System and method for reducing the effect of a radar interference signal

Abstract

A system and method are provided to reduce the effect of an interfering signal in a radar return signal for a frequency modulated continuous wave (FMCW) radar. Once the interfering signal is detected, an extent of the interfering signal is determined and the data that was corrupted by the interfering signal is not included in the processing of the radar return signal. This allows the radar to detect a target in the presence of the interfering signal. The system and method can benefit any FMCW radar that is within the range of an interfering radar source (e.g. another FMCW radar, a police radar gun, a pulse radar, etc.) operating in the same frequency band as the FMCW radar. An alternative arrangement provides a system and method for determining the frequency of the interfering signal and then avoiding transmitting power in that portion of the frequency spectrum where the interfering signal is present.

Description

STATEMENTS REGARDING FEDERALLY SPONSORED RESEARCH

Not applicable.

FIELD OF THE INVENTION

This invention relates generally to radar systems and methods and, more particularly, to a radar system and method that can reduce or eliminate an interfering signal that may be present in a radar return signal.

BACKGROUND OF THE INVENTION

As is known in the art, a frequency modulated continuous wave (FMCW) radar transmits an RF signal (also referred to herein as a radar signal) at a transmitted frequency that is continuously changing. In order to identify a range to a target, the FMCW radar measures a difference in frequency between a received radar signal, which is returned as an echo from the target, and the transmitted frequency. The difference in frequency is associated with a time delay between the transmitted signal and the received signal, i.e., a time that it takes the transmitted signal to reach the target and to return back to the radar.

In typical FMCW radar, for example, the frequency of the transmitted FMCW signal linearly increases from a first predetermined frequency to a second predetermined frequency in a so-called “chirp” signal. The chirp signal is often repeated at a repetition rate. FMCW radar has the advantages of high sensitivity, relatively low transmitter power, and good range resolution. In one conventional FMCW radar, the chirp signal varies substantially linearly from approximately 24.05 GHz to approximately 24.25 GHz.

A conventional FMCW radar uses a mixer, which mixes (i.e., multiplies) the transmitted and received signals. One of the outputs of the mixer is the above-described difference in frequency between the transmitted and received signals, which is also referred to herein as a “downconverted signal” or a “video signal,” which can have a “beat frequency.” The downconverted signal occurs at a frequency substantially lower than the frequency of the transmitted or received signals. The downconverted signal can be time sampled, for example, with an analog-to-digital (A/D) converter, and the time samples can be converted to the frequency domain, for example, with a fast Fourier transform (FFT), to provide a frequency spectrum. From the frequency spectrum, a variety of techniques can be used to identify a range to the target. Some such techniques are described in U.S. Pat. No. 6,577,269, issued Jun. 10, 2003.

It will be appreciated that the frequency spectrum contains not only a frequency of the downconverted signal, which corresponds to range to the target, but also contains noise. The noise is associated with a variety of noise sources including, but not limited to, electrical (i.e., thermal) noise sources and radar signal noise sources that may be present in the environment in which the FMCW radar is used. It will also be appreciated that there can be more than one target in a field of view of the radar system. Therefore, the time samples of the downconverted signal can include more than one beat frequency.

In order to locate a range to the target from the frequency spectrum, a frequency signal within the frequency spectrum is identified, the frequency of which is indicative of a range to the target. However, some types of interfering radar signals can greatly degrade the ability to find the frequency signal associated with the target within the frequency spectrum. For example, an interfering radar signal at sufficiently high power level and within the swept band (i.e., within the chirp frequency limits) of the FMCW radar can corrupt the time samples of the downconverted signal to such an extent that the resulting frequency spectrum is overwhelmed by the interfering signal, and therefore, the frequency signal associated with the target cannot be found in the frequency spectrum.

Referring now to FIG. 1, a graph 10 has a horizontal axis in units of frequency provided as FFT frequency bins and a vertical axis in units of dB in FFT counts (provided by FFT processing of time samples of the downconverted signal). A curve 12 has a peak 14 indicative of a frequency, f1, and a corresponding range to a target. The curve 12 also has a noise background 16. A curve 18 has no distinct peak that is clearly characteristic of a target. The curve 18 is indicative of the output of the FFT frequency domain processing when a received signal represented by the curve 12 also includes an interfering signal.

It will be appreciated that, even where the interfering signal is at a single frequency, the resulting processing of the FMCW radar system, including the above-described mixing, and the above-described FFT processing, results in a smearing of the single interfering signal frequency throughout the frequency spectrum. This is particularly true when the mixing process provides a mixing output signal corresponding to the interfering signal for only a brief time. An apparent increase in the noise level across some or all of the frequency spectrum reduces the signal to noise ratio (SNR) of the peak 14 (i.e., of the target) and greatly reduces the probability of detection of the FMCW radar. In the curve 18, either the peak 14 cannot be found, or the peak 14 cannot be accurately found.

One particular application of the FMCW radar is in an automotive radar system, for example, an automotive radar system used to detect an object in a blind spot next to a vehicle. Automobile radars often use the above-described frequency chirp extending, for example, from approximately 24.05 GHz to approximately 24.25 GHz. Conventional police radars used, for example, to detect speed of vehicles, operates within this band, for example, at approximately 24.197 GHz. In automobile applications, it is necessary to provide a radar system capable of accurately and reliably detecting objects, e.g., other vehicles, with minimal influence from interfering signals.

Accuracy and reliability of the automotive radar system are very important. Characteristics of the automotive radar system that contribute to accuracy and reliability include susceptibility of the radar system to noise, including interfering signals, and the overall precision with which received radio frequency (RF) signals are processed in the presence of the noise and interfering signals to detect objects. Susceptibility to noise, including interfering signals, can cause an automotive radar system to falsely detect an object (i.e., to increase a false alarm rate), and/or, can cause the vehicle radar system to miss a detection of an object (i.e., to reduce a probability of detection).

SUMMARY OF THE INVENTION

In accordance with an aspect of the present invention, a method of radar processing includes receiving an RF received signal and detecting the presence of an interfering signal within the RF return signal. The method further includes selecting a sweep range to avoid the interfering signal and transmitting a frequency modulated continuous wave (FMCW) RF signal having the sweep range.

