How to use more than two targets using MFSK waveform?

5 次查看（过去 30 天）
LeGeND
LeGeND 2020-8-27
Hi. I am trying to simulate multiple targets range and speed estimation using fmcw radar. I have come accross MFSK (https://www.mathworks.com/help/phased/examples/simultaneous-range-and-speed-estimation-using-mfsk-waveform.html) example that correctly gives results of 2 targets. I tried to follow the procedure and changed to detection of 4 targets. But I was getting 5 range and velocity values and all were not correct. Can someone tell the issue?
So far I have changed code to modeling of 4 (instead of 2) targets, added subplots. I was getting detection of 4 targets before CA-CFAR application that means it detected 4 targets. But after it, gives 5 targets and with incorrect values.
CODES:
This is the main file running. mySetup.m and mySimulate.m are given below. I changed targets in mySetup.m to 4; and added subplots in main file.
rng(2015);
[fmcwwaveform,target,tgtmotion,channel,transmitter,receiver,...
sensormotion,c,fc,lambda,fs,maxbeatfreq] = mySetup;
Nsweep = 2;
xr = mySimulate(Nsweep,fmcwwaveform,sensormotion,tgtmotion,...
transmitter,channel,target,receiver);
dfactor = ceil(fs/maxbeatfreq)/2;
fs_d = fs/dfactor;
fbu_rng = rootmusic(decimate(xr(:,1),dfactor),2,fs_d);
fbd_rng = rootmusic(decimate(xr(:,2),dfactor),2,fs_d);
sweep_slope = fmcwwaveform.SweepBandwidth/fmcwwaveform.SweepTime;
rng_est = beat2range([fbu_rng fbd_rng;fbu_rng flipud(fbd_rng)],...
sweep_slope,c)
mfskwaveform = phased.MFSKWaveform(...
'SampleRate',151e6,...
'SweepBandwidth',150e6,...
'StepTime',2e-6,...
'StepsPerSweep',1024,...
'FrequencyOffset',-294e3,...
'OutputFormat','Sweeps',...
'NumSweeps',1);
numsamp_step = round(mfskwaveform.SampleRate*mfskwaveform.StepTime);
sig_display = mfskwaveform();
spectrogram(sig_display(1:8192),kaiser(3*numsamp_step,100),...
ceil(2*numsamp_step),linspace(0,4e6,2048),mfskwaveform.SampleRate,...
'yaxis','reassigned','minthreshold',-60)
Nsweep = 1;
release(channel);
channel.SampleRate = mfskwaveform.SampleRate;
release(receiver);
receiver.SampleRate = mfskwaveform.SampleRate;
xr = helperFMCWSimulate(Nsweep,mfskwaveform,sensormotion,tgtmotion,...
transmitter,channel,target,receiver);
x_dechirp = reshape(xr(numsamp_step:numsamp_step:end),4,[]).';
fs_dechirp = 1/(2*mfskwaveform.StepTime);
xf_dechirp = fft(x_dechirp);
num_xf_samp = size(xf_dechirp,1);
beatfreq_vec = (0:num_xf_samp-1).'/num_xf_samp*fs_dechirp;
clf;
subplot(411),plot(beatfreq_vec/1e3,abs(xf_dechirp(:,1)));grid on;
ylabel('Magnitude');
title('Frequency spectrum for sweep 1');
subplot(412),plot(beatfreq_vec/1e3,abs(xf_dechirp(:,2)));grid on;
ylabel('Magnitude');
title('Frequency spectrum for sweep 2');
xlabel('Frequency (kHz)')
subplot(413),plot(beatfreq_vec/1e3,abs(xf_dechirp(:,3)));grid on;
ylabel('Magnitude');
title('Frequency spectrum for sweep 3');
xlabel('Frequency (kHz)')
subplot(414),plot(beatfreq_vec/1e3,abs(xf_dechirp(:,4)));grid on;
ylabel('Magnitude');
title('Frequency spectrum for sweep 4');
xlabel('Frequency (kHz)')
cfar = phased.CFARDetector('ProbabilityFalseAlarm',1e-2,...
'NumTrainingCells',16);
peakidx = cfar(abs(xf_dechirp(:,1)),1:num_xf_samp);
Fbeat = beatfreq_vec(peakidx);
phi = angle(xf_dechirp(peakidx,2))-angle(xf_dechirp(peakidx,1));
sweep_slope = mfskwaveform.SweepBandwidth/...
(mfskwaveform.StepsPerSweep*mfskwaveform.StepTime);
temp = ...
[1 sweep_slope;mfskwaveform.StepTime mfskwaveform.FrequencyOffset]\...
[Fbeat phi/(2*pi)].';
r_est = c*temp(2,:)/2
v_est = lambda*temp(1,:)/(-2)
%% mySetup.m %%
function [fmcwwaveform,target,tgtmotion,channel,transmitter,receiver,...
sensormotion,c,fc,lambda,fs,fr_max] = mySetup
% System parameter
fc = 77e9; % operating frequency
c = 3e8; % propagation speed
lambda = c/fc; % wavelength
tm = 0.001; % sweep time
range_res = 1; % range resolution
