Claims:CLAIMS
I/We Claim,
1. A method of enhancing a resolution in a radar system having an antenna aperture
comprising:
measuring a first set of radiation pattern corresponding to a first set of receiving antennas
by feeding a known radio frequency (RF) signal over the first set of receiving antennas, wherein
the first set of radiation pattern is due to an impairment in the radar system;
coherently combining an interpolated radiation pattern with a received radar signal
received by the set of receiving antenna when employed for an object detection, to generate a high
signal to noise ratio (SNR) received signal; and
iteratively combining the high SNR received signal with the interpolated signal to reduce
the error due to the impairment.
2. The method of claim 1, where in the impairment representing a second set of
radiations by a corresponding set of feed lines connecting the first set of receiving antennas to a
signal processing electronics in the radar system.
3. The method of claim 2, further comprising translating the first radiation pattern to
the interpolated radiation pattern using at least one of a frequency of operation, bandwidth, field
of view of the radar system when employed for the object detection.
4. The method of claim 3, wherein the first radiation pattern comprises a set of phase
angles corresponding to each antenna in the set of receiving antenna element and the interpolated
radiation pattern comprises a second set of phase angles interpolated from the first set of phase
angles.
5. The method of claim 4, further comprising, performing first Fast Fourier
Transform (FFT) operation on a RF signal received over the first set of receiving antennas when
employed for the object detection, to generate a range data and performing a second FFT
operation on the range data to generate a range and Doppler data.
wherein the received radar signal comprising the range and the Doppler data.
6. The method of claim 5, further comprising, iteratively combining the high SNR
received signal with the interpolated signal using a relation: ??????(??) = ??? - ?????? (??)????? such that
??????(??) successively reduced to zero by iteratively setting the values for x, in that Y represents the
13
high SNR received signal, Ami represents the interpolated signal, k is a value representing number
of iteration and x is a value determined at every iteration to minimize the residue ??????(??).
7. A method, system, and apparatus providing one or more features as described in
the paragraphs of this specification. , Description:Form 2
The Patent Act 1970
(39 of 1970)
AND
Patent Rules 2003
Complete Specification
(Sec 10 and Rule 13)
Title
A SYSTEM AND METHOD FOR GENERATING POINT
CLOUD DATA IN A RADAR BASED OBJECT DETECTION
Applicant(s) Steradian Semiconductors Private Limited
Nationality India
Address
216, Arcade at Brigade Metropolis, Whitefield Main Road,
Bangalore - 560048, Karnataka, India.
The following specification, particularly describes the invention and the manner in
which it is to be performed.
2
DESCRIPTION
FIELD OF INVENTION
[0001] Embodiments of the present disclosure relate to high resolution radar system and in
particular relates to a system and method for generating point cloud data in radar based object
detection.
RELATED ART
[0002] Radar systems are generally employed for object detection and increasingly used in
various automotive applications such as for driver assistance, obstacle detection, avoidance, and
navigation of drones/UAVs, for example. As is well known, radars can detect surrounding
obstacles or objects and send the relevant information like distance, relative position, and direction
and velocity of the object that are in motion to a controller (software or hardware) or to a decision
making unit in the automotive device.
[0003] In some applications antenna arrays are employed to transmit and receive radar signal. The
antenna array enables formation of an RF signal beam both for transmitting and receiving the
radar signal. In that, time shifted (phase shifted) radar signals are transmitted/received over the
antennas to steer the beam in desired direction as is well known in the art. A two or three
dimensional object shape and location is determined by steering the beam over a range/area.
[0004] Briefly, FIG. 1A illustrates a conventional technique for determining the range and angle.
In that, antenna array 101 transmits and receives the radar signal. In that, antenna array 101
transmits the radar signal provided by the radar transmitter 102. As is well known, the radar
transmitter 102 provides phase shifted version of a radar signal to the antenna array to form a
beam in a desired direction and the phase angle is adjusted to steer the beam over the desired area.
Similarly, the antenna array 101 receives the radar signal reflected from one or more objects and
provides the received signal to the radar receiver 103.
[0005] The radar receiver 103 may demodulate and perform signal processing like Fast Fourier
Transform (FFT) to extract range and Doppler. The range and Doppler is provided to the detector
104 that selects signals with signal-to-Noise ratio (SNR) higher than a preset threshold. The
selected signal is provided to the beam former 105 to form beam from the selected signals. The
beam provides the angle information. The detector 106 selects the beam that corresponds to local
peak. However such conventional radarsystem lacks resolution to detect objects with more
precision.