In accordance with another aspect of the present invention, a method of radar processing includes generating time samples associated with an RF return signal, generating an array having array values associated with the time samples, generating at least one threshold associated with the array values, and comparing the array values with the at least one threshold.

In accordance with another aspect of the present invention, a radar system includes an RF receiving module adapted to receive an RF received signal and adapted to provide a downconverted signal in response to the RF return signal. The system further includes an RF signal sampling module coupled to the RF receiving module and adapted to generate time samples associated with the downconverted signal. The system further includes an interference detector coupled to receive the time samples, adapted to detect an interfering signal, and adapted to select a sweep range to avoid the interfering signal. The system further includes an RF transmitting module adapted to transmit a frequency modulated continuous wave (FMCW) signal having the frequency sweep.

BRIEF DESCRIPTION OF THE DRAWINGS

The foregoing features of this invention, as well as the invention itself, may be more fully understood from the following description of the drawings in which:

FIG. 1 is a graph of an FMCW radar downconverted signal in the frequency domain, shown both without the presence of an interfering signal and with the presence of an interfering signal;

FIG. 2 is a block diagram of an FMCW radar system adapted to reduce an interfering signal in accordance with the present invention;

FIG. 3 is a graph showing maximum expected signal slope versus target size associated with time samples of a downconverted signal provided by the FMCW radar system of FIG. 2;

FIG. 4 is a graph showing the absolute value of the slope associated with time samples of a downconverted signal provided by the FMCW radar system of FIG. 2 in the presence of an interfering signal;

FIG. 4A is a graph showing time samples of a downconverted signal provided by the FMCW radar system of FIG. 2 in the presence of an interfering signal;

FIG. 5 is a graph showing time samples of a downconverted signal provided by the FMCW radar system of FIG. 2 in which the interfering signal is removed;

FIG. 6 is another graph of an FMCW radar system downconverted signal in the frequency domain, shown both without the presence of an interfering signal and with the presence of an interfering signal, shown both with and without the system and methods of the present invention;

FIG. 7 is a graph showing time samples of a downconverted signal in the presence of an interfering signal and an associated weighting function provided by an alternate FMCW radar system according to FIG. 2;

FIG. 8 is a flow chart showing a process for removing an interfering signal;

FIG. 9 is a flow chart showing further details of the process of FIG. 8;

FIG. 10 is a flow chart showing yet further details of the process of FIG. 8;

FIG. 10A is a flow chart showing further details of the process of FIG. 8; and

FIG. 10B is a flow chart showing alternate details of the process of FIG. 8.

DETAILED DESCRIPTION OF THE INVENTION

Before describing the radar system and method of the present invention, some introductory terms and concepts are discussed. As used herein, the term “frequency modulated continuous wave (FMCW) radar system” is used to describe a particular type of radar system that transmits an FMCW signal, also referred to herein as a “chirp” signal, that changes with time from an initial frequency to a final frequency. The FMCW radar system receives and processes a return signal from a target, which also has the chirp characteristics. As used herein, the terms “downconverted signal” and “video signal” are used to describe an output of a mixer circuit used in a receive portion of the FMCW radar system. The downconverted signal is representative of a frequency difference between the transmitted radar signal and the return signal from the target. Where a return signal substantially from but one target is received, the downconverted signal can have a “beat frequency” indicative of a largest amplitude frequency generated in the mixing process. As used herein, the term “radar signal” is used to describe a radio frequency (RF) signal transmitted by or received by a radar system.

Embodiments described below process time samples of a downconverted radar return signal to generate a slope (first derivative) array having slope values, which is used to detect and to determine an extent of an interfering signal. The embodiments described below show slope values to be absolute values. However, as used herein, the term “slope values” refers either to values associated with absolute value of slope or associated with non-absolute value of slope. As described below, some particular slope values are associated with an upper and a lower slope sample limit, and an upper and a lower slope sample extent (e.g., in FIG. 4). While slope is used in the illustrative embodiments below, it should be appreciated that other array types could be used to detect and to determine the extent of the interfering signal. For example, in other embodiment, the time samples can be processed to generate a power array having power values, resulting instead in an upper and a lower power sample limit, and an upper and a lower power sample extent. Other alternate array types are further described below.

Referring now to FIG. 2, an FMCW radar system 50 includes an FMCW transmitting system 64 coupled to a transmitting antenna 54. The FMCW radar system 50 also includes an FMCW receiving system 66 coupled to a receiving antenna 56. The FMCW receiving system 66 includes a radar receiving module 68 to provide a downconverted signal 73 and a radar signal sampling module 74 (analog-to-digital (A/D) converter) to provide time samples 76 associated with the downconverted signal 73. The radar system 50 also includes an interference detector 78 (having an array processor 78a) coupled to receive the time samples 76 and to process the time samples 76 to identify an interfering signal. The radar system 50 also includes an interference extent processor 88 (having a limit processor 88a) to determine an extent of the interfering signal and an interference removal processor 92 (having one or both of a selection processor 92a and a weighting processor 92b) to remove the interfering signal from the time samples associated with the radar return signal, providing a processed signal 94.

In operation, the FMCW radar system 50 generates a radar chirp 58 via a transmitting antenna 54, which echoes from a target 52, returning a target echo 60 (also referred to herein as a return signal), which is received by a receiving antenna 56. The receiving antenna 56 can be the same antenna or a different antenna from the transmitting antenna 54. An interferer 53 can be present, which generates an interfering radar signal 62 also received by the receiving antenna.

Both the target echo 60 and the interfering radar signal 62 are downconverted by a mixer 70, providing the downconverted signal 73. The downconverted signal is sampled by the A/D converter 74, providing the time samples 76 to the interference detector 78. The time samples 76 can include an interfering signal associated with the interfering radar signal 62.

Operation of the interference detector 78 is described in greater detail in conjunction with FIGS. 4, 8 and 9. However, in operation, in one particular embodiment, the interference detector 78 processes the time samples 76 to detect an interfering signal within the time samples 76 by generating a slope array having slope array values associated with the time samples and by comparing the slope array values with at least one threshold.

