Phase-coded FMCW MIMO radar with cross-correlation residual cancellation

Abstract

Aspects of the present disclosure are directed to systems and methods of operating a MIMO radar system for automotive applications, including transmitting FMCW signals from first and second transmitters; wherein the signals each comprise a plurality of chirps, and each signal comprises phase modulation of the FMCW signal within each chirp by a respective CDMA phase code to produce respective first and second PM-FMCW signals; receiving a composite signal, wherein the composite signal comprises at least one reflection of the first PM-FMCW signal and the second PM-FMCW signal; and in the digital domain, processing a plurality of down-converted and pre-processed groups of samples of the composite signal by: applying a frequency response correction; applying a group delay filter; the CDMA phase code is removed; the 2D-FFT is applied; and removing a phase code cross-correlation residual from the corresponding transformed demodulated sample group by calculating an estimated residual and subtracting the estimated residual from the corresponding transformed demodulated signal.

Description

TECHNICAL FIELD

[0001] The present disclosure relates generally to MIMO FMCW radar systems and methods. BACKGROUND

[0002] Most multiple-input multiple-output (MIMO) multiplexing schemes for automotive radar have so far relied on either time discrimination (time division multiple access - TDMA) or speed discrimination (Doppler division multiple access - DDMA). However, these schemes do not easily scale to a large number of simultaneously active transmitters, partly due to the reduction of explicit speed separation.

[0003] One alternative is chirped code division multiple access (CDMA). This does not limit the explicit speed, but degrades the dynamic range of the system typically in proportion to the number of active transmit (Tx) elements due to linear processing on the receiver. Specifically, the dynamic range or orthogonality ratio (OR), which can be defined as the ratio between the peak level and the average side lobe level and provides an indication of the ease of peak detection, decreases with the number of transmitters K. Specifically, for a code length Lc, the orthogonality ratio can be approximated as:

[0004] OR = 10 log10((K - 1) / Lc) (1)

[0005] Using equation 1, it can be seen that a typical system would have an average side lobe 33 dB below the peak with 2 transmit elements and a code length of 2048, but this would drop to only 21 dB for 16 transmit elements. For such CDMA systems, it has been considered to improve the dynamic range by means of residual cancellation, but for a fixed frame rate, the number of chirps and the duration of each chirp must be balanced. Since increasing the code length requires increasing the number of chirps, this is constrained by the requirement to have a sufficient chirp length. SUMMARY

[0006] According to a first aspect of the disclosure, there is provided a method of operating a multiple-input multiple-output, MIMO, radar system for automotive applications, the method comprising: transmitting a first frequency-modulated continuous-wave, FMCW, signal from a first transmitter; transmitting a second FMCW signal from a second transmitter; wherein the first and second signals each comprise a plurality of chirps, and each signal comprises a FMCW signal within each chirp being phase-modulated by a respective code-division multiple access, CDMA, phase code to produce respective first and second PM-FMCW signals; receiving, at a receiver, a composite signal comprising at least one reflection of each of the first and second PM-FMCW signals; and in the digital domain, processing a plurality of down-converted and pre-processed sample groups of the composite signal by: applying a frequency response correction to the down-converted and pre-processed sample groups to provide corrected sample groups; applying a group delay filter to align the CDMA phase codes of the corrected sample groups to provide time-aligned sample groups; removing the respective CDMA phase codes from the time-aligned sample groups to provide respective demodulated sample groups; transforming the respective demodulated sample groups to a range-Doppler domain by applying a two-dimensional fast Fourier transform, 2D-FFT; and removing phase code cross-correlation residuals from the respective transformed demodulated sample groups by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals. The residual cancellation allows weaker targets to be unmasked instead of being hidden by cross-correlation residuals from stronger targets. This in turn can enable PM-FMCW to be used with acceptable levels of processing for sufficient performance, that is, the processing can be implemented within a sufficiently short effective time frame. It will be appreciated that although only a first transmitter is recited above, there can be further transmitters, and the phase code cross-correlation residuals can be derived from more than a single cross-correlation.

[0007] In one or more embodiments, removing the respective CDMA phase codes comprises multiplying the group-corrected frequency-response-corrected sample groups by a conjugate of the PM-FMCW signal.

[0008] In one or more embodiments, the first and second signals each comprise a repeating frame of a plurality of chirps, and each chirp within the frame has a different CDMA phase code.

