;FR}{\mathrm{\&alpha;}}^{\left(q\right)}+(1-\mathrm{\&lambda;})F\tilde{R}{\mathrm{\&beta;}}^{\left(q\right)}]},$

Wherein, a is the amplitude of signal, establishes a=1 in this example, and λ is for connecting the weights of two criterions, 0≤λ≤1;

(5d) error delta of the q-1 time iteration of calculating and the q time iteration:

δ=||s ^(q)-s ^(q-1)|| ₂，

Wherein, || || ₂The 2-norm of expression vector, if error delta less than preset threshold ρ=0.8, then stops iteration, obtaining length is the radar phase modulated signal vector s=s of L ^(q)Otherwise, make q=q+1, repeating step (5b)-(5d) continues iteration, up to error delta less than preset threshold ρ=0.8.

Effect of the present invention can further specify by following l-G simulation test:

1. simulated conditions:

Suppose radar emission signal bandwidth B=300MHz, frequency domain sample frequency f=0.5MHz, the power spectrum of radar target, clutter and noise is all known, and be Gauss's spectrum, the length of the power spectrum density vector n of the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and noise is M, M=B/f=600, length L=600 of radar phase modulated signal vector s, radar emission signal energy E=2055, Lagrange multiplier γ=0.2

2. emulation content:

Emulation 1, in weights λ=1 o'clock, output letter miscellaneous noise ratio SCNR is with the situation of change of iterations in the iterative process of diplomatic copy inventive method, and the result is as shown in Figure 3.As can be seen from Figure 3, along with the increase of iterations, output letter miscellaneous noise ratio SCNR constantly increases, and more and more approaches output letter miscellaneous noise ratio maximal value SCNR _Max

Emulation 2 in weights λ=0 o'clock, receives mutual information MI between echo and target with the situation of change of iterations in the iterative process of diplomatic copy inventive method, the result as shown in Figure 4.As can be seen from Figure 4, along with the increase of iterations, the mutual information MI that receives between echo and target constantly increases, and more and more near the mutual information maximal value MI between radar reception echo and target _Max

Emulation 3 is respectively λ=1 at weights, λ=0.3, and λ=0 o'clock utilizes the inventive method emulation to obtain the phase modulated signal of radar, as shown in Figure 5.As can be seen from Figure 5, when weights λ=0.3, the radar phase modulated signal that utilizes the inventive method to obtain is the trade-off of weights λ=1 and weights λ=0 o'clock two radar phase modulated signals.

3. weights are made as λ=1 respectively, λ=0.3, λ=0 o'clock, emulation obtains exporting letter miscellaneous noise ratio SCNR and radar and receives mutual information MI between echo and target, and the result is as shown in table 1:

Letter miscellaneous noise ratio SCNR under the different weights of table 1 and mutual information MI

?	λ=1	λ=0.3	λ=0	Maximal value
					SCNR(dB)	28.59	28.45	28.18	28.62
MI(nats)	411.8	437.5	442.4	443.5

As can be seen from Table 1, in weights λ=1 o'clock, output letter miscellaneous noise ratio SCNR is minimum with peaked difference, and along with reducing of weights λ, output believes that miscellaneous noise ratio SCNR and peaked difference are increasing; In weights λ=0 o'clock, the mutual information MI that radar receives between echo and target is minimum with peaked difference, and along with the increase of weights λ, mutual information MI and peaked difference between radar reception echo and target are increasing; In weights λ=0.3 o'clock, output letter miscellaneous noise ratio SCNR and to receive mutual information MI between echo and target be trade-off under weights λ=1 and weights λ=0 two kind of the situation.

To sum up, radar emission Design of Signal method of the present invention has been considered the influence of clutter, is applicable to multiple criteria radar emission Design of Signal, satisfies the job requirement of multitask radar system, helps to improve the serviceability of radar total system.