bw = range2bw(range_res,c); % bandwidth
sweep_slope = bw/tm; % sweep slope
range_max = 200;
fr_max = range2beat(range_max,sweep_slope,c);
v_max = 230*1000/3600;
fd_max = speed2dop(2*v_max,lambda);
fb_max = fr_max+fd_max;
fs = max(2*fb_max,bw);
fmcwwaveform = phased.FMCWWaveform('SweepTime',tm,'SweepBandwidth',bw,...
'SampleRate',fs,'SweepDirection','Triangle');
car_dist = 50;
car_speed = 96*1000/3600;
car_rcs = db2pow(min(10*log10(car_dist)+5,20));
car2_dist = 60;
car2_speed = 69*1000/3600;
car2_rcs = db2pow(min(10*log10(car_dist)+5,20));
truck_dist = 55;
truck_speed = -70*1000/3600;
truck_rcs = db2pow(min(10*log10(truck_dist)+5,20));
truck2_dist = 65;
truck2_speed = -40*1000/3600;
truck2_rcs = db2pow(min(10*log10(truck2_dist)+5,20));
tgtpos = [[car_dist;0;0],[car2_dist;0;0],[truck_dist;0;0],[truck2_dist;0;0]];
tgtvel = [[car_speed;0;0],[car2_speed;0;0],[truck_speed;0;0],[truck2_speed;0;0]];
tgtmotion = phased.Platform('InitialPosition',tgtpos,'Velocity',tgtvel);
tgtrcs = [car_rcs,car2_rcs,truck_rcs,truck2_rcs];
target = phased.RadarTarget('MeanRCS',tgtrcs,'PropagationSpeed',c,...
'OperatingFrequency',fc);
channel = phased.FreeSpace('PropagationSpeed',c,...
'OperatingFrequency',fc,'SampleRate',fs,'TwoWayPropagation',true);
ant_aperture = 6.06e-4; % in square meter
ant_gain = aperture2gain(ant_aperture,lambda); % in dB
tx_ppower = db2pow(5)*1e-3; % in watts
tx_gain = 9+ant_gain; % in dB
rx_gain = 15+ant_gain; % in dB
rx_nf = 4.5; % in dB
transmitter = phased.Transmitter('PeakPower',tx_ppower,'Gain',tx_gain);
receiver = phased.ReceiverPreamp('Gain',rx_gain,'NoiseFigure',rx_nf,...
'SampleRate',fs);
radar_speed = 60*1000/3600;
sensormotion = phased.Platform('Velocity',[radar_speed;0;0]);
%% mySimulate.m %%
function [xr,xr_unmixed] = mySimulate(Nsweep,waveform,...
radarmotion,carmotion,transmitter,channel,cartarget,receiver)
% This function helperFMCWSimulate is only in support of FMCWExample. It
% may be removed in a future release.
% RSWEEP =
% helperFMCWSimulate(NSWEEP,WAVEFORM,RADARMOTION,CARMOTION,TRANSMITTER,
% CHANNEL,CARTARGET,RECEIVER) returns the simulated sweep train RSWEEP.
%
% The input parameters are:
% NSWEEP: number of sweeps
% WAVEFORM: waveform object
% RADARMOTION: platform object for the radar
% CARMOTION: platform object for target car
% TRANSMITTER: transmitter object
% CHANNEL: propagation channel object
% CARTARGET: target car object
% RECEIVER: receiver object
%
% The rows of RSWEEP represent fast time and its columns represent slow
% time (pulses). When the pulse transmitter uses staggered PRFs, the
% length of the fast time sequences is determined by the highest PRF.
% Copyright 2010-2016 The MathWorks, Inc.
release(waveform);
release(radarmotion);
release(transmitter);
release(receiver);
release(carmotion);
release(channel);
release(cartarget);
if isa(waveform,'phased.MFSKWaveform')
sweeptime = waveform.StepTime*waveform.StepsPerSweep;
else
sweeptime = waveform.SweepTime;
end
Nsamp = round(waveform.SampleRate*sweeptime);
xr = complex(zeros(Nsamp,Nsweep));
xr_unmixed = xr;
Ntgt = numel(cartarget.MeanRCS);
for m = 1:Nsweep
% Update radar and target positions
[radar_pos,radar_vel] = radarmotion(sweeptime);
[tgt_pos,tgt_vel] = carmotion(sweeptime);
% Transmit FMCW waveform
sig = waveform();
txsig = transmitter(sig);
% Propagate the signal and reflect off the target
rxsig = complex(zeros(Nsamp,Ntgt));
for n = 1:Ntgt
rxsig(:,n) = channel(txsig,radar_pos,tgt_pos(:,n),radar_vel,tgt_vel(:,n));
end
rxsig = cartarget(rxsig);
% Dechirp the received radar return
rxsig = receiver(sum(rxsig,2));
xd = dechirp(rxsig,sig);
xr_unmixed(:,m) = rxsig;
xr(:,m) = xd;
end

请先登录，再回答此问题。

回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Waveform Design and Signal Synthesis 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by