[0009] In other embodiments, the first and second signals each comprise a repeating frame of a plurality of chirps, and each chirp within the frame has the same CDMA phase code. In such embodiments, each chirp within the frame can have a different second CDMA code.

[0010] Accordingly, the CDMA codes can be repeated within a frame, such that the CDMA phase codes do not occupy the entire duration of a frame; for example, the CDMA phase codes can be spread across a single chirp, and a separate "outer code" can be used to distinguish between chirps in a frame.

[0011] In one or more embodiments, each chirp comprises a plurality of chips, and has a phase code comprising a combination of respective phases of each of the chips. Typically, each chirp can have between four and 16 chips, or even up to 64 or more chips per chirp.

[0012] In one or more embodiments, the frequency modulation is linear modulation. Linear frequency modulation can provide for either up-chirp or down-chirp, without limitation.

[0013] In one or more embodiments, removing the phase code cross-correlation residue from each respective transformed demodulated signal comprises a residue cancellation iteration comprising: applying a detection mask to the respective transformed demodulated signal, thereby isolating high-probability target points; applying a 2D inverse FFT, 2D-IFFT, resulting in an isolated target signal; estimating respective estimated cross-correlation components in or within each of the other transformed demodulated signals from the transformed isolated target signal, and combining them to produce an estimated residue signal; applying a 2D FFT to the estimated residue signal, resulting in a transformed estimated residue signal; and subtracting the transformed estimated residue from the respective transformed demodulated signal. It will be appreciated that under this arrangement, a separate 2D FFT is applied to each Tx. In particular, the application of the group delay portion and the application of the code in fast and slow time allow for the use of two dimensions, and enable longer codes with higher orthogonality ratios. This can improve detection of weak targets, and make the technique more suitable for dense traffic scenarios and a larger number of transmit antennas.

[0014] In one or more embodiments, estimating respective estimated cross-correlation components in each of the other transformed demodulated signals from the transformed reduced target signal comprises: multiplying the other transformed demodulated signals by pre-computed cross-correlation coefficients between the other signals and the signal.

[0015] In one or more embodiments, the method can additionally comprise a further residue cancellation iteration. Increasing the number of residue cancellation iterations improves the level of detection of targets; however, this should typically be traded off or balanced against the increased processing resources required for multiple iterations. The number of iterations for any given application can be determined by the specifics of the application, the device, and in particular the DSP used, etc.

[0016] According to a second aspect of the disclosure, there is provided a CDMA automotive MIMO radar system, the CDMA automotive MIMO radar system comprising: a plurality of transmitters each configured to transmit a radar frame, the radar frame comprising a plurality of chirps having a frequency modulation therein, wherein the frame from each transmitter comprises a code division multiple access, CDMA, encoding, each of the plurality of chirps comprising a phase modulation within according to the CDMA encoding; a plurality of receivers; a respective RF front-end processor for each of the plurality of receivers and configured to provide a group of down-converted and pre-processed digital samples to a digital signal processor; and a digital signal processor configured to apply a frequency response correction to the group of down-converted and pre-processed samples to provide a corrected group of samples; apply a group delay filter to align the CDMA phase codes of the corrected group of samples to provide a time-aligned group of samples; remove respective CDMA phase codes from the time-aligned group of samples to provide respective groups of demodulated samples; transform the respective groups of demodulated samples to a range-Doppler domain by applying a two-dimensional fast Fourier transform, 2D-FFT; and remove phase code cross-correlation residuals from the respective groups of transformed demodulated samples by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals.

[0017] In one or more embodiments, removing the respective CDMA phase codes comprises multiplying the group-corrected frequency-response-corrected group of samples with a conjugate of the PM-FMCW signal. That is, multiplying it with the phase code used to modulate the chirps to obtain the PM-FMCW signal.

[0018] In one or more embodiments, the first and second signals each comprise a repeating frame of a plurality of chirps, and each chirp within the frame has a different CDMA phase code. In one or more embodiments, each chirp comprises a plurality of chips, and has a phase code comprising a combination of respective phases of each of the chips. In one or more embodiments, the frequency modulation is a linear modulation.

