GB2521098A

GB2521098A - High-resolution radar

Info

Publication number: GB2521098A
Application number: GB0724152.4A
Authority: GB
Inventors: Raymond Bailey
Original assignee: Thales Holdings UK PLC
Current assignee: Thales Holdings UK PLC
Priority date: 2007-12-06
Filing date: 2007-12-06
Publication date: 2015-06-17
Anticipated expiration: 2027-12-06
Also published as: GB0724152D0; GB2521098B

Abstract

A processing system for high-resolution stretch-compression radar, in which a ramp signal of steadily varying frequency is transmitted and a similar, corresponding de-ramp signal mixed with the radar return signal received, the de-ramped signal is then down converted by at least one IF stage; the system comprising means for receiving demodulated digital radar return signals following the de-ramping and down conversion to video frequency of signals received by the radar; means for storing data representative of corrections for unwanted artefacts in the ramp and de-ramp signals and in the transmit-receive path due to hardware in the radar system including the antenna and microwave transmission parts; and a processor arranged to correct the demodulated input radar return signals by first applying the stored corrections for de-ramp signal artefacts, then removing residual video phase between the different frequency components in the composite radar return signal corresponding to different target ranges, and then applying the stored corrections for the transmitted ramp signal and transmit-receive path artefacts.

Description

I

111GM-RESOLUTION RADAR This invention relates to high-resolution stretch-compression radar and in particular to a processing system for such a radar and to a method of operating the radar and of calibrating the radar.

Very high-resoli4ion image forming radar systems requirethe use of very high bandwidth signals. A method of achieving very high time-bandwidth products within a single pulse is called "stretch-compression". However, residual errors within the signals used may result in a poor impulse response, which is detrimental to the final image quality.

The purpose of the present invention is to overcome the problem of poor impulse response in such high-resolution radar systcms.

One method for overcoming this problem would be to exclude the possibility of errors from the radar system but this would be prohibitively expensive, especially over the wIde range of operating environments expected, for example, of military radar.

Accordingly, the present invention provides a processing system for high-resolution stretch-compression radar, in which a ramp signal of steadily varying frequency is transmitted and a similar, corresponding de-ramp signal mixed with the radar return signal received, the de-ramped signal is then down converted by at least one IF stage; the system comprising: means for receiving demodulated digital radar return signals following the de-ramping and down conversion to video frequency of signals received by the radar; means for storing data representative of corrections for unwanted artefacts in the ramp and de-raxnp signals and in the transmit-receive path due to hardware in the radar system including the antenna and microwave transmission parts; and a processor arranged to correct the demodulated input radar return signals by first applying the stored corrections for de-rainp signal artefacts, then removing residual video phase between the different frequency components in the composite radar return signal corresponding to different target ranges, sd then applying the stored corrections for the transmitted ramp signal and transmit-receive path artefacts.

Preferably, the processor is arranged to carry out the corrections in the time domain and then to transform the corrected radar return signal into the frequency domain to provide an output for display.

The invention also provides a method of correcting radar return signals in high resolution stretch-compression radar, in which a ramp signal of steadily varying frequency is transmitted and a similar, corresponding de-ramp signal is mixed with the radar return signal received, the de-ramped signal then being down-converted by at least one IF stage; the method comprising receiving demodulated digital radar return signals following the de-ramping and down-conversion to video frequency of signals received by the radar; and correcting the demodulated input radar return signals by first applying stored data representative of corrections for unwanted artefacts in the ramp and de-ramp signals, and then removing residual video phase between the different frequency components in the composite radar return signal corresponding to different target ranges, and then applying stored data representative of corrections for the transmitted ramp signal and artefacts in the transmit-receive path due to hardware in the radar system including the antenna and microwave transmission paths.

The present invention overcomes the problem of poor impulse response by managing the errors inherent in the radar system. In accordance with the invention, these inherent errors are quantified so that corrective action is applied after the event, i.e. on radar return signals.

To explain further the natute of the problem, the stretch-compression technique requires the generation of very accurate linear frequency ramps. Periodic phase modulation brings about range sidelobes, and a tolerable magnitude of residual phase errors is a maximum of about 0.06 radian. However, 3 to 4 radians of phase distui±ance can be created within the ramp generating and amplification process in a typical high-resolution stretch-compression radar system. In addition, phase droop across the pulse gives rise to geolocation errors. Finally, the inherent group delay functions within the filters employed in the radar system generate quadratic and higher order waveform modulation, which broadens the range point-spread-fimction of the compressed waveform. To compound this problem, all these errors change as a function of temperature of the hardware.