3
[0006] In particular, when the radar system is employed for imaging (often referred to as imaging
radar), a high range as well as a high angle resolution (both azimuth and elevation) is desirable to
get the shape and contour of a 3D object. For example, high angle resolution in a radar system
enables representing a 3D object with a larger number of points for more precise detection.
However, the beam width is limited by the antenna radiation pattern such as main lobe, side lobes
etc., as is well known in the art. Some of the conventional techniques employed for detecting
more objects (increase angle resolution) are described below for reference.
[0007] FIG. 1B illustrates another conventional technique. As shown there correlator 127 and
Capon beamformer 128 are additionally employed between the detector 104 and 106. The
correlator 127 makes use of the selected signal from detector 104 over multiple frames to
determine the covariance. For example, the correlator 127 may perform correlation of the data
received from the detector 104 with the data received over prior K frames. The correlated data is
provided to the Capon beamformer 128 for generating the beam.
[0008] Due to correlation, the side lobes errors are removed to an extent there by increasing the
resolution as is well known in the art. However, such conventional technique requiresbuffering the
data over K frames there by increasing the response time, at least. FIG.1C illustrates yet another
conventional technique. In that, the source estimator 137 performs data covariance over K
snapshots, find Eigen values for signal and noise subspace, and estimate number of sources. The
MUSIC Pseudo spectra beam former 138 performs beam forming employing well known MUSIC
algorithm.
[0009] As may be appreciated, both conventional techniques of FIG. 1B and 1C employ multiple
snapshots of the received data and hence are referred to as data dependent beam forming. The
conventional techniques fail to adequately remove errors due to some impairment in the system, at
least.
SUMMARY
[0010] According to an aspect, method of enhancing a resolution in a radar system having an
antenna aperture comprises measuring a first radiation pattern corresponding to a first set of
receiving antennas by feeding a known radio frequency (RF) signal over the first set of receiving
antennas, wherein the first set of radiation due to an impairment, coherently combining an
interpolated radiation pattern with a received radar signal received by the set of receiving antenna
when employed for an object detection, to generate a high signal to noise ratio (SNR) received
4
signal, and iteratively combining the high SNR received signal with the interpolated signal to
reduce the error due to the impairment.
[0011] According to another aspect the impairment is a set of radiation by the corresponding set
of feed lines connecting the first set of receiving antennas to a signal processing electronics
(transceiver) in the radar system. Further method comprises, translating the first radiation pattern
to the interpolated radiation pattern using at least one of a frequency of operation, bandwidth, and
field of view of the radar system when employed for the object detection.
[0012] Several aspects are described below, with reference to diagrams. It should be understood
that numerous specific details, relationships, and methods are set forth to provide full
understanding of the present disclosure. Skilled personnel in the relevant art, however, will
readily recognize that the present disclosure can be practiced without one or more of the specific
details, or with other methods, etc. In other instances, well-known structures or operations are not
shown in detail to avoid obscuring the features of the present disclosure.
BRIEF DESCRIPTION OF DRAWINGS
[0013] FIG. 1A illustrates one conventional technique for determining the range and angle.
[0014] FIG. 1B illustrates another conventional technique.
[0015] FIG. 1C illustrates yet another conventional technique.
[0016] FIG. 2 is a block diagram of an example system 200 (environment) in which various
aspects of the present invention may be seen.
[0017] FIG. 3 is an example radar transceiver for object detection and recognition in an
embodiment.
[0018] FIG. 4A is a block diagram illustrating example impairment in the radar system.
[0019] FIG. 4B illustrates an example smear in the phase response due to the radiation from feed
lines.
[0020] FIG. 4C illustrates an example real target 491 and smear 499 due to the feed line radiation.
[0021] FIG. 5A is a block diagram illustrating the manner in which the resolution of a radar
system is improved in an embodiment.
[0022] FIG. 5B illustrates an example impairment in the radar system.
[0023] FIG. 6 is an example precision object detector (POD) in an embodiment.
DETAILED DESCRIPTION OF THE PREFERRED EXAMPLES
5
[0024] FIG. 2 is a block diagram of an example radar system 200 (environment) in which various
aspects of the present invention may be seen. The environment is shown comprising objects 210,
Radio Frequency (RF) transceiver 220, processor 230, output device 240 and memory 250. Each
element in the system 200 is further described below.