[0019] In one or more embodiments, the processor is configured to remove the phase code cross-correlation residuals from each respective transformed demodulated signal (that is, each respective channel) comprises performing a residual cancellation iteration by: applying a detection mask to the respective transformed demodulated signal, thereby isolating high-probability target points and removing a remaining portion of the signal; applying a 2D inverse FFT, 2D-IFFT, resulting in an isolated target signal; estimating respective estimated cross-correlation components of each other transformed demodulated signal from the isolated target signal and combining them to produce an estimated residual signal; and applying a 2D FFT to the estimated residual signal to produce a transformed estimated residual signal; and subtracting the transformed estimated residual from the respective transformed demodulated signal.

[0020] In one or more embodiments, the processor is configured to perform a plurality of residual cancellation iterations.

[0021] According to a third aspect of the disclosure, there is provided a method of operating a multiple-input multiple-output, MIMO, radar system for automotive applications, the method comprising: transmitting a first frequency-modulated continuous-wave, FMCW, signal from a first transmitter; transmitting a second FMCW signal from a second transmitter; wherein the first and second signals each comprise a plurality of chirps, and each signal comprises a FMCW signal within each chirp being phase-modulated by a respective code-division multiple access, CDMA, phase code to produce respective first and second PM-FMCW signals; receiving, at a receiver, a composite signal comprising at least one reflection of the first PM-FMCW signal and the second PM-FMCW signal; processing, in the digital domain, a plurality of groups of down-converted and pre-processed samples of the composite signal by: applying a frequency response correction to the groups of down-converted and pre-processed samples to provide corrected samples; applying a group delay filter to align the CDMA phase codes of the corrected samples to provide time-aligned samples; removing the respective CDMA phase codes from the time-aligned samples to provide respective groups of demodulated samples; transforming the respective groups of demodulated samples to a range-Doppler domain by applying a two-dimensional fast Fourier transform, 2D-FFT; and removing phase code cross-correlation residuals from the respective groups of transformed demodulated samples by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals.

[0022] In one or more embodiments, removing the respective CDMA phase codes comprises multiplying the group-corrected frequency-response-corrected sample groups with a conjugate of the PM-FMCW signal, in other words, with the phase code used to modulate the chirp to obtain the PM-FMCW signal.

[0023] In one or more embodiments, the first and second signals each comprise a plurality of repeated frames of chirps, and each chirp within a frame has a different CDMA phase code. BRIEF DESCRIPTION OF DRAWINGS

[0024] Reference will now be made to the drawings, in which:

[0025] Figure 1 A radar system according to one or more embodiments is shown schematically in a high-level block diagram;

[0026] Figure 2 A radar system of Figure 1 is shown in more detail;

[0027] Figure 3 A radar system of Figure 1 is shown in yet more detail;

[0028] Figure 4 A process flow diagram is shown in accordance with embodiments of the present disclosure;

[0029] Figure 5 A simplified block diagram of a residual cancellation method is shown in accordance with embodiments of the present disclosure;

[0030] Figure 6 A flow diagram of the method of Figure 5 is shown;

[0031] Figure 7 A block diagram of a general residual cancellation method is shown in accordance with embodiments of the present disclosure;

[0032] Figure 8 Examples of fast time phase history before and after applying a group delay filter compared to the transmitted signal are shown; and

[0033] Figure 9 Some statistical analysis of achievable dynamic range in accordance with embodiments of the present disclosure is shown.

[0034] It should be noted that the drawings are diagrammatic and not drawn to scale. Relative dimensions and proportions of parts of these drawings have been shown exaggerated or reduced in size, for the sake of clarity and convenience in the drawings. The same reference numerals are generally employed to denote the same elements throughout the drawings. DETAILED DESCRIPTION

[0035] As described in the introduction, the use of CDMA in frequency modulated continuous wave (FMCW) radars is typically limited to applying codes between chirps (in so-called "slow time") and the code length is thus constrained by the number of chirps per frame. The present disclosure relates to providing codes for CDMA within individual chirps (that is, in so-called "fast time") and addresses the orthogonality ratio problem that ensues.

[0036] Since the available code length within individual chirps can not be sufficient, the encoding is typically extended across chirps as well as within individual chirps. Joint slow time and fast time encoding can achieve better separability of the transmitted channels. As will be discussed in more detail below, although two-dimensional residual cancellation adds some signal processing, two-dimensional iterative residual cancellation of codes in conjunction with this can improve the achievable dynamic range of the system.