Operation of the interference extent processor 88 is described in greater detail in conjunction with FIGS. 4A, 8, and 10. However, in operation, in one particular embodiment, the interference extent processor 88 identifies a lower slope sample limit and an upper slope sample limit associated with slope array values that exceed the at least one threshold. The interference extent processor 88 expands these limits by using sample guard band values to provide the interfering signal extent (e.g., time extent).

Operation of the interference removal processor 92 is described in greater detail in conjunction with FIGS. 5, 8, 10, and 10A. However, the interference removal processor can select a contiguous group of time samples within the time samples 76, not including time samples within the interfering signal extent identified by the interference extent processor 88. In one particular embodiment, the interference removal processor 92 can pad the selected time samples, for example, zero pad the selected time samples. Essentially, the contiguous group of time samples is the time samples that will not be removed.

Where the interference detector 78 does detect an interfering signal, the radar return signal processor 82 operates on a signal 94. However, where the interference detector 78 does not detect an interfering signal, the radar return signal processor 82 operates on a signal 80, which can be the same as the time samples 76.

The signal 94 from which the interfering signal is removed, or alternatively, the signal 80 having no interfering signal, is processed, for example, by the radar return signal processor 82, to generate a radar system detection and/or classification of a target, represented by signal 84.

In some embodiments, the FMCW transmitting system 64 is adapted to generate a radar frequency sweep over a fixed sweep range, for example, from approximately 24.05 GHz to approximately 24.25 GHz. However, in other embodiments, the FMCW transmitting system 64 is adapted to generate a radar frequency sweep over a reduced sweep range responsive to an interfering signal. The reduced sweep range is discussed more fully below.

Referring now to FIG. 3, a graph 100 has a horizontal scale in units of dBsm (decibels re: 1 square meter), corresponding to a radar cross section of a variety of target sizes. A vertical scale corresponds to a maximum expected slope associated with time samples of a downconverted signal (e.g., time samples 76, FIG. 2). A curve 102 is indicative of the maximum expected slope for different targets having different radar cross sections at any range within predetermined range limits.

Slope absolute value associated with the time samples 76 (FIG. 2) can be provided for each time sample j=1:N−1 (N=number of time samples associated with a chirp) as:slope(j)=|time sample(j+1)−time sample(j)|It will be recognized that slope(j) is an array having array values, and more particularly, a slope array having slope values, each associated with the time samples 76. The number of slope values in slope(j) can be the same as or different from the number of time samples 76 (FIG. 2). In one particular embodiment, the number of slope values in slopes) is less than the number of time samples 76.

The curve 102 can be generated in the following way. A maximum frequency associated with the time samples (corresponding to a maximum target range) can be computed as follows.

$f_{\max }=\frac{2·R_{\max }·B}{c·\Delta t}$ where:

R_max=maximum range;

c=speed of light;

Δt=frequency sweep time;

f_max=maximum received frequency; and

B=bandwidth of the frequency sweep.

Conversely, a range that corresponds to each of the frequencies less than f_max can be calculated as follows.

$R=\frac{c·\Delta t·f}{2·B}$

A free space range loss at each of the ranges (i.e., frequencies) can be calculated as follows:

$\mathrm{rangeloss}\left(f\right)=\frac{1}{{\left(4\pi \right)}^2·R^4}$

A frequency dependent gain of a receiving amplifier, ampgain(f), for example, a receiving amplifier associated with the down converter 68 of FIG. 2, is system dependent. In some embodiments, the gain of the receiving amplifier, ampgain(f), is selected to reverse compensate for the above free space range loss.

A maximum expected slope associated with the time samples can be computed over all possible frequencies (i.e., ranges) as follows:maximum expected slope(j)=|RangeFreqSlope(j)rangeloss(j)ampgain(j)RCS|where j=1 to the highest frequency (FFT bin number). RangeFreqSlope(j) is a maximum slope of a received signal at a range corresponding to the jth range based on the system parameters of sample rate and sample time. The frequency used to find the jth element of RangeFreqSlope(j) is found by replacing R_max in the above equation (for f_max) with the value of the jth range. RCS is a radar cross section of the target.

The above equation for maximum expected slope(j) provides a maximum expected slope array having maximum expected slope values. Any one of the maximum expected slope values, representative of slope at the variety of ranges, can be a maximum slope value, maxslopevalue, associated with a given radar cross section shown in the curve 102. It will be appreciated that the maxslopevalue can occur for the target at any range (i.e., for any frequency) due to dependence upon the amplifier gain, ampgain(f), and is not necessarily at the closest range.

It will become apparent below that the maxslopevalue selected from the array, maximum expected slope(j), can be used as a threshold against which slope array values can be compared.

Referring now to FIG. 4, a graph 120 includes a horizontal scale in units of slope sample number, wherein the slope sample number is associated with the time samples 76 (FIG. 2) according to the above equation for slope(j). In one particular embodiment, there are two hundred fifty five slope samples (and two hundred fifty six time samples). However, in other embodiments the number of slope samples can be less than or greater than two hundred fifty five and the number of time samples can be greater than or less than two hundred fifty six. The graph includes a vertical scale in units of slope absolute value in accordance with the equation for slope(j) given above.

A curve 126 is representative of slope absolute values associated with the time samples 76 (FIG. 2) according to the above equation for slope(j). The curve 126 connects slope samples, slope(j), which are not individually shown. The curve 126 includes curve portions 126a, 126b representative of the time samples 76 having relatively small slopes and a curve portion 126c representative of the time samples 76 having a relatively high slope. It will become apparent than the curve portion 126c is indicative of an interfering signal.

A first threshold 122, referred to herein as a maximum slope threshold, corresponds to the above-identified maxslopevalue. As described above, the maxslopevalue is the maximum expected slope for a target having a predetermined radar cross section, at whatever range yields the maximum slope. Therefore, it should be apparent that the curve 126 should rarely cross above the maximum slope threshold 122 except for the presence of an interfering signal.