Use of the present invention can substantially mitigate these problen s by correcting for the errors at appropriate stages in the signal processing.

In order that the invehtion maybe better understood, a preferred embodiment will now be described, by way of example only, with reference to the accompanying diagrammatic drawings in which: Figure 1 is a block diagram of a conventional high-resolution stretch-compression radar system; Figure 2 i a diagram illustrating graphically5 by plotting frequency against time or range, the de-ramp process used in the radar system of Figure 1; Figure 3 shows eight waveforms in sequence resulting from the operation of a preferred embodiment of the present invention; Figure 4 is a diagram illustrating video de-skew processing i.e., video phase removal from the radar return signals in the preferred embodiment; and Figure 5 is a diagram illustrating the operation of the preferred embodiment of the invention, in terms both of calibrating the radar system and operating the calibrated system.

A conventional high-resolution stretch-compression radar system will first be described with reference to Figures 1 and 2. An X-band waveform generator generates a microwave signal ramping from 8.805 to 10.125 0Hz, and this signal Tx is sent to a transmitter which sends microwaves through a duplexer to an antenna. Signals received in the antenna pass through the duplexer and are received in the band 8.805 to 10.125 0Hz. These received signals are combined in a de-rarnp mixer with a de-ramp waveform varying in frequency from 6.843 to 8.287 0Hz. This de-ramp waveform is generated bymixing a de-rarnp waveform varying in frequency from 8.743 to 10.187 0Hz in a mixer with a root sinusoid signal of 1.9 0Hz and then passing the mixed signal through a filter to select the lower side band.

The de-ramped signal is down converted in a primary IF filter with a bandwidth of 124 MHz and a frequency of 1.9 0Hz, and the down converted signal is mixed with a root sinusoid signal of 1.8 0Hz for further down conversion to a second IF with a centre frequency of 100 MHz. It is this further down converted signal which is processed in a filter and analogue to digital converter block which produces the output radar return signal.

Thus in this example the ramp bandwidth is 1.32 GHz and the receiver bandwidthis 124 MHz.

Amplitude and phase errors ae introduced at all stages of the process. Of particular relevance to the present invention is the generation of errors in the transmitted ramp signal defined by the X-band waveform generator which has its own sources of error; and also the further distortions or artefacts introduced by the transmitter, the duplexer, the antenfla and the various connecting waveguides and cables. These further distortions, apart from the transmitted ramped signal errors, are referred to as transmit-receive path errors or artefacts.

The conventional stretch-compression process is described for example in "Introduction to Airborne Radar" by George W. Stimson, Scitech Publishing, Inc., 1998, at page 166, and is also illustrated in Figure of the accompanying drawings.

The received stretch waveform from multiple targets at different ranges is mixed with a dc-ramp waveform. This results in different fixed frequency signals corresponding to targets at different ranges. In this example, two signals at 10 MHz and 100 MHz are created. The frequency of the received and de-ramped scattered signals is directly proportional to the range from the start of the swath and the slope of the signal ramp.

These fixed frequency signals are displaced in time with respect to each other, and each has amplitude and phase errors associated with the transmit ramp.

-A preferred embodiment of the invention will now be described with reference to Figures 3 to 5. In the example shown in Figure 3, the radar system is operated in a hypothetical situation assuming only two scatterers at a range of 20 km separated by 500m.

With reference to Figure 3, waveform WI is the nearer scatterer return, with transmitted errors, and with a bandwidth of 1.3 GHz. It is the first scatterer arriving at the radar. The step fimctions on the signal represent the errors acquired during the creation of the transmitted waveform. These errors are from the waveform generator, the high power amplifier and the reflections created by the radome.

The signal from the second scatterer is shown as W2. This is delayed relative to the first scatterer by the product of the propagation rate and twice the different in range.

Waveform W3 shows the actual signal being offered to the de-ramp mixer in the radar.

It is the composite scatterer returns received by the antenna and processed in a low noise amplifier (not shown) between the antenna duplexer and the de-rarnp mixer. The signal at this point has a bandwidth of 1.32 GHz, and its bandwidth is next reduced by mixing it with the de-ramp signal as shown in Figures 1 and 2.

Waveform W4 shows the composite scatterer returns at the output of the de-ramp mixer. Additional phase errors have been introduced by the de-ramp process. It is this signal that, after further down conversion, is digitised and plated in memory accessible by a post-processing system. This signal has both range-and non range-dependent errors, which both have to be corrected.

Waveform W5 shows the scatterer returns after reference ramp error removal.

This process will be described below with reference to Figure 5. The errors imposed during the de-ramp process are removed by complex number division, using the error coefficients that are stored in the processing system.