[0037] It should be noted that the application of codes defined by phase modulation (PM) is in addition to the frequency modulation of the FMCW signal, that is to say in combination with the frequency modulation of the FMCW signal and not instead of the frequency modulation of the FMCW signal. Thus, the signals according to the present disclosure can be described as PM-FMCW signals. Thus, the present disclosure does not relate to pure PMCW signals. Pure PMCW signals would require a completely different system architecture than the system architecture of FMCW automotive radars.

[0038] The skilled person will be aware that, in the context of radar systems, a chirp can be defined as a continuous wave signal or tone which has a frequency modulated over time of the chirp, typically from a lower starting frequency to a higher ending frequency for an “up-chirp” or from a higher starting frequency to a lower ending frequency for a “down-chirp”. At the end of the chirp, the frequency can be immediately reset to the starting frequency to provide a “sawtooth” in frequency. A set of such chirps can be described as a frame.

[0039] Thus, according to aspects of the present disclosure, the transmitted signal combines linear frequency modulation (LFM), which generates a “chirp”, with a further phase modulation over fast time, that is to say within the chirp. The code applied to an individual chirp is typically different from the code applied to other chirps. Each chirp can be considered as a set of sub-elements or chips, and in the case of binary phase encoding, each chip has one of two relative phases. The skilled person will appreciate that the term “phase” used herein can refer to a relative phase compared to a co-frequency signal, such that if the signals are in phase, the relative phase is 0°, if the signals are quadrature, the relative phase is 90°, if the signals are anti-phase, the relative phase is 180°, and so on, depending on the context. Although binary encoding is the simplest encoding conceptually, in practical applications the sharp transitions associated with binary encoding can be problematic, and alternative encoding schemes such as higher-level encoding or GMSK (Gaussian Minimum Shift Keying) can be preferred as the skilled person will be familiar. The present disclosure does not rely on any particular type of encoding of the phase code. Although in some embodiments the code applied to an individual chirp is typically derived from the code applied to other chirps within the frame, in other embodiments each chirp of the frame has the same CDMA phase code. In such embodiments, a second CDMA code is typically applied across the chirps of the frame. The combination of the CDMA phase code and the second CDMA code uniquely identifies the chirps from each transmitter across the frame.

[0040] The encoding is different for each Tx element of the system; Tx elements can be on the same integrated circuit (IC) or chip, or across multiple cascaded ICs; the code is used to distinguish signals from each Tx element that are received together on each Rx element, thus creating a virtual MIMO channel. A group delay filter in the receive processor is used to time-align the signals from each of the range cells. As is familiar to those skilled in the art, the entire range (i.e., the maximum range at which the radar can detect reflections from a target) is divided or segmented into an appropriate number of sub-ranges or range cells. The group delay filter provides compensation for the time delay associated with multiple targets at multiple different distances from the radar. This simplifies or reduces the processing complexity in the rest of the receive processing flow, particularly the demodulation processing complexity used to remove the code from the signal. Signal encoding and decoding are typically performed in both fast and slow time, allowing CDMA codes to be removed within each chirp and across chirs in the same processing stage. Subsequently, code residuals resulting from cross-correlation are eliminated to improve dynamic range. Code residual elimination can be a single-pass elimination process or can be applied via multiple passes or iterations.

[0041] Now turn to the attached image. Figure 1 A radar system 100 according to one or more embodiments is schematically illustrated in a high-level block diagram. The system includes a chirp generator 110 that applies frequency modulation to a continuous signal or tone provided from a reference local oscillator (LO) 112. In automotive applications, the tone can typically be in one of the frequency bands assigned to the automotive radar, such as 76 to 81 GHz, and approximately 140 GHz. The chirp generator can modulate the tone over a bandwidth of, for example, 250-750 MHz, but for some applications, the modulation can be wider, such as up to 4 GHz. The FMCW signal generated by the chirp generator 110 is applied to a Tx modulator module 120. The Tx modulator 120 imposes phase modulation onto the chirp according to the CDMA system (described in more detail below). The CDMA code is unique for each transmitter channel. The signal for each of the K transmitters is amplified via corresponding power amplifiers (PAs) 132, 134 and transmitted from the corresponding transmit antenna TX0 142 to TX0 142. K-1 Broadcast 144.