A second threshold 124, referred to herein as a mean slope threshold, is representative of a mean slope along the entire curve 126, including the curve portion 126c, scaled by a constant factor, K. The mean slope threshold can be computed, for example as:

$K*\mathrm{MeanValue}=\frac{\sum _{j=1}^{255}\mathrm{Slope}\left(j\right)}{255}*K$

In one particular embodiment, the constant factor, K, has a value of three. The factor, K, is selected to provide a low false alarm rate, i.e., a low rate at which the curve 126 crosses the second threshold 124. The factor, K, is also selected to provide a good probability of detection, i.e., presence of an interfering signal is most often detected by techniques described below. It should be apparent that the curve 126 should rarely cross above the mean slope threshold 124 except for the presence of an interfering signal.

While the mean slope threshold 124 is shown to be lower than the maximum slope threshold 122, the opposite is also possible. In an alternate embodiment, computation of the mean slope threshold excludes the curve portion 126c, for example, to the extent that the samples have crossed the maximum slope threshold.

The interfering signal is identified, for example, by the interference detector 78 of FIG. 2, which can generate the curve 126 and the first and second thresholds 122, 124. To this end, in one particular embodiment, detection of an interfering signal is made in accordance with the curve 126 crossing (i.e., exceeding) both the maximum slope and mean slope thresholds 122, 124. In another embodiment, detection of an interfering signal is made in accordance with the curve 126 crossing (i.e., exceeding) only the maximum slope threshold 122. In yet another embodiment, detection of an interfering signal is made in accordance with the curve 126 crossing (i.e., exceeding) only the mean slope threshold 124.

In other embodiments, detection of an interfering signal is made in accordance with the curve 126 equaling, rather than exceeding, both the maximum slope and mean slope thresholds 122, 124. In other embodiments, detection of an interfering signal is made in accordance with the curve 126 equaling, rather than exceeding, only the maximum slope threshold 122. In other embodiments, detection of an interfering signal is made in accordance with the curve 126 equaling, rather than exceeding, only the mean slope threshold 124. In other embodiments, detection of an interfering signal is made in accordance with the curve 126 equaling, rather than exceeding, only the maximum slope threshold 122 and the curve 126 exceeding the mean slope threshold 124. In other embodiments, detection of an interfering signal is made in accordance with the curve 126 equaling, rather than exceeding, only the mean slope threshold 124 and the curve 126 exceeding the maximum slope threshold 122. As used herein, the term “reaching” when referring to a curve and a threshold, refers to the curve equaling or exceeding the threshold.

In some arrangements, the identification can be made in accordance with a selected number of slope values exceeding (or equaling) the maximum slope threshold 122 and the mean slope threshold 124. The selected number can be any number greater than one. For example, in one particular embodiment, to detect an interfering signal, at least two slope values must exceed (or equal) the two thresholds, a first and a last slope value. In some embodiments, the selected number of slope values exceeding (or equaling) the thresholds is dynamically determined.

Once the interfering signal is detected by way of a reaching of one or both of the maximum slope threshold 122 and the mean slope threshold 124, an extent of the interfering signal is determined, for example, by the interference extent processor 88 of FIG. 2. To this end, reachings of the threshold can be identified, where the reachings correspond to the curve 126 (i.e., the slope) reaching the mean slope threshold 124.

The portion 126c of the curve 126, representative of a slope of an interfering signal, crosses the average threshold 124 at two points 130, 132 and also, in some instances, at intermediate points, of which a point 143 is but one example. The point 130 is associated with an earlier slope sample and is referred to herein as a lower slope sample limit. Similarly, the point 132 is associated with a later slope sample and is referred to herein as an upper slope sample limit.

A first guard band value having a selected number of slope sample values is subtracted from the lower slope sample limit 130 to achieve a point 136, referred to herein as a lower slope sample extent. A second guard band value is added to the upper slope sample limit 132 to achieve the point 138, referred to herein as an upper slope sample extent. In one particular embodiment, the first and second guard band values are the same, for example, five samples. However, in other embodiments, the first and second guard band values can be greater than or less than five samples, and they can be the same or different guard band values. In some embodiments, the first and second guard band values are dynamically determined, for example, in accordance with a shape of the curve portion 126c. Dynamic selection is described more fully below in conjunction with FIG. 10.

The first and second guard band values are selected in order to achieve subsequent removal of time samples associated with the slope samples between the points 136 and 138. The first and second guard band values are selected to avoid time samples associated with any residual part of the curve portion 126c (for example, leading and trailing edges of the curve portion 126c that are below the mean slope threshold 124) from remaining once eliminated from time samples as described below.

While the points 130, 132, 136, 138, 143 are described in conjunction with reaching (equaling or exceeding) of the mean slope threshold 124, in other embodiments, the reaching is determined in accordance with reaching the maximum slope threshold 122.

While the curve 126 represents a slope array having slope values, in particular slope absolute values, other types of arrays having other array values can also be used. As described further below, the array and associated array values can include, but are not limited to, a slope (i.e., first derivative) array having slope values (slope absolute values as shown in FIG. 4), a higher order derivative array having higher order derivative values, a power array having power values, and a time sample absolute value array having time sample absolute values. For each array type, corresponding thresholds comparable to the maximum slope threshold 122 and the mean slope threshold 124 can be established and used in much the same way as described above. For example, in another embodiment using a higher order derivative array, a maximum higher order derivative threshold and a mean higher order derivative threshold could be used in place of the maximum slope threshold 122 and the mean slope threshold 124. In yet another embodiment using a power array, a maximum power threshold and a mean power threshold could be used in place of the maximum slope threshold 122 and the mean slope threshold 124. In yet another embodiment using a time sample absolute value array, a maximum time sample absolute value threshold and a mean time sample absolute value threshold could be used in place of the maximum slope threshold 122 and the mean slope threshold 124. Accordingly, for any array type, sample limits comparable to the upper and lower slope sample limits 132, 130, respectively, and sample extents comparable to the upper and lower slope sample extents 138, 136, respectively could be identified.

While the curve 126 corresponds to slope values described to be slope absolute values, it should be appreciated that, in other arrangements, the slope values can instead be non-absolute slope values, which may have two polarities.