Next, a video phase removal dispersive filter is applied to the radar return signal.

This is described below with reference to Figures 4 and 5. Because each scatterer has a different frequency, all the signals are made to shift in time so that they align in the time domain. The result is waveform W6.

The third main correction process is the removal of transmitted errors, i.e. errors introduced in the transmit-receive path, and the result of this process is shown as waveform W7. Waveform W7 shows the signal now comprising two sinusoids, one from each scatterer. The removal of the transmitted errors is done by applying stQred coefficients, described below with reference to Figure 5. Finally, the waveform W7 is transformed, as shown in waveform W8, from the time domain to the frequency domain, and the two scatterers appear clearly as compressed returns. The relationship between the frequency and spatial domains is one of simple scaling and unit conversion.

Because these two scatterers which are represented by individual frequencies have passed through the IF amplifier, further phase and amplitude distortion has occurred. In the final image, this appears as a spatial effect. This is corrected by multiplying the waveform W8 by the inverse of the amplifier amplitude and phase responses, and this is described below with reference to Figure 5.

The radar apparatus embodying the invention differs from the basic radar architecture shown in Figure. 1 by virtue of signal processing carried out in the filter and ADC block of Figure 1, and this is shown in Figure 5. The radar apparatus is operable in a number of different modes, and in particular it may be calibrated in one mode and operated to generate radar returns, in at least one other mode.

The processor shown in Figure 5 has internal memory arranged to store calibration data or correction data, which is used in the normal operation of the radar to correct the radar return signals. The memory stores samples of the reference ramp signal; IF filter correction coefficients; samples taken from the transmit signal Tx; and exo-calibration loop coefficients for each transmit mode and for each receive mode.

Reference ramp error removal and calibration will first be described. The basic requirement for reference ramp calibration is to record the primary transmit and receive signals that are intended to be used during the data collection process in the air. These data are then used within the post data collection processing, either directly, within a matched filter, or as extracted error terms in the form of coefficients. Although the error signals of interest occupy a low bandwidth of several hundred kHz, they modulate a wide band ramp of some 1.4 GHz, yet the system bandwidth through to the digitizer is only 124 MHz. This bandwidth limitation is overcome by recording the waveforms of interest in sections of 124 M}Iz, or "ramplets". Thus the wide band ramp data are collected in sections or pieces, and the data for each of these ramplets are concatenated to produce the fhll ramp data set. As shown in the top row of Figure 5, fifleen reference ramplets are taken as samples and are stored in memory. The processor dc-ramps each ramplet to baseband, applies predetermined phase and amplitude corrections for the IF amplifier involved in the down conversion, and stitches the ramplets together by phase matching adjacent ramplets at overlaps and the splicing coefficient sections into one contiguous data set. The combined data are then filtered to remove artefacts and the resulting data are stored for the process, in the operational mode, of removing reference ramp phase errors. As shown in the lower part of Figure 5, sampled radar data are converted from real numbers to complex numbers, down converted and IQ demodulated, before the refereàce ramp phase errors are removed.

In the operational mode, following removal of the reference ramp phase errors, the processor corrects for video phase, as illustrated in Figure 4, using video phase removal filter coefficients calculated to support the radar range of the returns. A matched filter programmed with these filter coefficients processes the video to correct the video phase differences.

The correction for transmit-receive path errors will now be described, as this is the next sequential stage of error correction in the normal operational mode. As shown in the third, fourth and fifth rows from the top in Figure 5, fifteen transmitted signal ramp samples are taken from the memory and are separately de-ramped to reduce them to baseband; these signals are then corrected for phase and amplitude to compensate for the aftefacts of the IF amplifier responsible for the down conversion. Adjacent ramplets are phase matched at the overlaps and are spliced into one contiguous data set which is then filtered to remove artefacts. In a parallel processing path, exo-calibration loop coefficients for each transmit and each receive mode are retrieved from memory and are selected and combined according to the radar mode. Filters are applied to these combined data to remove artefacts. The filtered data are then combined with the filtered data derived from the transmit ramp samples to provide transmit-receive path coefficients which are used in the removal of transmit-receive path and amplitude errors from the radar return signal output from the video phase removal filter.

The exo-calibration loop coefficients relate to the duplexer, the antdnna and the radome primarily: these components are outside the calibration loop, but they affect the impulse performance of the system. In particular the antenna and the radorne are significant in that there are spatial dimensions of the same order as the resolution of the radar.