[0042] On the receiving side of radar system 100, a group of L antennas Rx0 to Rx0 are shown at positions 152, 154…156. L-1The receive antennas are each connected to a respective RF front end (RFE) 162, 164... 166 of the receiver front end 160. The respective RFEs process the signals in the analog domain (again as will be described in more detail below), and convert the processed signals to groups of digital samples. The groups of digital samples from each RFE are passed to a digital signal processor DSP 170 or equivalent for further signal processing, again as will be described in more detail below. The skilled person will appreciate that the DSP functionality can be implemented within one processor, unit or device, or can be distributed across more than one processor, unit or device.

[0043] Turning to Figure 2 , Figure 2 The radar system 200 can correspond to the radar system shown in Figure 1 The Tx modulator 120 comprises a phase shift generator 222 which receives a signal from the reference LO 112 and applies a further respective time dependent phase shift to the signal to be amplified by the respective power amplifier 132... 134 at a respective mixer 226... 228 where m denotes the slow time "chirp index" within a chirp, and t is the "fast time". This is generally referred to herein as phase encoding or phase code; thus, the phase encoding is applied both within a chirp and across multiple chirps of each radar frame. The phase shift or phase rotation can be applied by known processing techniques.

[0044] Figure 2 More detail is also shown of one 162 of the RF front ends of the receiver front end 160. In particular, the signal received at the receive antenna Rx0 152 is amplified by a low noise amplifier 262 (LNA) before being mixed at a mixer 264 with a copy of the unmodulated chirp generated by the chirp generator 110. Mixing the amplified received signal with the unmodulated chirp from the transmitter can be seen as "down-converting" the signal, and this is widely employed in wireless communications and radar applications to recover information or signals. However, in the case of FMCW radar applications, the result of mixing the amplified received signal with the (unmodulated) transmitted signal is to isolate the frequency difference (which can also be referred to as the beat frequency) between the transmitted signal and the received signal: since the transmitted signal is a chirp which typically has a linear frequency modulation (LFM), the beat frequency corresponding to the frequency difference between the transmitted signal and the received signal depends on the total time of flight of the transmitted signal to and from the reflector (or target), which in turn depends on the distance between the radar and the reflector.

[0045] The analog signal is then cleaned up, e.g. filtered by a high pass filter (HPF) 266 for direct leakage suppression, followed by a low pass filter (LPF) 268 for anti-aliasing, and removes any frequencies and components from the mixed signal, leaving only the frequency difference or beat signal. This is then converted to the digital domain by means of an analog-to-digital converter (ADC) 270, which samples the data at a predefined acquisition interval. The ADC must have a sufficient sampling rate to sufficiently capture the frequency and phase of the received signal. It typically has a sampling rate of at least 10 times the frequency of the reference oscillator 112. The digital samples of the signal are then passed to the DSP 170 together with corresponding pre-processed digital signals from other receive channels originating from the respective antennas 154...156.

[0046] The processing in the DSP 170 can be referred to Figure 3 In summary, Figure 3 corresponding to Figure 1 and Figure 2 a radar system 300 of the radar system shown in

[0047] First, at 310, the frequency response (amplitude and phase) of the digitized analog incoming signal is corrected for proper signal decoding. Then, at 312, a group delay filter is applied to achieve proper alignment of the responses of all targets, independent of their distance, for proper decoding in fast time. The signal is replicated or copied K times to produce K signals, each of which can be processed according to a different transmit channel Tx. Then, at 314, the code is removed, typically by multiplication with the conjugate of the transmitted code in the corresponding Tx channel, and at 316, a two-dimensional fast Fourier transform (2D FFT) from time samples and channels to range and velocity (“Doppler”) is applied. The output is used in an iterative 2D residual cancellation block 318, where 2D residual cancellation is applied at least once, and can be applied multiple passes to improve the dynamic range or OR of the CDMAMIMO. The cleaned up output can then be derived for Direction of Arrival (DOA) processing, detection and creation of a point cloud, as shown at 320.

[0048] Thus, the received signals are first pre-processed and digitized in the analog domain in the receiver front end 160; then, they are processed in the digital domain, e.g. within the DSP 170.