Referring now to FIG. 4A a graph 150 includes a horizontal scale in units of time sample number, wherein the time sample number is associated with the time samples 76 (FIG. 2). As shown, there are two hundred fifty six time samples. However, in other embodiments the number of time samples can be less than or greater than two hundred fifty six. The graph 150 also includes a vertical scale in units of time sample amplitude as represented by digital counts 76 (FIG. 2) provided, for example, by the analog-to-digital (A/D) converter 74 of FIG. 2.

A curve 152 is representative of the time samples 76 associated with the downconverted signal 73 (FIG. 2). The curve 152 connects the time samples, which are not individually shown. The curve 152 includes curve portions 152a, 152b representative of the time samples 76 having no interfering signal and a curve portion 152c representative of the time samples 76 having an interfering signal.

It will be apparent that the curve portion 152c generally aligns with the curve portion 126c of FIG. 4; i.e., the interfering signal represented in time by the curve portion 152c has a relatively high slope, represented by the curve portion 126c. Therefore, the interfering signal lower slope limit extent 136 and the interfering signal upper slope limit extent 138, which correspond to slope sample numbers, are associated with points 162, 164, respectively, corresponding to time sample numbers. The point 162 is referred to herein as an interfering signal lower time sample extent and the point 164 is referred to herein as an interfering signal upper time sample extent.

As described above, the points 136 and 138 have guard band values applied. Therefore, the points 162, 164 form a region 160 having boundaries 160a, 160b surrounding the interfering signal, without leaving a residual part of the interfering signal outside of the region 160.

It will become apparent from discussion below that time samples between the interfering signal lower time sample extent 162 and the interfering signal upper time sample extent 164 can be removed from the group of two hundred fifty six time samples before further processing.

Referring now to FIG. 5, in which like elements of FIG. 4A are shown having like reference designations, a graph 180 includes a horizontal scale in units of time sample number, wherein the time sample number is associated with the time samples 76 (FIG. 2). As shown, there are two hundred fifty six time samples. However, in other embodiments the number of time samples can be less than or greater than two hundred fifty six. The graph includes a vertical scale in units of time sample amplitude as represented by digital counts 76 (FIG. 2) provided, for example, by the analog-to-digital (A/D) converter 74 of FIG. 2.

The curve 152 is shown having only the curve portion 152a. The lower time sample extent represented by the point 162 terminates the curve 152 at the boundary 160a, effectively eliminating the curve portions 152c and 152b of FIG. 4A. From the lower time sample extent 162 to at least time sample number two hundred fifty five, the curve 152 is extended with zero values 182. In some embodiments, the curve 152 is further extended with zero values to provide more data points for a FFT processing. The zero values added to the curve 152 are referred to herein as “zero padding.” In other embodiments, the curve 152 can be extended with other values, more generally referred to herein as “padding.”

Referring now to FIG. 6, in which like elements of FIG. 1 are shown having like reference designations, a graph 200 has a horizontal axis in units of frequency provided as FFT frequency bins and a vertical axis in units of dB in FFT counts (provided by FFT processing of time samples 76 (FIG. 2) of the downconverted signal 73 (FIG. 2)). As described above, the curve 12 is representative of a frequency domain signal having a peak 14 indicative of a beat frequency, f1, and a corresponding range to a target. The curve 12 also has a noise background 16. The curve 18 has no distinct peak. The curve 18 is representative of a frequency domain signal indicative of the output of the FFT frequency domain processing when a received signal represented by the curve 12 also includes an interfering signal. In the curve 18, either the peak 14 cannot be found, or the peak 14 cannot be accurately found.

A curve 204 is representative of a resulting frequency domain signal provided by the above-described techniques shown in FIGS. 4-6. Some time samples used to generate the curve 204 are the same as the time samples used to compute the curve 18, however, for the curve 204, time samples associated with the interfering signal have been removed by the above-described techniques.

The curve 204 has a peak 208. The curve 204 also has a noise region 212. Comparing the curve 12 achieved in the presence of a target and no interfering signal with the curve 204 achieved in the presence of the target and the interfering signal, which is removed by the above-described technique, it can be seen that the peak 204 substantially aligns with the peak 14, but is wider. It can also be seen that the noise background 212 is higher than the noise background 16. It will be understood that both the wider peak 208 and the higher noise background 212 achieved in the presence of the target and the interfering signal, which is removed, tend to reduce an accuracy of the FMCW radar system compared to that of curve 12 achieved by the FMCW radar system in the presence of the target and no interfering signal. However, it can also be seen that the peak 208, indicative of a range to the target, can be identified, while a corresponding peak cannot be identified in the curve 18 achieved in the presence of the target and an interfering signal, which is not removed by the above-described technique. Therefore, the above-described technique is able to greatly improve the accuracy of the radar system in the presence of both a target and an interfering signal.

Referring now to FIG. 7, in which like elements of FIG. 4A are shown having like reference designations, a graph 220 includes a horizontal scale in units of time sample number, wherein the time sample number is associated with the time samples 76 (FIG. 2). The graph 220 also includes a vertical scale in units of time sample amplitude as represented by digital counts 76 (FIG. 2) provided, for example, by the analog-to-digital (A/D) converter 74 of FIG. 2.

As described above, the curve 152 is representative of the time samples 76 associated with the downconverted signal 73 (FIG. 2). The curve 152 includes the curve portions 152a, 152b representative of the time samples 76 having no interfering signal and the curve portion 152c representative of the time samples 76 having an interfering signal.

The interfering signal lower time sample extent 162 and the interfering signal upper time sample extent 164 define the region 160.

A curve 222 having a notch 222a is representative of a mathematical weighting function that can be applied to the time samples. It will be recognized that, when applied to the time samples represented by the curve 152, time samples in the region 160 would tend to be reduced, i.e., the interfering signal 152b would tend to be reduced, and two hundred fifty six windowed samples would still remain. In some embodiments, the two hundred fifty six windowed samples could be padded, for example, with another two hundred fifty six values, to provide additional samples to be used in subsequent FFT processing.

