PULSE COMPRESSION METHOD, INCLUDING A CORRECTION OF NON-LINEARITY IN RADARSIGNALS, AND SYSTEM FOR PERFORMING THE SAME
Field of the Invention This invention relates to a method and system for correcting the image effects of signal non-linearity in radar. More particularly, though not exclusively, this invention relates to a method and system for correcting the image effects of transmit signal frequency ramp non-linearity in radar.
Background of the Invention Deramp is a known signal pulse compression technique and known signal demodulation technique. It has long been used as a method of pulse compression in airborne radar.
An acknowledged advantage of deramp has been its ability to be implemented in analogue technology. However, a known drawback with analogue circuitry is that in use, such circuitry introduces frequency-dependent signal phase shifts on account of non-linearities associated therewith, and that it can be difficult to correct for errors in signal processing attributed to such non- linearities. For example, it is typically not possible to correct for signal ramp non-linearity effects using simple wide bandwidth generators. Deramp continues to be used in the field of analogue technology because it can for example reduce data volume in high resolution, low platform velocity airborne environments. However, in more recent times, deramp is being less used as digital signalling techniques take over. This is partly because the demand for fine radar resolutions requires high compression ratios which in turn requires a degree of ramp linearity that is difficult to achieve with deramp.
To date, deramp has not been used in space radars for a number of reasons, namely: (a) high efficiency, non-linear electronics is used instead, (b) high platform velocity typically results in low pulse to echo window length ratio which is less suited to deramp, and (c) long synthetic apertures typically require
phase stability of the compressed pulse. On the other hand, interest in deramp is now resurfacing with the emergence of high resolution spotlight modes of operation, multi-mode radar, improved signal processing techniques and digital generation of arbitrary signal waveforms. Objects and Summary of the Invention
The present invention aims to overcome or at least substantially reduce some of the above-mentioned drawbacks.
It is the principal object of the present invention to provide an improved radar system and method which has a focussed imagery capability using a cheap and realisable form of signal waveform generator.
In broad terms, the present invention resides in the concept of generating signals in a radar system which depend on the characteristics of an incident transmission signal, and processing the signals in an inventive fashion which takes account of the degree of non-linearity in the transmission signal so as to produce well focussed imagery in the system.
According to a first aspect of the present invention, therefore, there is provided a method of correcting the image effects of transmit signal frequency ramp non-linearity in a radar system, said method comprising aligning in time a number of deramped radar return signals from a number of target features at various ranges, and applying a correction related to the non-linearity to each of the resultant aligned signals in time domain, the alignment of the signals being based upon the performance of a predetermined frequency-dependent shift operation in Fourier transform domain.
According to a second aspect of the present invention, there is provided a method of correcting the image effects of transmit signal frequency ramp non- linearity in a radar system, said method comprising the steps of: (a) combining a time-delayed copy of an original non-linear transmission signal with a number of deramped radar return signals representative of a number of target features at various ranges; (b) applying a first correction to the signals to take account of non-linearity effects in the original transmission signal; (c) Fourier transforming the corrected signals enabling the return signals to be separated in frequency;
(d) applying a predetermined frequency-dependent shift operation to the signals; (e) reverse Fourier transforming the resultant signals into time domain, providing an alignment in time domain of said number of radar return signals from the various range features; (f) applying a second correction to the signals in dependence upon the degree of non-linearity in the transmission signal; and (g) Fourier transforming the resultant signals so as to provide said number of corrected radar return signals in range domain.
In accordance with an exemplary embodiment of the invention which will be described hereafter in detail, the frequency-dependent shift operation comprises deriving a phase-varying function representative of a linear transmit signal, and combining said derived phase-varying function with each of the signals such as to compensate for signal errors associated with any signal phase variations.
Advantageously, the phase-varying function is a quadratic phase-varying function and said function is multiplied to each of the signals enabling errors associated therewith to be corrected. In this way, the residual video phase inherent in the use of deramp modulation can be effectively removed.
Preferably, the first correction for application to the signals comprises deriving a difference between a predetermined linear transmission signal and the non-linear transmission signal to be used, and applying a weighting factor related to the difference between said linear and non-linear signals to each of the signals.