[0049] In particular, reference is made to Figure 4And considering the simplest case where there are only two Tx elements 142, 144 (each Tx element transmits a PM-FMCW signal reflected from the target to provide a single receiver with a front-end 162 with first and second reflected signals, the front-end 162 converts and pre-processes the reflected signals as described above), the down-converted pre-processed versions of the first and second reflected signals are processed in the digital domain by: applying a frequency response correction to the down-converted pre-processed versions of the first and second reflected signals at step 410 to provide a corrected sample group; applying a group delay filter to align the phase codes of the corrected sample group at step 420 to provide a time-aligned sample group; removing the respective CDMA phase codes from the time-aligned sample group at step 430 to provide respective demodulated signals; transforming the respective demodulated signals to the range-Doppler domain (where range corresponds to the “fast frequency” and Doppler corresponds to the “slow frequency”) by applying a two-dimensional fast Fourier transform 2D-FFT at step 440; and removing the phase code cross-correlation residuals from the respective transformed demodulated signals by calculating the estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals at step 450.

[0050] Now consider the case where there are three or more Tx elements Figure 5 and Figure 6 ; Figure 5 A block diagram illustrating the process of removing the phase code cross- correlation residuals from the respective transformed demodulated signals in the simplest case where the received signals include reflections from only two Tx elements is shown; Figure 6 A corresponding flowchart is shown for the case of three or more Tx elements. The cross-correlation residuals arise from the imperfect orthogonality of the phase codes: since they are not infinite in length, the phase codes are not perfectly orthogonal, and so the signal in the first channel partially leaks into the other channels, producing a “residual” of the first channel signal in the other channels.

[0051] In summary, a first set of targets is identified and isolated, and it is assumed that these targets provide a reflection component to the signal from each of the transmitted signals. Individual targets in the first set of targets can also be referred to as high probability target points or strong spectral peaks. For each channel, the impact of those targets on each of the remaining channels can be determined (or at least estimated) and the "residuals" removed from the signal. The original signal, or more accurately, the original sample group (as the processing is in the digital domain), is replaced with the sample group with the residuals removed, resulting in a "cleaner" sample group. Since the resulting sample group includes lower contamination (the residuals from the already identified targets have been removed), if the same detection mask that was used to identify the first set of targets is reapplied, additional targets can now be revealed. The process can be applied iteratively by estimating and removing additional cross-correlation residuals (and so on) produced by those additional targets.

[0052] At step 610, and as shown at 510 on the block diagram, a first set of targets is identified. This can be done, for example, by incoherent integration as shown at 512, and then applying a detector at 514. This can be done by known techniques, for example by using a Constant False Alarm Rate (CFAR) detector. The resulting detection mask includes the first set of targets. The first set of targets is represented in the time domain as As shown at 620. The residuals in the second channel are then determined from the signal in the first channel of the identified target at 530 and 630 (which results from the imperfect orthogonality of the channel phase codes), just as the corresponding residuals in the first channel The cross-correlation coefficients between the two channels can be pre-computed and As shown at 532 and 632. It will be appreciated that and are the conjugate transpose of each other, and the cross-correlation coefficients are equivalent. Applying another 2D FFT transforms the residuals and into the frequency-Doppler (i.e., range-velocity) domain as and As shown at 540 and 640. The residuals can then be subtracted from the original sample group to eliminate the residuals associated with the identified targets from the data.

[0053] The skilled person will appreciate that Figure 5 The example shown in FIG. 5 with only two transmitters represents the simplest case. Figure 7 A more representative case in an example with K transmitters is shown in FIG. 6. The cross-correlation coefficients between the two channels are Figure 7 andFigure 5 Comparing, the main difference is that for each transmitter there are residual elements from more than one other channel. To eliminate the residual elements from each other channel, these residual elements must be combined, which is done in the time domain as shown at 532 for each channel. Again, the skilled person will appreciate that since channels are shown, there are two cross-correlation residual elements to be combined in each mixer 532; in the general case with K channels, K-1 cross-correlation elements residual elements will be mixed in each mixer 532. Recall that K is at least 2, and typically K is 8 or 16.

[0054] For a more complete understanding of the present disclosure, the method described above will now be considered in more detail with a more complete mathematical basis: the MIMO radar 300 has K simultaneously active channels transmitting M linear frequency modulated chirps bursts:

[0055]

[0056] where f c denotes the carrier frequency, β is the chirp slope, m is the chirp (slow time) index and t is the fast time, and k = 0,.., K-1 is the transmitter index. The phase encoding φ k (m, t) is applied in addition to the chirp, and can be seen as a two-dimensional array per Tx showing the fast time modulation applied per sequential chirp index m in fast time.