The curve 222 represents but one example of a weighting function that could be used. Other weighting functions can also be used. For example, in some embodiments, a weighting function can be generated based on a combination of windowing functions. For example, in one particular embodiment, in which a group of two hundred fifty six time samples (e.g., 76, FIG. 2) have fifty samples corrupted by an interfering signal, a weighting function can be constructed by multiplying a two hundred fifty six sample Hamming windowing function with a (1-Cosine) weighting function in the region of the fifty corrupted samples. In another embodiment, the weighting function can be the (1-Cosine) weighting function in the region of interference of the fifty samples without combination with another windowing function.

It should be appreciated that FIGS. 8-10B show flowcharts corresponding to the below contemplated technique which would be implemented in the FMCW radar system 50 (FIG. 2). The rectangular elements (typified by element 252 in FIG. 8), herein denoted “processing blocks,” represent computer software instructions or groups of instructions. The diamond shaped elements (typified by element 258 in FIG. 8), herein denoted “decision blocks,” represent computer software instructions, or groups of instructions which affect the execution of the computer software instructions represented by the processing blocks.

Alternatively, the processing and decision blocks represent steps performed by functionally equivalent circuits such as a digital signal processor circuit or an application specific integrated circuit (ASIC). The flow diagrams do not depict the syntax of any particular programming language. Rather, the flow diagrams illustrate the functional information one of ordinary skill in the art requires to fabricate circuits or to generate computer software to perform the processing required of the particular apparatus. It should be noted that many routine program elements, such as initialization of loops and variables and the use of temporary variables are not shown. It will be appreciated by those of ordinary skill in the art that unless otherwise indicated herein, the particular sequence of blocks described is illustrative only and can be varied without departing from the spirit of the invention. Thus, unless otherwise stated the blocks described below are unordered meaning that, when possible, the steps can be performed in any convenient or desirable order.

Referring now to FIG. 8, a process 250 begins at block 252, where a radar RF signal is received, for example, with the FMCW receiving system 66 of FIG. 2.

At block 253, the received radar RF signal is processed, providing a downconverted radar signal, e.g., the downconverted radar signal 73 of FIG. 2.

At block 254, the downconverted radar signal 73 is time sampled, for example by the A/D converter 74 of FIG. 2 to provide time samples 76 (FIG. 2).

As described above, an interfering signal can be present within the time samples. At block 256 the time samples are processed to detect an interfering signal, for example, with the interference detector 78 of FIG. 2. The time samples can be processed, for example, in accordance with techniques described in conjunction with FIGS. 4 and 4A. The processing of the time samples is also further described in conjunction with FIG. 9.

A decision is made at decision block 258 as to whether an interfering signal has been detected. If an interfering signal is detected at block 258, the process continues to block 260, where an extent of the interfering signal is determined, for example, with the interference extent processor 88 of FIG. 2. The determination can be made, for example, in accordance with techniques described in conjunction with FIGS. 4 and 4A. The determination is also further described in conjunction with FIG. 10.

At block 262, the identified interfering signal is removed from the time samples generated at block 254, for example, by the interference removal processor 92 of FIG. 2. The removal can be made, for example, in accordance with techniques described in conjunction with FIGS. 5 and/or 7. The removal is also further described in conjunction with FIGS. 10A and 10B.

At block 264, a signal 94 (FIG. 2) from which the interfering signal is removed, is processed, for example, with the radar return signal processor 82 of FIG. 2 to generate a radar system detection and/or classification of a target, represented by signal 84 of FIG. 2. The processing at block 264 can include, for example, a frequency domain conversion, e.g., an FFT.

In one particular embodiment, the radar system detection and/or classification of a target can be associated with a vehicle side object detection system, for example, as described in U.S. Pat. No. 6,577,269, issued Jun. 10, 2003. However, the above described system and techniques are not limited to a vehicle application.

Referring now to FIG. 9, a process 300, which can be performed, for example by the interference detector 78 of FIG. 2, begins at block 302 where a slope array having slope array values is generated in accordance with time samples generated at block 254 of FIG. 8. The slope array is represented, for example, by the curve 126 in FIG. 4.

At block 304, a first threshold is generated. In one particular embodiment, the first threshold corresponds to the maximum slope threshold 122 of FIG. 4. At block 306, a second threshold is generated. In one particular embodiment, the second threshold corresponds to the mean slope threshold 124 of FIG. 4.

At block 308, the slope array values generated at block 302 are compared against the first and/or second thresholds generated in blocks 304, 306.

At decision block 310, if at least a selected number of slope array values generated at block 302 reach both the first and second thresholds, then existence of an interfering signal is identified at block 312, i.e., the interfering signal is detected. The selected number can be any number greater than one. In one particular embodiment, the selected number is two. In some embodiments, the selected number is dynamically determined.

As described in conjunction with FIG. 4, in the particular embodiment shown, detection of an interfering signal is made in accordance with the slope array values (curve 126 of FIG. 4) reaching (exceeding or equaling) both the first and second thresholds (maximum slope and mean slope thresholds 122, 124) generated in blocks 304 and 306. In another embodiment, detection of an interfering signal is made in accordance with the slope array values reaching (exceeding or equaling) only the first threshold (maximum slope threshold 122). In yet another embodiment, detection of an interfering signal is made in accordance with the slope array values reaching (exceeding or equaling) only the second threshold (mean slope threshold 124).

If, at decision block 310, none of the slope array values reach the first and second threshold, then at block 314, no interfering signal is detected.

It will be understood that the process 300 can be performed on data corresponding, for example, to one radar chirp, and the process 300 can be repeated for each subsequent radar chirp, or for selected ones of successive radar chirps.

The process 300 is described in conjunction with a slope array having slope values, a corresponding maximum slope threshold, and a corresponding mean slope threshold. However, as described, for example, in conjunction with FIG. 4, in other embodiments, other array types having other array value types can be used along with corresponding maximum and mean thresholds associated with the other array value types.

Referring now to FIG. 10, a process 350, which can be performed, for example, by the interference extent processor 88 of FIG. 2, begins at block 352 where a lower slope sample limit is identified in accordance with an interfering signal. At block 354 an upper slope sample limit is identified. The upper and lower slope sample limits are represented, for example, by the points 130, 132 of FIG. 4.

At block 356 a first sample guard band value is selected and at block 358, a second guard band value is selected. The sample guard band value selection is described above in conjunction with FIGS. 4 and 4A. The guard band value selection can be static or dynamic.