Preferably, the second correction for application to the signals comprises analysing the characteristics of the aligned signals, comparing said analysed signal characteristics with the characteristics of the original transmission signal, fitting a model function to the signals in dependence upon said comparison and deriving a set of parameters therefrom, enabling the degree of extraneous signal distortion to be established.
Conveniently, the level of signal distortion associated with non-linearity effects in the system can be modelled using a low order polynomial function of phase. In such a model, signal characteristics such as signal frequency, rate of
change of frequency and higher derivatives thereof can be related to the derived coefficients of the model polynomial function. Further, an optional signal amplitude correction can be effected with the signal phase correction, if desired, providing a suppression/reduction of unwanted signal sidelobe effects. Conveniently, the method of the invention comprises effecting a predetermined co-ordinate transformation on the signals, for example converting signals in frequency domain to signals in range domain, or vice versa, prior to applying the frequency-dependent shift operation.
Conveniently, the signal spectrum analysis in the method of the invention is performed with digital Fast Fourier Transform (FFT) means.
The present invention further extends to a system adapted and arranged to carry out the above described method. More particularly, such a system comprises means for combining a time-delayed copy of an original non-linear transmission signal with a number of deramped radar return signals representative of a number of target features at various ranges; means for applying a first correction to the signals to take account of non-linearity effects in the original transmission signal; means for Fourier transforming the corrected signals enabling the return signals to be separated in frequency; means for applying a predetermined frequency-dependent shift operation to the signals; means for reverse Fourier transforming the resultant signals into time domain, providing an alignment in time domain of said number of radar return signals from the various range features; means for applying a second correction to the signals in dependence upon the degree of non-linearity in the transmission signal; and means for Fourier transforming the resultant signals so as to provide said number of corrected radar return signals in range domain.
Advantageously, the present invention can be embodied in hardware or software.
Further, according to another aspect of the invention there is provided a method of determining signal ramp non-linearity in a radar system, said method comprising the steps of: (a) generating a digital representation of a predetermined original linear signal, (b) generating a time-delayed copy of the
generated signal of step (a); (c) digitally combining the generated signals of steps (a) and (b); (d) converting the combined signal into analogue form and passing the analogue signal through a predetermined non-linear transmit path of the radar system; (e) mixing the resultant non-linear signal in a predetermined way using the deramp demodulation characteristics of the radar system's receive path; (f) digitally analysing various combinations of signals arising from the various signals produced in steps (a) to (e); and (g) fitting a model function to the various signals in dependence upon said digital analysis in step (f), and deriving a set of model coefficients therefrom, enabling the degree of non-linearity distortion in the system to be evaluated.
In this connection it is to be noted that the signal analysis in step (f) conveniently comprises analysing various signal products between the original transmission signal and the delayed mixer signal, and between the original mixer signal and the delayed transmission signal. It is to be appreciated that the present invention provides a cheap and simplified solution for correcting for non-linearity effects in radar and that it has utility for various applications, for example for SAR and other space-based applications.
The above and further features of the invention are set forth with particularity in the appended claims and will be described hereinafter with reference to the accompanying drawing.
Brief Description of the Drawing
Figures 1(a) to (j) schematically show various steps of a preferred method of the present invention. Detailed Description of an Exemplary Embodiment
In this specification, the term "transmit signal frequency ramp" is taken to mean a linear frequency-modulated (FM) ramp transmission signal, and the term "deramp" is used to mean removing this characteristic in the demodulation by demodulating with a signal with an inverse FM ramp.
Referring to Figure 1 , there is schematically shown various steps of a preferred method of the invention for correcting image effects of ramp non- linearity in radar.
In relation to Figure 1 , it is to be understood that the skilled man in the relevant art would readily appreciate that the steps to be described hereinafter can be implemented using conventional hardware/electronics circuitry as deployed in conventional signal demodulation techniques for example.