[0025] RF transceiver 220 transmits a radar (RF) signal over a desired direction(s) and receives a
reflected radar signal that is reflected by the object 210. In one embodiment, the RF transceiver
220 may employ multiple (one or more) receiving antennas to receive the reflected RF signal and
multiple (one or more) transmitting antenna for transmitting the radar signal. Accordingly, the
transceiver 220 may employ these multiple transmitting/receiving antennas in several of multiple
input and multiple output (MIMO) configurations to form desired transmitting and receiving RF
signal beam (often referred to as Beam forming) to detect objects from the reflected signal. The
objects 210 may comprise a terrain, terrain projections, single object, cluster of objects, multiple
disconnected objects, stationary object, moving object, live objects etc.
[0026] Processor 230 conditions and processes the received reflected RF signal to detect one or
more objects (for example 210) and determines one or more properties of the objects. The
properties of the object thus determined (like shape, size, relative distance, velocity etc.) are
provided to the output device 240. In an embodiment, the processor 230 comprises signal
conditioner to perform signal conditioning operations and provides the conditioned RF signal for
digital processing. The memory 250 may store RF signal like samples of the reflected RF signal
for processing. The processor 230 may temporarily store received data, signal samples,
intermediate data, results of mathematical operations, etc., in the memory 250 (such as buffers,
registers). In an embodiment, the processor 230 may comprise group of signal processing blocks
each performing the specific operations on the received signal and together operative to detect
object and its characteristics/properties.
[0027] The output device 240 comprises navigation control electronics, display device, decision
making electronic circuitry and other controllers respectively for navigation, display and further
processing the received details of the object. Accordingly, the system 200 may be deployed as part
of unmanned vehicles, driver assistant systems, for obstacle detection, navigation and control, and
for terrain mapping.
[0028] In an embodiment, the RF transceiver 220, processor 230, and memory 250 are
implemented as part of an integrated circuit integrated with other functionality and/or as a single
6
chip integrated circuit with interfaces for external connectivity like the output device 140. The
manner in which the transceiver 220 and the processor 230 (together referred to as Radar
transceiver) may be implemented in an embodiment is further described below.
[0029] FIG. 3 is an example radar transceiver for object detection and recognition in an
embodiment. The radar transceiver 300 is shown comprising transmitting antenna 310, transmitter
block 315, receiving antenna array 320, mixer 325, filter 330 Analog to digital convertor (ADC)
340, Range FFT 350, Doppler FFT 360 and precision object detector 370. Each element is
described in further detail below.
[0030] The transmitting antenna array 310 and the transmitter 315 operate in conjunction to
transmit RF signal over a desired direction. The transmitter 315 generates a radar signal for
transmission and provides the same to the transmitting antenna array310 for transmission. The
transmitting antenna array 310 is employed to form a transmit beam with an antenna aperture to
illuminate objects at suitable distance and of suitable size. Various known beam forming
techniques may be employed for changing the illuminated region. In one embodiment, the
transmitter 215 may generate a radar signal comprising sequence of chirps.
[0031] The receiving antenna array 320 comprises antenna elements each element capable of
receiving reflected RF signal. The receiving antenna array 320 is employed to form an aperture to
detect objectswith a desired resolution (for example object of suitable size).The RF signal
received on each element corresponding to one transmitted chirp represents one snapshot. The
received signal (the sequence of snapshots corresponding to the sequence of chirps transmitted) is
provided to the mixer 325.
[0032] The Mixer 325 mixes RF signal received on each of M antenna elements with the
transmitted chirp (local oscillator frequency) to generate an intermediate frequency signal (IF
signal). In that the mixer 325 may comprise M number of complex or real mixers to mix each RF
signal received on the corresponding M antenna elements. Alternatively, the mixer 325 may
comprise a fewer mixers multiplexed to perform desired operation. The N number of intermediate
frequency (IF) signal is provided on path 323 to filter 330. The filter 330 passes the IF signal
attenuating the other frequency components (such as various harmonics) received from the mixer.
The filter 330 may be implemented as a pass band filter to pass a desired bandwidth (in
conjunction with chirp bandwidth BW). The filtered IF signal is provided on path 334 to ADC
340.
7
[0033] The ADC 340 converts IF signal received on path 334 (analog IF signal) to digital IF
signals. The ADC 340 may sample the analog IF signal at a sampling frequency Fs and convert
each sample value to a bit sequence or binary value. In one embodiment the ADC 340 may
generate 256/512/1024 samples per chirp signal. The digitized samples of IF signal (digital IF
signal) is provided for further processing.