A dynamic selection can depend upon a variety of factors. For example, in some embodiments, an interfering signal having a relatively small extent can result in a relatively small sample guard band value and an interfering signal having a relatively large extent can result in a relatively large sample guard band value. For another example, in some embodiments, an interfering signal having relatively small array values can result in a relatively small sample guard band value and an interfering signal having relatively large array values can result in a relatively large sample guard band value.

At block 360, the first sample guard band value is subtracted from the lower slope sample limit to provide an interfering signal lower slope sample extent. Similarly, at block 362, the second sample guard band value is added to the upper slope sample limit to provide an interfering signal upper slope sample extent. The interfering signal upper and lower slope sample extents are represented, for example, by the points 138, 136, respectively, of FIG. 4.

At block 364, an interfering signal lower time sample extent is identified in accordance with the interfering signal lower slope sample extent. Similarly, at block 366, an interfering signal upper time sample extent is identified in accordance with the interfering signal upper slope sample extent. The upper and lower time sample extents are represented, for example, by the points 164, 162, respectively, of FIG. 4A, which are seen to coincide with the upper and lower slope sample extents 138, 136, respectively, of FIG. 4.

It will be understood that the process 350 can be performed on data corresponding, for example, to one radar chirp, and the process 350 can be repeated for each subsequent radar chirp, or for selected ones of successive radar chirps.

The process 350 is described in conjunction with slope sample limits and extents. However, as described, for example, in conjunction with FIG. 4, in other embodiments, other array sample limits and extents associated with other types of array values can be used.

Referring now to FIG. 10A, a process 400, which can be performed, for example, by the interference removal processor 92 of FIG. 2, begins at block 402, where a contiguous group of time samples are selected from within time samples 76 of FIG. 2, excluding time samples between the interfering signal lower time sample extent and the interfering signal upper time sample extent identified at blocks 362 and 364 of FIG. 10. The selected time samples can correspond, for example, to the curve portion 152a of FIGS. 4A and 5. The selected time samples can correspond instead, for example, to the curve portion 152b of FIGS. 4A and 5. In some embodiments, the selected time samples are selected to be a largest group of contiguous time samples excluding time samples between the interfering signal lower time sample extent (e.g., 152a).

At block 404, the selected group of time samples are padded, for example, with the zero padding 182 shown in FIG. 5. However, in other embodiments, the selected group of time samples are padded with a value other than zero. In still other embodiments, the selected group of time samples are not zero-padded.

The zero padded time samples can then be further processed, for example by the radar return signal processor 82 of FIG. 2. The radar return signal processor 82 can, for example, perform an FFT on the zero padded time samples. However, in one particular embodiment, the zero padded time samples can be weighted at block 406 prior to FFT processing, for example with a conventional windowing function such as a Hanning or Hamming windowing function.

It will be understood that the process 400 can be performed on data corresponding, for example, to one radar chirp, and the process 400 can be repeated for each subsequent radar chirp, or for selected ones of successive radar chirps.

The process 400 is described in conjunction with a slope array. However, as described, for example, in conjunction with FIG. 4, in other embodiments, other types of arrays can be used.

Referring now to FIG. 10B, a process 450, which can be performed, for example, by the interference removal processor 92 of FIG. 2, represents an alternate method to that of FIG. 10A. The method 450 begins at block 452, where a conventional or a non-conventional weighting function is generated in accordance with the interfering signal lower time sample extent and the inferring signal upper time sample extent identified at blocks 362 and 364 of FIG. 10. The weighting function can correspond, for example to the weighting function 222 of FIG. 7. The weighting function is selected to provide a reduced interfering signal once the weighting function is applied to the time samples (76, FIG. 2).

At block 454, the weighting function is applied to the time samples (i.e., the time samples are multiplied by the weighting function), including the time samples between the interfering signal lower time sample extent and the interfering signal upper time sample extent.

The weighted time samples can then be further processed, for example by the radar return signal processor 82 of FIG. 2. The radar return signal processor 82 can, for example, perform an FFT on the weighted time samples generated at block 454.

It will be understood that the process 450 can be performed on data corresponding, for example, to one radar chirp, and the process 450 can be repeated for each subsequent radar chirp, or for selected ones of successive radar chirps.

The process 450 is described in conjunction with a slope array. However, as described, for example, in conjunction with FIG. 4, in other embodiments, other types of arrays can be used.

The method of interference rejection described above is not only applicable to a radar return signal having a continuous wave (CW) interfering signal but is also applicable to a radar return signal having any in-band interfering signal. The system and method described above encompass removal of corrupted data samples from a time sampled radar return signal in any FMCW radar operating at any frequency and having any bandwidth, with any type of interfering signal. It should be apparent that, because the interfering signal can be identified and removed in each chirp return signal, the interfering signal need not be at the same frequency in each chirp return signal, nor must it have the same characteristics. There can also be more than one interfering signal within the chirp return signal.

While a slope array having slope array values is described in embodiments herein, as described above, the array and associated array values can include, but are not limited to, the slope (i.e., first derivative) array having the slope values (slope absolute values as shown in FIG. 4) described, for example, by slope(j)=|time sample (j+1)−time sample(j)|, a higher order derivative array having higher order derivative values, a power array having power values described, for example, by power(j)=[time sample (j)]², and a time sample absolute value array having time sample absolute values described, for example, by absolute value (j)=|time sample (j)|.

It will be apparent that, when using a different type of array than the slope array (e.g., curve 126, FIG. 4), the above described first and second thresholds (e.g., 122, 124, FIG. 4) are no longer associated with slope values and the slope sample limits and interfering signal slope sample extents (e.g., points 136, 130, 132, 138, FIG. 4) are no longer associated with slope sample numbers. Instead the first and second thresholds are associated with a maximum array value and a mean array value, respectively, according to the type of array used. Similarly, the array sample limits and interfering signal array sample extents are associated with array sample numbers, according to the type of array used.

The above-described systems and techniques provide an ability to both detect an interfering signal (e.g., 62, FIG. 2) and to remove the interfering signal from time samples (e.g., time samples 76, FIG. 2). However, once detected, the interfering signal can be removed or otherwise avoided in other ways, which are described below.

