Matlab雷达回波数据模拟Word文件下载.docx

1、amplitudegridNp = 20; % 20 pulsesjkl = 0:（Np-1）; % pulse index array, 慢时间采样的序列，注意第一个PRI标记为0是为了慢时间起始时刻从零开始PRF = 10.0e3; % PRF in HzPRI = （1/PRF）; % PRI in secT_0 = PRI*jkl; % relative start times of pulses, in secg = ones（1,Np）; % gains of pulsesT_out = 12 40*1e-6; % start and end times of range wind

2、ow in sec， 这个就是接收窗的时间宽度TrecT_ref = 0; % system reference time in usec， T_ref = 0指T_0=0时，r_at_T_0 = ri ；当T_0 = 0时，r_at_T_0 = ri - vi*T_0（j）fc = 10e9; % RF frequency in Hz; 10 GHz is X-bandfprintf（nWe are simulating %g pulses at an RF of %g GHz,Np,fc/1e9）nand a PRF of %g kHz, giving a PRI of %g usec.,

3、PRF/1e3,PRI/1e-6）nThe range window limits are %g to %g usec.n, . T_out（1）/1e-6,T_out（2）/1e-6）% Compute unambiguous Doppler interval in m/sec% Compute unambiguous range interval in metersvua = 3e8*PRF/（2*fc）; %第一盲速rmin = 3e8*T_out（1）/2;rmax = 3e8*T_out（2）/2;rua = 3e8/2/PRF;nThe unambiguous velocity i

4、nterval is %g m/s.,vua）nThe range window starts at %g km.,rmin/1e3）nThe range window ends at %g km.,rmax/1e3）nThe unambiguous range interval is %g km.nn,rua/1e3）% Define number of targets, then range, SNR, and% radial velocity of each. The SNR will be the actual SNR of the target in% the final data;

5、 it will not be altered by relative range.Ntargets = 4;del_R = （3e8/2）*（ 1/fs ）/1e3; % in kmranges = 2 3.8 4.4 4.4*1e3;SNR = -3 5 10 7; % dBvels = -0.4 -0.2 0.2 0.4*vua; % in m/sec% From SNR, we compute relative RCS using the idea that SNR is proportional% to RCS/R4. Students will be asked to deduce

6、 relative RCS.rel_RCS = （10.（SNR/10）.*（ranges.4）;rel_RCS = db（rel_RCS/max（rel_RCS）,powernThere are %g targets with the following parameters:,Ntargets）for i = 1:Ntargets fprintf（n range=%5.2g km, SNR=%7.3g dB, rel_RCS=%7.3g dB, vel=%9.4g m/s ranges（i）/1e3,SNR（i）,rel_RCS（i）,vels（i） ）end% Now form the

7、range bin - pulse number data mapdisp（disp（. forming signal componenty = radar（s,fs,T_0,g,T_out,T_ref,fc,ranges,SNR,vels）; % y是337-by-20的矩阵% add thermal noise with unit powerdisp（. adding noise%randn（seed,77348911）;My,Ny = size（y）;nzz = （1/sqrt（2）*（randn（My,Ny） + J*randn（My,Ny）; % 产生方差为1的复高斯白噪声y = y

8、 + nzz;%create log-normal （ground） clutter with specified C/Nand 具体原理不清楚，需要时套用此格式即可！% log-normal standard deviation for amplitude, uniform phase% Clutter is uncorrelated in range, fully correlated in pulse #disp（. creating clutterCN = 20; % clutter-to-noise ratio in first bin （dB）SDxdB = 3; % in dB

9、（this is NOT the sigma of the complete clutter）ncc=10 .（SDxdB*randn（My,Ny）/10）;ncc = ncc.*exp（ J*2*pi*rand（My,Ny） ）;% Force the power spectrum shape to be Gaussiandisp（. correlating and adding clutterG = exp（-（0:4）.2/1.0）;G = G;zeros（Ny-2*length（G）+1,1）;G（length（G）:-1:2）;for i=1:My ncc（i,:）=ifft（G.*

10、fft（ncc（i,:）;end% rescale clutter to have desired C/N ratiopcc = var（ncc（:）;ncc = sqrt（10（CN/10）/pcc）*ncc;% 10*log10（var（ncc（:）/var（nzz（:） % check actual C/N% Now weight the clutter power in range for assume R2 （beam-limited） losscweight = T_out（1）*（T_out（1） + （0:My-1）*（1/fs）.（-1）;cweight = cweight*

11、ones（1,Np）;ncc = ncc.*cweight; % var（ncc）可以看出20列clutter的方差均在30左右y = y + ncc;My,Ny=size（y）;d=（3e8/2）*（0:My-1）*（1/fs） + T_out（1）/1e3; % T_out（1）/1e3是接收窗的起始时刻plot（d,db（y,voltage）distance （km）amplitude （dB）grid% Save the data matrix in specified file.% Save the student version in the mystery file.% Also