As shown in the Figure, the first step (step (a)) involves controllably generating a time-delayed (deramp) copy 1 of an original transmission signal and feeding back the generated signal copy 1 through a receive path (not shown) of the deployed circuitry, and also collecting via said circuitry, various deramped radar return signals 2 ("echo returns") from various target features (not shown) at various ranges. Conveniently, as shown in Figure 1 (a) of the embodiment, the echo returns 2 from the various ranges are converted to various CW tones having different, predefined range-dependent frequencies.
Significantly, the time-delayed generated signal copy in step (a) is required to be of wider bandwidth and of longer duration than the original transmission signal. This is conveniently achieved, for example, by chopping the transmission signal from a signal generator or by digitally extending the time-delayed deramp signal. Further, in the embodiment, the transmission signal is desirably a linear ramp frequency modulation (FM) waveform (typically,
of quadratic phase to better than -π over half the pulse duration) and the
8 resultant deramp FM rate is desirably an exact, phase-locked delayed copy of the transmission signal. In practical terms, it is difficult to produce a perfectly linear transmission signal and part of the inventive process, as will be described hereinafter, is thus particularly concerned with determining and correcting for any non-linearity in the transmission signal. In the embodiment, a reproducible near linear transmission signal can be readily achieved with a digital waveform generator, for example. The second step (step (b)) of the method in the embodiment involves mixing the collection of echo returns with the time-delayed (deramp) signal copy
and then digitising the resultant signal, and then, by using processor means in the circuitry, partially correcting for the known non-linearity of the regenerated signal by multiplying by the difference between the perfectly Linear FM signal (as desired) and the actual non-linear transmission signal used such as to provide the output in Figure 1(b).
Preferably, the following multi-step ((A) to (G) step) procedure is used to determine the difference in step (b) between the perfectly linear FM signal and the actual non-linear signal:
(A) Generate an ideal digital representation of the required linear FM signal;
(B) Generate a second, time-delayed exact digital copy of the original linear FM signal;
(C) Precisely digitally sum the two signals of (A) and (B);
(D) Convert the digitally-combined signal to analogue and pass the analogue signal through the non-linear transmit path of the radar;
(E) Mix the now non-linear signal with itself using the known deramp demodulation characteristics of the radar receive path;
(F) Digitally analyse the resulting signal products between the original transmission signal and the delayed mixer signal, and between the original mixer signal and the delayed transmission signal; and
(G) Fit a prescribed parameterised model function to the various signals in dependence upon the digital analysis in step (F), and extract a set of model coefficients therefrom, the derived coefficients being a measure of the non-linearity distortion in the radar system.
Thus, it is to be understood that the above described multi-step procedure for determining the transmission signal non-linearity cleverly takes advantage of two digitally combined time-displaced signals, information on the degree of non-linearity being obtained from the non-alignment in time of the respective cross signals.
The third step (step (c)) of the method in the embodiment involves
Fourier Transforming the partially corrected signal such as to separate out the various range returns in frequency. The result 5 of this operation is shown in
Figure 1 (c). This achieves an imperfect signal compression because of the effects of the non-linearity of the transmitted signal.
The fourth step (step (d)) of the method in the embodiment is optional insofar as it is not essential to the invention. As shown, the fourth step comprises performing a co-ordinate transformation on the signals, enabling the number of signals 5 in frequency domain (see Figure 1(c)) to be converted into a corresponding number of signals 6 in range domain. The effect of this transformation step is shown in Figure 1 (d). Conveniently, the transformation of step (d) is used to enhance the system's overall image formation capability.
The fifth and sixth steps (steps (e) and (f)) of the method in the embodiment comprise (i) providing a quadratic phase-varying signal function (linear ramp) 10, as defined by the ideal transmit signal waveform (see Figure 1 (e)), and then (ii) applying a deterministic frequency-dependent shift to the rate of change of the signal phase by means of multiplying the quadratic phase- varying function to each of the signals. In this way, the so-called "residual video phase" inherent in the use of deramp demodulation is effectively removed. The result of this frequency-dependent shift operation is shown in Figure 1(f). It is, however, to be appreciated that other forms of phase-varying signal function could alternatively be used to effect the same shift operation, if desired.