As described above, one particular application of the FMCW radar is in an automotive radar system, for example, in an automotive radar system used to detect an object in a blind spot next to a vehicle. Automobile radars can use a frequency chirp extending, for example, from approximately 24.05 GHz to approximately 24.25 GHz, a frequency sweep of approximately two hundred MHz. With this arrangement, once time samples (e.g., 76, FIG. 2) are transformed to the frequency domain with an FFT or the like, target detections at resulting frequencies correspond to ranges to the target. Methods and apparatus for target detections are described, for example, in U.S. Pat. No. 6,577,269, issued Jun. 10, 2003. Processing described within U.S. Pat. No. 6,577,269, has an FMCW frequency sweep over a two hundred MHz sweep range, and results in a target range resolution of better than 0.6 meters per FFT bin (i.e., a 0.6 meter range resolution).

Conventional police radars used, for example, to detect speed of vehicles, operate, for example, at approximately 24.197 GHz, which is within the above-described automotive FMCW radar sweep range. Therefore, the time samples 76 collected when the transmitter is radiating close to 24.197 GHz will be influenced the most by the interfering signal. This interfering signal, if it is processed by an FFT, tends to spread into many FFT frequency bins.

In one arrangement, upon detection of the interfering signal, the frequency sweep, which normally sweeps from approximately 24.05 GHz to approximately 24.25 GHz, is altered so as to sweep through a different, e.g., smaller, sweep range. For example, in one particular embodiment, the frequency sweep is altered to sweep instead from approximately 24.05 GHz to approximately 24.195 GHz, for a total sweep of approximately one hundred forty five MHz. With this arrangement, it will be understood that the interfering signal from a police radar at approximately 24.197 MHz is entirely avoided and does not appear in the time samples 76 or in the FFT spectrum once the time samples 76 are transformed to the frequency domain.

It will be understood that the above-described reduction in the sweep range of the transmitted FMCW signal results in a larger value of target range per FFT bin, i.e., in a coarser range resolution. In order to obtain the original range resolution while using the reduced FMCW sweep range, values in the new FFT bins can be interpolated, which result in an improved range resolution that can be as good as the original range resolution using the original frequency sweep.

In some embodiments, the reduced sweep range is selected statically. In other words, once any interfering signal is detected, the system changes to transmit a predetermined reduced sweep range for a predetermined period of time. However, in other embodiments, the reduced sweep range is selected in accordance with a detected frequency of the interfering signal.

In still further embodiments, the interfering signal can be passively detected, without use of a transmitted RF signal. Similar to the embodiment described above, in response to the detected interfering signal, in some embodiments, the reduced sweep range is selected statically. In other words, once any interfering signal is detected, the system changes to transmit a predetermined reduced sweep range for a predetermined period of time. However, in other embodiments, the reduced sweep range is selected in accordance with a passively detected frequency of the interfering signal. In some embodiments, there can be a plurality of reduced sweep ranges, each having a start and a stop frequency in accordance with an identified frequency of a detected interfering signal.

All references cited herein are hereby incorporated herein by reference in their entirety. Having described preferred embodiments of the invention, it will now become apparent to one of ordinary skill in the art that other embodiments incorporating their concepts may be used. It is felt therefore that these embodiments should not be limited to disclosed embodiments, but rather should be limited only by the spirit and scope of the appended claims.

Claims

1. A method of radar processing, comprising: receiving an RF received signal having a frequency modulated continuous wave (FMCW) frequency sweep with a sweep range defining a frequency band;detecting a frequency of an interfering signal within the RF received signal;selecting a reduced frequency range less than the sweep range of the received RF signal and within the frequency band to avoid the frequency of the interfering signal; andprocessing the received RF signal within the reduced frequency range and excluding parts of the frequency band not within the reduced frequency range.

2. The method of claim 1, wherein the detecting the frequency of the interfering signal comprises: generating time samples associated with the RF received signal;generating an array having array values derived from the time samples;generating at least one threshold associated with the array values; andcomparing the array values with the at least one threshold.

3. The method of claim 1, further comprising: transmitting a frequency modulated continuous wave (FMCW) RF signal having the reduced sweep range.

4. The method of claim 2, further including detecting that the interfering signal exists if at least a selected number of array values reach the at least one threshold.

5. The method of claim 2, wherein the array values comprise a selected one of slope values, higher order derivative values, or time sample absolute values.

6. The method of claim 2, wherein the generating at least one threshold comprises generating at least one of a first threshold related to a maximum expected array value or a second threshold related to a mean of the array values.

7. The method of claim 2, wherein the array values comprise slope values derived from differences of the time samples.

8. A radar system, comprising: an RF receiving module configured to receive an RF received signal having a frequency modulated continuous wave (FMCW) frequency sweep with a sweep range defining a frequency band, wherein the RF receiving module is configured to provide a downconverted signal in response to the RF return signal;an RF signal sampling module coupled to the RF receiving module and configured to generate time samples associated with the downconverted signal;an interference detector coupled to receive the time samples, configured to detect a frequency of an interfering signal and configured to select a reduced frequency range less than the sweep range of the received RF signal and within the frequency band to avoid the frequency of the interfering signal; anda processor configured to process the received RF signal within the reduced frequency range and excluding parts of the frequency band not within the reduced frequency range.

9. The system of claim 8, wherein the interference detector includes an array processor configured to generate an array having array values derived from the time samples, configured to generate at least one threshold associated with the array values, and configured to compare the array values with the at least one threshold.

10. The system of claim 8, wherein the at least one threshold includes at least one of a first threshold related to a maximum expected array value or a second threshold related to a mean of the array values.

11. The system of claim 8, further comprising an RF transmitting module configured to transmit a frequency modulated continuous wave (FMCW) RF signal having the reduced sweep range.

12. The system of claim 9, wherein the array processor is further configured to determine the interfering signal exists if at least the selected number of array values reach the at least one threshold.

13. The system of claim 9, wherein the array values comprise a selected one of slope values, higher order derivative values, or time sample absolute values.

14. The method of claim 9, wherein the array values comprise slope values derived from differences of the time samples.