Claims (5)

1. the multiple criteria radar emission Design of Signal method based on the phase place modulation comprises the steps:

1) respectively the power spectrum of radar target, clutter and noise is carried out frequency domain sample with the frequency domain sample frequency f, obtain the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, the length of these three vectors is M, M=B/f, wherein, B is the radar emission signal bandwidth;

2) according to the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, SCNR designs the radar emission signal for criterion with maximization output letter miscellaneous noise ratio, try to achieve the power spectrum density vector d of Optimal Signals, the length of vectorial d is M;

3) according to the power spectrum density vector h of radar target, the power spectrum density vector c of clutter and the power spectrum density vector n of noise, for criterion designs the radar emission signal, try to achieve the power spectrum density vector of Optimal Signals with the mutual information MI between maximization radar reception echo and target

Vector

Length be M;

4) according to the Optimal Signals power spectrum density vector d and the vector that obtain under the different criterions

Set up multiple criteria signal optimizing function J:

$J=\mathrm{\&lambda;}{\left|\right|F^{H}s-\mathrm{Ra}\left|\right|}_2^2+(1-\mathrm{\&lambda;}){\left|\right|F^{H}s-\tilde{R}\mathrm{\&beta;}\left|\right|}_2^2,$

Wherein,

Represent vectorial 2-norm square, [] ^HThe operation of expression conjugate transpose, λ represents to connect the weights of two criterions, 0≤λ≤1, R is that dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

d _mBe m element among the Optimal Signals power spectrum density vector d,

For dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

Wherein,

Be Optimal Signals power spectrum density vector

In m element, m=1,2 ... M, α are that length is the auxiliary vector of letter miscellaneous noise ratio of M, and β is that length is the auxiliary vector of mutual information of M, the radar phase modulated signal vector of s for finding the solution, F is that dimension is the discrete fourier transform matrix of L * M, wherein the element representation of the capable m row of l is as follows:

${\left[F\right]}_{l,m}=\frac{1}{\sqrt{m}}{e}^{\left[j2\mathrm{\&pi;}\frac{(l-1)(m-1)}{M}\right]},$

Wherein, j represents imaginary number,

L=1,2 ..., L, m=1,2 ..., M, L≤M;

5) find the solution radar phase modulated signal vector s, make multiple criteria signal optimizing function J minimum:

(5a) iteration initialize signal vector s is set at random ⁽⁰⁾

(5b) establish iterations q=1, in the q time iteration, utilize the signal vector s in the iteration q-1 time ^(q-1), try to achieve the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q)

(5c) utilize the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q), calculate the signal vector s in the iteration the q time ^(q)

(5d) error delta of the q-1 time iteration of calculating and the q time iteration, if error delta less than preset threshold ρ=0.8, then stops iteration, obtaining length is the radar phase modulated signal vector s=s of L ^(q)Otherwise, make q=q+1, repeating step (5b)-(5d) continues iteration, up to error delta less than preset threshold.

2. method according to claim 1, wherein said step 2) with maximization output letter miscellaneous noise ratio SCNR for criterion design the radar emission signal, try to achieve the vectorial d of power spectrum density of Optimal Signals, carry out as follows:

2a) set up maximization output letter miscellaneous noise ratio SCNR criterion function:

$\max \mathrm{SCNR}=\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\frac{h_m{d}_m}{n_m+c_m{d}_m},$

Wherein, n _mBe m element in the power spectrum density vector n of radar noise, h _mBe m element among the power spectrum density vector h of radar target, c _mBe m element among the power spectrum density vector c of clutter, d _mM element among the Optimal Signals power spectrum density vector d that finds the solution for needs, m=1,2 ..., M;

2b) according to maximization output letter miscellaneous noise ratio SCNR criterion function, the value d of m element among the rated output spectral density vector d _m, undertaken by following formula:

$d_m=\max [\frac{\sqrt{n_m{h}_m/\mathrm{\&gamma;}}-n_m}{c_m},0],$

Wherein, max[x, y] represent to get x, the big value between the y, γ is Lagrange multiplier, its value satisfies following formula: Wherein, E represents the radar emission signal energy;

2c) value of usefulness all elements is formed the power spectrum density vector d=[d of Optimal Signals ₁..., d _m..., d _M] ^T, wherein, [] ^TThe expression matrix transpose operation.

