|On this page…|
The phased.Transmitter object enables you to model key components of the radar equation including the peak transmit power, the transmit gain, and a system loss factor. You can use phased.Transmitter together with radareqpow, radareqrng, and radareqsnr, to relate the received echo power to your transmitter specifications.
While the preceding functionality is important in applications dependent on amplitude such as signal detectability, Doppler processing depends on the phase of the complex envelope. In order to accurately estimate the radial velocity of moving targets, it is important that the radar operates in either a fully coherent or pseudo-coherent mode. In the fully coherent, or coherent on transmit, mode, the phase of the transmitted pulses is constant. Constant phase provides you with a reference to detect Doppler shifts.
A transmitter that applies a random phase to each pulse creates phase noise that can obscure Doppler shifts. If the components of the radar do not enable you to maintain constant phase, you can create a pseudo-coherent, or coherent on receive radar by keeping a record of the random phase errors introduced by the transmitter. The receiver can correct for these errors by modulation of the complex envelope. The phased.Transmitter object enables you to model both coherent on transmit and coherent on receive behavior.
The transmitter object has the following modifiable properties:
PeakPower — Peak transmit power in watts
Gain — Transmit gain in decibels
LossFactor — Loss factor in decibels
InUseOutputPort — Track transmitter's status. Setting this property to true outputs a vector of 1s and 0s indicating when transmitter is on and off. In a monostatic radar, the transmitter and receiver cannot operate simultaneously.
CoherentOnTransmit — Preserve coherence among transmitter pulses. Setting this property to true (the default) models the operation of a fully coherent transmitter where the pulse-to-pulse phase is constant. Setting this property to false introduces random phase noise from pulse to pulse and models the operation of a non-coherent transmitter.
PhaseNoiseOutputPort — Output the random pulse phases introduced by non-coherent operation of the transmitter. This property only applies if the CoherentOnTransmit property is false. By keeping a record of the random pulse phases, you can create a pseudo-coherent, or coherent on receive radar.
Construct a transmitter with a peak transmit power of 1000 watts, a transmit gain of 20 decibels (dB), and a loss factor of 0 dB. Set the InUseOutPutPort property to true to record the transmitter's status.
htx = phased.Transmitter('PeakPower',1e3,'Gain',20,... 'LossFactor',0,'InUseOutputPort',true)
Construct a pulse waveform for transmission. In this example, use a 100-microsecond linear FM pulse with a bandwidth of 200 kHz. Use the default sweep direction and sample rate. Set the PRF to 2 kHz.
hpuls = phased.LinearFMWaveform('PulseWidth',100e-6,'PRF',2e3,... 'SweepBandwidth',2e5,'OutputFormat','Pulses','NumPulses',1);
Obtain the pulse waveform using the step method of the waveform object. Transmit the waveform using the step method of the transmitter object, hpuls. The output is one pulse repetition interval because the NumPulses property of the waveform object is equal to 1. The pulse waveform values are scaled based on the peak transmit power and the ratio of the transmitter gain to loss factor. The scaling factor is sqrt(htx.PeakPower*db2pow(htx.Gain-htx.LossFactor)).
wf = step(hpuls); [txoutput,txstatus] = step(htx,wf); t = unigrid(0,1/hpuls.SampleRate,1/hpuls.PRF,'[)'); subplot(211) plot(t,real(txoutput)); axis tight; grid on; ylabel('Amplitude'); title('Transmitter Output (real part) - One PRI'); subplot(212) plot(t,txstatus); axis([0 t(end) 0 1.5]); xlabel('Seconds'); grid on; ylabel('Off-On Status'); set(gca,'ytick',[0 1]); title('Transmitter Status');
To model a coherent on receive radar, you can set the CoherentOnTransmit property to false and the PhaseNoiseOutputPort property to true. You can output the random phase added to each sample with step.
To illustrate this process, the following example uses a rectangular pulse waveform with five pulses. A random phase is added to each sample of the waveform. Compute the phase of the output waveform and compare the phase to the phase noise returned by the step method.
For convenience, set the gain of the transmitter to 0 dB, the peak power to 1 W, and seed the random number generator to ensure reproducible results.
hrect = phased.RectangularWaveform('NumPulses',5); htx = phased.Transmitter('CoherentOnTransmit',false,... 'PhaseNoiseOutputPort',true,'Gain',0,'PeakPower',1,... 'SeedSource','Property','Seed',1000); wf = step(hrect); [txtoutput,phnoise] = step(htx,wf); phdeg = radtodeg(phnoise); phdeg(phdeg>180)= phdeg(phdeg>180)-360; plot(wf); title('Input Waveform'); axis([0 length(wf) 0 1.5]); ylabel('Amplitude'); grid on; figure; subplot(2,1,1) plot(radtodeg(atan2(imag(txtoutput),real(txtoutput)))) title('Phase of the Output'); ylabel('Degrees'); axis([0 length(wf) -180 180]); grid on; subplot(2,1,2) plot(phdeg); title('Phase Noise'); ylabel('Degrees'); axis([0 length(wf) -180 180]); grid on;
The first figure shows the waveform. The phase of each pulse at the input to the transmitter is zero. In the second figure, the top plot shows the phase of the transmitter output waveform. The bottom plot shows the phase added to each sample. Focus on the first 100 samples. The pulse waveform is equal to 1 for samples 1–50 and 0 for samples 51–100. The added random phase is a constant –124.7 degrees for samples 1–100, but this affects the output only when the pulse waveform is nonzero. In the output waveform, you see that the output waveform has a phase of –124.7 degrees for samples 1–50 and 0 for 51–100. Examining the transmitter output and phase noise for samples where the input waveform is nonzero, you see that the phase output of step and the phase of the transmitter output agree.