The seventh step (step (g)) of the method in the embodiment comprises Fourier Transforming the signals back to the original time domain ('range frequency' domain) wherein all the original signal returns from the various range features are made to be coincident in time as a result of the applied "residual video phase" correction. The resultant signal alignment 13 in time domain is shown in Figure 1 (g).
The eighth step of the method in the embodiment comprises applying a deterministic correction to each of the aligned signals, this correction being determined from the known transmission signal non-linearity.
More particularly, the deterministic correction involves analysing the aligned signal characteristics and comparing the analysed signal characteristics with the characteristics of the original (non-linear) transmission signal using conventional signal analysis techniques, and then fitting a model phase correction function to the signals in dependence upon the results of the comparison.
Preferably, as shown in Figure 1(h), the model phase correction function is a low order polynomial function of phase 15, the derived coefficients of the polynomial function being related to the signal characteristics such as signal frequency/rate of change of frequency and higher derivatives thereof. In this way, by applying (multiplying) the derived phase correction function to each of the signals, it is possible to correct simultaneously for the signal phase error in each signal return. In addition, a spectral weight (amplitude) correction can be optionally applied to the signals, if desired, enabling an effective suppression/reduction of any unwanted signal sidelobe effects. A typical form of a spectral weight (amplitude) correction 16, as can be applied to the signals, is displayed in Figure 1 (i). Note that other different types of weighting could be used instead to suppress the signal sidelobes, if desired, and that it is generally impractical to apply precise weighting to suppress the sidelobes. The final step comprises Fourier Transforming the modified signals to range domain such as to recover a now perfectly range-compressed function 20. The result 20 of this signal transformation is shown in Figure 1 (j).
Conveniently, the system/method of the invention allows signal distortion to be continuously and effectively monitored. Advantageously, the system/method of the invention is reliant only upon the generation of a replica signal (as opposed to a calibration signal). The circuitry for use in the invention is also simplified and can be implemented at reasonable cost - it could, for example, be a simple coupler round central (conventional) electronics.
Advantageously, application of the deramp technique is at the first Rx downconverter of the system to reduce bandwidth to carrier ratio in RF/baseband conversion.
Conveniently, the method/system of the invention also provides an effective mode of signal distortion determination in which the effect of leakage path can be separated and be taken into account.
It is to be appreciated that the deramp technique for use in the present invention finds utility for various space-based and ground applications, for example for the following kinds of time warping technique: (1 ) Radar-altimeter pulse compression;
(2) SPECAN SAR processing;
(3) Spotlight mode azimuth ambiguity suppression;
(4) Spotlight - stripmap co-ordinate transformation; and
(5) Squinted spotlight de-skewing. Having regard to the foregoing, the inventors have also made the following key observations:
(1 ) the deramp technique can be applied to the case of a high resolution stand-off SAR on a light aircraft, such as a UAV;
(2) Deramp is compatible for use with chirp scaling processing techniques;
(3) Deramp can be used to provide an efficient chirp scaling technique for motion compensation;
(4) Deramp allows efficient real time motion compensation and azimuth pre-sum filtering onboard a UAV, achieving high data compression with low mass and power demands;
(5) Deramp is compatible with non-linearity correction; and
(6) Deramp is ideal for application of high compression encoding schemes.
Having thus described the present invention by reference to a preferred embodiment, it is to be appreciated that the embodiment is in all respects exemplary and that modifications and variations are possible without departure from the spirit and scope of the invention. For example, the precision of the embodiment could possibly be improved, if desired, by using different, predetermined kinds of correction function to take account of the signal distortion effects in the radar system.
Further it is to be appreciated that while an intermediary co-ordinate transformation step is included in the described embodiment, it is equally possible for this step to be omitted or to be replaced by a different combination of intermediary co-ordinate transformation steps such as to provide the same inventive technical effect, if desired.
Furthermore, it is also to be appreciated that while the signals in the embodiment are pulsed, it is equally possible for the signals to be continuous emissions over prolonged periods.
It is also to be appreciated that the present invention finds utility for various space-based applications and ground applications - for example, the signals are typically of the low power, short range kind in ground penetration applications.