As shown in the bottom right-hand section of Figure 5, the next stage in the signal processor for radar returns is to apply a zero range phase rate off-set, by calculating a mean phase rate from the time domain data, and applying a stored calibration zero loop back phase rate to determine the zero range phase rate off-set to be applied. Taylor windows are then applied, using pre-stored Taylor window coefficients, to suppress range sidelobes The resulting signal is then transformed to the frequency or spatial domain, depending on the mode of operation, as illustrated in the change from waveform W7 to W8 in Figure 3.

Pre-stored IF filter correction coefficients are interpolated using temperatUre information from the recorded data, and are then applied in the next correction process, to correct for IF phase and amplitude errors. The output from this process block is complex range output data. These are used to generate the radar display.

It will be understood that the present invention is applicable to different types of radar, and moreover that the frequency ranges and the nature of the artefacts will differ from system to system. The way in which the artefacts are corrected may also be accomplished differently, whether using filter coefflcients or stored algorithms or data concordances such as look-up tables. Separate processes maybe provided for the calibration mode, i.e., for deriving the stored correction data or algorithms.

The ramp signal is usually an X-band waveform, and its bandwidth is preferably between 1 and 5 0Hz, preferably centred on a frequency of between 5 and 115 0Hz.

The present invention is particularly useflul in synthetic aperture radar, but it is applicable to other types of radar as well.

Claims

CLAIMS: 1. A processing system for high-resolution stretch-compression radar, in which a ramp signal of steadily varying frequency is transmitted and a similar, corresponding dc-ramp signal mixed with the radar return signal received, the de-rainped signal is then down converted by at least one IF stage; the system comprising: means for receiving demodulated digital radar return signals following the de-ramping and down conversion to video frequency of signals received by the radar; means for storing data representative of corrections for unwanted artefacts in the ramp and de-ramp signals and in the transmit-receive path due to hardware in the radar system including the antenna and microwave transmission parts; and a processor arranged to correct the demodulated input radar return signals by first applying the stored corrections for de-ramp signal artefacts, then removing residual video phase between the different frequency components in the composite radar return signal corresponding to different target ranges, and then applying the stored corrections for the transmitted ramp signal and transmit-receive path artefacts.
2. A processing system according to Claim 1, in which the processor is arranged to carry out the corrections iii the time domain and then to transform the corrected radar return signal into the frequency domain to provide an output for display.
3. A processor system according to Claim 1 or Claim 2, in which the storing means is arranged to store IF filter correction coefficients as a function of different temperatures, and the processor is arranged to apply these correction coefficients to the radar return signals after the correction for transmit-receive path artefacts.
4. A processing system according to any preceding claim, in which the ramp signal is an X-band wavefonn.
5. A processing system according to Claim 4, in which the ramp signal has a bandwidth of between 1 and 5 GHz.
6. A processing system according to Claim 7, in which the ramp signal is centred on a frequency of between S and 15 GHz.
7. A processing system according to any preceding claim, in which the storing means stores the said data representative of corrections for artefacts.
8. A high resolution stretch-dompression radar, comprising a waveform generator for generating a ramp signal of steadily varying frequency and a transmitter for transmitting the ramp signal, and a receive path for mixing a dc-ramp signal derived from the ramp signal with a radar return signal received in an antenna, the de-ramped signal then being down converted by at least one IF stage; and comprising a processing system according to any preceding claim, connected to receive the down-converted radar return signals.
9. A processing system substantially as described herein with reference to the accompanying drawings.
10. Radar apparatus substantially as described herein with reference to the accompanying drawings.
11. A method of correcting radar return signals in high resolution stretch-compression radar, in which a ramp signal of steadily varying frequency is transmitted and a similar, corresponding de-ramp signal is mixed with the radar return signal received, the de-ramped signal then being down-converted by at least one IF stage; the method comprising receiving demodulated digital radar return signals following the de-ramping and down-conversIon to video frequency of signals received by the radar; and correcting the demodulated input radar return signals by first applying stored data representative of corrections for unwanted artefacts in the ramp and dc-ramp signals, and then removing residual video phase between the different frequency components in the composite radar return signal corresponding to different target ranges, and then applying stored data representative of corrections for the transmitted ramp signal and artefacts in the transmitreceive path due to hardware in the radar system including the antenna and microwave transmission paths.
12. A method according to Claim 11, comprising carrying out the said corrections in the time domain and then transforming tie corrected radar return signals into the frequency domain to display them on a display.
13. A method of calibrating a processing system according to Claim 9, comprising sampling signals from the radar in operation and storing corresponding correction data including calibration coefficients in the storing means.
14. A method of operating a radar, substantially as described herein with reference to the accompanying drawings.
15. A method of calibrating a radar, substantially as described herein with reference to the accompanying drawings.