[0034] The Range Fast Fourier transform (FFT) 350, performs P point FFT on the digital IF
samples to generate plurality of ranges of the plurality objects 210. For example, range FFT 350
performs FFT on digital IF signal corresponding to each chirp. The Range FFT 350 produces
peaks representing the ranges of the objects.
[0035] The Doppler FFT 360 performs FFT operation on the ranges across N chirps. The peaks in
the Doppler FFT represent the Doppler of the objects or the velocity of the objects. The ranges
and Doppler of the objects are provided to the precision object detector 370. The precision object
detector 370 detects objects with higher resolution for the RF signal received on the receiving
antenna array 320 with resolution corresponding to the antenna aperture. In one embodiment, the
precision object detector 370 reduces the error due to impairments in the system 300 and enhances
the resolution. Accordingly, example impairment in a radar system is first described below.
[0036] FIG. 4A is a block diagram illustrating example impairment in the radar system. The block
diagram is shown comprising transceiver pins 410A-N, antenna pins 420A-N, antenna pitch 430A
and 430B and feed lines 440A-N. Each element is described in further detail below.
[0037] The transceiver pins 410 A-N provide means for electrical connectivity between the
transceiver and the antenna array. The transceiver pins 410A-N represents the interface point on
an integrated circuit or on a device implementing transmitter block/receiver block. The electronics
and/or the processing elements to generate radar signal for transmission, in full or in part may be
implemented in the transmitter block (for example block 320) so that the radar signal, ready for
transmission, (to be radiated over antenna array) are presented on selected transceiver pins 410AN.
Similarly, the electronics and/or the processing elements to receive radar signal that are
reflected from the object, in full or in part may be implemented in the receiver block (for example
blocks 325, 330, 340, 350 etc.,) so that the radar signal, as received, ready over antenna array are
presented on selected transceiver pins 330A-N for processing.
[0038] The antenna pins 420A-N are the interface points connecting the antenna elements in the
antenna array (for example 310). The antenna array elements are arranged in a definite pattern and
8
spacing with a pitch 430A and/or 430B. The pattern and the pitch are often determined based on
frequency of RF signal, Radiation pattern and the aperture desired. Accordingly, the antenna array
is positioned such that, the signal radiated is towards the space in which the objects are required to
be detected. The antenna pitch 430A and 430B may be set to different distance. For example, the
antenna elements configured to transmit may be spaced with a pitch 430A and the antenna
elements configured to receive the reflected signal may be spaced apart with a pitch 430B.
[0039] The feed lines 440A-N are conductive paths electrically connecting the transceiver pins
410A-N with the corresponding ones of antenna pins 420A-N. The pitch of the antenna pins and
pitch of the transceiver pins may be different. For example, the pitch 415 of the transceiver pins
may be set to 0.5mm while the pitch of the antenna pin may be set to 2mm. Further, location and
position of the antenna array and transceiver, and other aspects may cause the feed line to bend
and curve and have different length. The feed line 440A-N may also radiate signal and cause
interference in both transmitted and received RF signal.In particular, at millimeter wave
frequencies the feed lines 440A-Nradiate and corrupt the radiation behavior of the antenna
element in the array 310. This means, the gain and phase response of each antenna will not be
identical to each other.
[0040] As may be appreciated, the convention systems discussed in FIG.1A, 1B and 1C, estimate
the angle by considering that the phase responses across sensors (antenna) are proportional to the
path lengths (treating each antenna element as point source as its geometric centre). When the feed
lines are radiating and are not identical, this consideration is violated.For example, in a uniform
linear array, the phase response deviates from being ideal linear and hence creates smearing (loss
of resolution) & spurious targets. FIG. 4B illustrates an example smear in the phase response due
to the radiation from feed lines. FIG. 4B is a graph, in that X-axis represents receiver antenna
elements and Y-axis represents the phase angle of the signal radiated. As may be seen, the points
460A-C representthe phase angle of the radar signal received only by antenna elements. The
points 470A-C represent the phase angle deviation due to radiation of the feed line. The FIG. 4C
illustrates an example real target 491 and smear 499 due to the feed line radiation. In one
embodiment, the precision object detector 370 reduces the smear and improves the resolution.
[0041] FIG. 5A is a block diagram illustrating the manner in which the resolution of a radar
system is improved in an embodiment. In block 510, a non ideality in the radar system is
9
measured. The non ideality in the radar system may comprise a radiation pattern due to feed lines
that may be measured by subjecting the radar system 300 to a known signal.