12、 save all parameter value displays in corresponding filedata_file=file,.mat;mystery_file=file,_mys.matlisting_file=file,.liseval（save ,data_file, J T W fs s Np PRF PRI T_out fc vua rmin rmax rua Ntargets ranges vels SNR rel_RCS y）;save -v6 ,mystery_file, J T W fs s Np PRF T_out fc yfid=fopen（listing

13、_file,wfprintf（fid,rDESCRIPTION OF DATA IN FILE ,file,.mat AND _mys.matrrfprintf（fid,rPulse length = %g microsecondsr,T/1e-6）;Chirp bandwidth = %g Mhzr,W/1e6）;Sampling rate = %g Msamples/secr,fs/1e6）;rWe are simulating %g pulses at an RF of %g GHz,Np,fc/1e9）;rand a PRF of %g kHz, giving a PRI of %g

14、usec.,PRF/1e3,PRI/1e-6）;rThe range window limits are %g to %g usec.r T_out（1）/1e-6,T_out（2）/1e-6）;rThe unambiguous velocity interval is %g m/s.,vua）;rThe range window starts at %g km.,rmin/1e3）;rThe range window ends at %g km.,rmax/1e3）;rThe unambiguous range interval is %g km.rr,rua/1e3）;rThere are

15、 %g targets with the following parameters: Ntargets）; fprintf（fid,r ranges（i）/1e3,SNR（i）,rel_RCS（i）,vels（i） ）;endfclose（fid）;fprintf（nnData is in file ,data_file）nStudent data is in file ,mystery_file）nListing is in file ,listing_file,nn）_用到的函数function y = radar（ x, fs, T_0, g, T_out, T_ref, fc, r,

16、snr, v ）% RADAR simulate radar returns from a single pulse or burst of identical pulses usage: R = radar（ X, Fs, T_0, G, T_out, T_ref, Fc, R, SNR, V ） X: baseband single pulse waveform （complex vector） Fs: sampling frequency of input pulse in Hz T_0: start time（s） of input pulse（s） sec （number of pu

17、lses in burst assumed = length（g） ） G: complex gain（s） of pulse（s）， 即慢时间，各个PRI对应的脉冲的前的加权 20-by-1 T_out: 2-vector T_min,T_max defines output window delay times w.r.t. start of pulse T_ref: system reference time, needed to simulate burst returns. THIS IS THE t=0 TIME ! Fc: center freq. of the radar. R

18、: vector of ranges to target（s） meters （number of targets assumed = length（r） ） SNR: vector of target SNRs （unit noise power assumed） This will be SNR *after* allowing for R4 V: vector of target velocities （optional） in m/sec （positive velocities are towards the radar）% note（1）: VELOCITY in meters/s

19、ec ! distances in m, times in sec, BW in Hz.% note（2）: assumes each pulse is constant （complex） amplitude note（3）: will accomodate up to quadratic phase pulses note（4）: vector of ranges, R, allows DISTRIBUTED targets （c） jMcClellan 7/28/90 Modified by M. A. Richards, August 1991J = sqrt（-1）;c = 3e8;

20、 % velocity of light in m/secMx = length（x）;delta_t = 1/fs; % sampling interval （sec）t_y = T_out（1）:delta_t:T_out（2） ; % output sampling times （sec），接收窗的宽度的等间隔采样 337-by-1T_p = Mx*delta_t; % length of input pulse （sec），基带信号chirp的脉冲持续时间，即Te% Assume zero velocities （stationary targets） if no velocity%

21、vector providedif nargin 7 v = zeros（r）;end% ensure that all vectors are column vectorsx=x（: g=g（: T_0=T_0（: r=r（: snr=snr（: v=v（:% determine the quadratic phase modulation parameters for% later interpolation of pulse samplest_x = delta_t*0:（Mx-1）x_ph = unwrap（angle（x）; %基带chirp信号的相位，可以看出x_ph是个抛物线q

22、= polyfit（t_x,x_ph,2）; %目的是用 q = ax2 + bx + c 逼近x_ph% check result using correlation coefficientxfit = polyval（q,t_x）; % 看看用 q = ax2 + bx + c 拟合的xfit与x_ph的一致程度if （x_ph*xfit）/norm（x_ph）/norm（xfit） = T_out（2） | tmax = T_out（1）nEcho from target #%g at range %g km,i,ri）nis COMPLETELY OUT OF the range windownon pulse #%g.n,j） else % Figure out which sample locations in the output grid contain % reflected pulset_vals = t_y - tau; %t_y是接收窗的区间tau = 2*r_at_T_0/（c+vi） ，表示tau时刻目标位于r_at_T_0 ，tau即书上的t0 = 2R0/c % 用t_vals 而不是t_y是为了用 t_vals与 0, T_p这两个上下界相比较，可以想象若用t_y应该和 tau 和 T_p + tau比较