3. method according to claim 1 receives mutual information MI between echo and target for criterion designs the radar emission signal with the maximization radar in the wherein said step 3), tries to achieve the power spectrum density vector of Optimal Signals

Carry out as follows:

3a) set up the mutual information MI criterion function that maximizes between radar reception echo and target:

$\max \mathrm{MI}\underset{m=1}{\overset{M}{\mathrm{\&Sigma;}}}\ln (1+\frac{h_m\widetilde{d_m}}{n_m+c_m\widetilde{d_m}}),$

Wherein, ln () expression is the logarithm at the end with e, n _mBe m element in the power spectrum density vector n of radar noise, h _mBe m element among the power spectrum density vector h of radar target, c _mBe m element among the power spectrum density vector c of clutter, The Optimal Signals power spectrum density vector of finding the solution for needs

In m element, m=1,2 ..., M;

3b) according to the mutual information MI criterion function between maximization radar reception echo and target, rated output spectral density vector

In the value of m element

Undertaken by following formula:

${\tilde{d}}_m=\max [\sqrt{{\left(\frac{2c_m{n}_m+h_m{n}_m}{2(c_m^2+h_m{c}_m)}\right)}^2+\frac{n_m{h}_m/\widetilde{\mathrm{\&gamma;}}-n_m^2}{c_m^2+h_m{c}_m}}-\frac{2c_m{n}_m+h_m{n}_m}{2(c_m^2+h_m{c}_m)},0],$

Wherein,

Be Lagrange multiplier, its value satisfies following formula:

Wherein, E represents the radar emission signal energy;

3c) value of usefulness all elements is formed the power spectrum density vector of Optimal Signals

4. method according to claim 1, the signal vector s that utilizes q-1 iteration to obtain in the wherein said step (5b) ^(q-1)Try to achieve the auxiliary vectorial α of the letter miscellaneous noise ratio in the iteration the q time ^(q)With the auxiliary vectorial β of mutual information ^(q), undertaken by following formula:

${\mathrm{\&alpha;}}^{\left(q\right)}=e^{j\arg \left[{\left[{\left(\mathrm{FR}\right)}^H\left(\mathrm{FR}\right)\right]}^{-1}{\left(\mathrm{FR}\right)}^H{s}^{(q-1)}\right]},$

${\mathrm{\&beta;}}^{\left(q\right)}=e^{j\arg \left[{\left[{\left(F\tilde{R}\right)}^H\left(F\tilde{R}\right)\right]}^{-1}{\left(F\tilde{R}\right)}^H{s}^{(q-1)}\right],}$

Wherein, arg[] expression gets phase operation, [] ^-1Representing matrix is inverted, and R is that dimension is the diagonal matrix of M * M, and m diagonal element is in the matrix

d _mBe m element among the Optimal Signals power spectrum density vector d, For dimension is the diagonal matrix of M * M, m diagonal element is in the matrix

Wherein,

Be Optimal Signals power spectrum density vector

In m element, m=1,2 ..., M, F are that dimension is the discrete fourier transform matrix of L * M, the element representation of the capable m row of this discrete fourier transform matrix l is as follows:

${\left[F\right]}_{l,m}=\frac{1}{\sqrt{m}}{e}^{\left[j2\mathrm{\&pi;}\frac{(l-1)(m-1)}{M}\right]},$

Wherein, j represents imaginary number,

L=1,2 ..., L, m=1,2 ..., M, L≤M.

5. method according to claim 1, the wherein auxiliary vectorial α of the described utilization letter of step (5c) miscellaneous noise ratio ^(q)With the auxiliary vectorial β of mutual information ^(q), calculate the signal vector s in the iteration the q time ^(q), undertaken by following formula:

$s^{\left(q\right)}=a{e}^{j\arg [\mathrm{\&lambda;FR}{\mathrm{\&alpha;}}^{\left(q\right)}+(1-\mathrm{\&lambda;})F\tilde{R}{\mathrm{\&beta;}}^{\left(q\right)}]},$

Wherein, a is the amplitude of signal, and λ is for connecting the weights of two criterions, 0≤λ≤1.

Images