[0042] FIG. 5B illustrates an example impairment in the radar system. In that, the angles ?1, ?2,
?3,…..?N, represent the primary angles measured across N receiving antenna elements 560A-N in
the array respectively. A reference signal with a known phase angle, frequency, and bandwidth for
example, may be fed to the antenna elements to measure the primary angles. Accordingly,?1, ?2,
?3,…..?N, may represent an object at particular angle from the radar antenna array. The angles f1,
f2, f3… fk represent the angles measured due to feed lines radiations. These angles f1, f2, f3… fk
may result in (represent) a spurious object. Accordingly, for the reference signal, angles ?1, ?2,
?3,…..?N, represent the correct/desired object (primary object) and angles f1, f2, f3… fk represent
the impairment (non ideality). These impairments need to be cancelled or corrected for accurate
and high resolution object detection.
[0043] The measured non ideality may be quantized and/or digitized to store in a memory. In
block 520, the measured non ideality is interpolated (interpolated non ideality data) corresponding
to the operating conditions like desired resolution, frequency of operation, RF bandwidth, and
Field of View (FOV), etc. As may be appreciated, the stored impairments (angles f1, f2, f3… fk)
correspond to transmission/reception of the reference signal. However, in real time, the radar
system may be subjected to transmit and receive signal with different frequency, bandwidth etc.
Accordingly, a new set ofangles ?1, ?2, ?3… ?k (interpolated non ideality data) may be generated
through interpolation techniques that corresponds to the operating frequency, bandwidth and
FOV, for example.
[0044] In block 530, the interpolated non ideality data is coherently combined with the received
data to enhance the signal to noise ratio. In block 540, the interpolated non ideality data is
iteratively combined to determine the angle of the objects. For example, a sparse signal processing
technique may be employed that chooses the direction(s) of the incoming data by iteratively
reducing the residual error as a function of incoming data and interpolated non ideality data.
[0045] In block 550, the points representing the objects are determined from the angles. For
example, objects and its angle may be selected by using one or more technique such as using a
relative threshold from the output having maximum strength, by iterating for pre-defined number
of iteration(s) and/or by selecting a solution of particular iteration if residual error is smaller than
pre-defined threshold.
10
[0046] As a result, the radar system 300 overcomes at least some of the dependency like,
dependency on linear phase assumption across MIMO antenna array, dependency on RF start
frequency and RF sweep bandwidth. Further, the radar system 200/300 achieves better resolution
in angle over data-independent/data-dependent beam forming approaches discussed with reference
to FIG 1A-1C.In the radar system 300, there is no need to estimate number of signal sources apriori
like in sub-space based approaches. Thus, angle is estimated from a single snapshot of the
data unlike as in sub-space methods and data-dependent beamformer methods. The manner in
which the precision object detector 370 may be implemented employing the stored non-idealities
is further described below.
[0047] FIG. 6 is an example Precision Object Detector (POD) in an embodiment. The POD 601 is
shown comprising memory 610, interpolator 620, data selector 630, direction selector 640 and
point cloud generator 650. Each element is described in further detail below.
[0048] The memory 610, stores the non-ideality data measured under known/standard conditions.
For example, the non-ideality data may comprise, the reference signal information (such as, RF
frequency, bandwidth, sweep speed, etc.) that is employed for measuring the non ideality data, the
primary angles (?1,?2, ?3,…..?N, for example) and non ideality angles (f1, f2, f3… fk,for example).
The non ideality data may be stored in digital form as a matrix. For example, when the antenna
array is two dimensional, the non ideality angles may be stored as a two dimension matrix.
[0049] The interpolator620 interpolate the non ideality angles to generate an error metrics
(interpolated non ideality) corresponding to operating conditions. For example, the interpolator
may adjust the non ideality angles f1, f2, f3… fk to new set of angles (error metric) ?1, ?2, ?3… ?k
based on at least one of the frequency of operation, bandwidth, etc. In one embodiment, the
interpolator may generate new set of angles in proportionate to the frequency of operation.
[0050] The data selector 630 receives the range and Doppler data corresponding to each antenna
element. The range and Doppler data may comprise peaks representing an object distance and the
relative velocity. The data received on path 631 by the data selector 630 corresponds to data on
path 367. As may be appreciated, the data on path 367 represents the radar signal received on the
antenna array and subjected Fast Fourier Transforms (FFT) by range FFT 350 and subsequently
the Doppler FFT 360. Thus, the data on path 631 comprises peaks representing range and velocity.
[0051] The data selector 630 performs coherent combining of the data on path 631 and the
interpolated non ideality received on path 623. The coherent combining operation increases the