[0057] The observed scenario can be represented by a linear combination of multiple point-like targets embedded in system noise (for notational simplicity, the derivation is listed per single Rx channel, the following equations can be generalized to multiple Rx channels):

[0058]

[0059] The received signal is mixed with an unmodulated copy of the chirp to obtain the beat signal, which can be described by the following expression (where a narrowband approximation is applied):

[0060]

[0061] Here, the range R1of each target is embedded in its time delay:

[0062] τ i = 2R i / c; (5)

[0063] The velocity v i produces a corresponding Doppler frequency

[0064]

[0065] and the angle is given by the spatial frequency The signals are represented (for simplicity, Tx is represented as a uniform linear array - ULA).

[0066] Finally, n(m,t) defines the thermal noise of the system.

[0067] The first step of the signal processing consists in using a group delay filter and aligning the code for the data captured in a predefined acquisition interval. It is assumed that the modulated signal is sufficiently narrow band (typically the modulation bandwidth is lower than 10% of the ADC sampling frequency, B φ ≤ 0.1f s and its spectrum decays rapidly outside its frequency band, then the group delay filter should align the signal so that the effect of the time delay is attenuated:

[0068]

[0069] Note that the group delay filter can be implemented by various different procedures, for example, via an FFT in the frequency domain and a phase correction or via a digital filter implementation. Figure 8 An example of the fast time phase history before (at 810) and after (at 820) applying the group delay filter is shown compared to the transmitted signal (830). Note that before applying the group delay filter, the Rx signal is not aligned with the Tx signal. Without applying the group delay filter, this would lead to incorrect decoding and low dynamic range. After applying the group delay filter, the Rx signal is aligned with the Tx signal. Figure 8 In the example shown, the phase encoding φ k (m,t) is a GMSK modulation over fast time, with 32 chips per chirp.

[0070] The signal copy is then copied the same number of times as the number of Tx used in the CDMA mode, and the virtual Tx is represented by:

[0071]

[0072] The The virtual channel thus has a signal with a term corresponding to the target response of this virtual Tx channel (first sum in the equation below) and a cross-correlation term from the other Tx channels (second sum in the equation below):

[0073]

[0074] In this expression

[0075]

[0076] is the code k and ​2D cross-correlation between them. Note that during the presence of the cross-correlation term of the strong signal, weak targets near the strong target can be hidden by the cross-correlation response of the strong target.

[0077] Applying a 2D FFT to the acquired data provides a range-Doppler representation of the scene:

[0078]

[0079] In this expression, v denotes the fast time frequency (range) and f denotes the Doppler frequency, which is proportional to the target radial velocity; furthermore, is the cross-correlation function of the 2D FFT.

[0080] According to the present disclosure, the proposed 2D residual estimation algorithm has the steps already outlined above Figure 5 and Figure 6 with respect to the steps outlined above. This will now be restated with appropriate mathematical derivations. First, at 610, a detector such as Constant False Alarm Rate (CFAR) is applied to the combination of channels (e.g., via non-coherent summation). Then the detection mask M(f,v) in range and Doppler is obtained:

[0081]

[0082] Next, at 620, then the time-domain representation of the detected targets for each Tx channel is obtained by applying a 2D IFFT to the product of the current range-Doppler plot and the detection mask:

[0083]

[0084] Next, at 630, for each Tx channel, the impact of the cross-correlation from the other active CDMA-coded Tx channels is evaluated in the time domain: although the simplest case of only two Tx elements (and thus two virtual channels, making ) is considered, this can be generalized to any suitable value of :

[0085]

[0086] At 640, it is transformed to the frequency domain by applying a 2D FFT, and at 650, it is subtracted from the original range-Doppler plot:

[0087]

[0088] In case multiple iterations are applied, this range-Doppler plot can be used as the output of the residual cancellation technique, or it can be used as the input for the next iteration.

[0089] Turning now to Figure 9 which shows some statistical analysis of achievable dynamic range according to embodiments disclosed herein. For the statistical simulation, a frame length of 128 chirps was used, and a signal sampling rate of N = 512 (i.e., 512 samples per chirp provided by each ADC, which corresponds to a 12.8 ps acquisition time at Fs = 40 MHz). A noiseless scenario with 32 targets was considered, where the MIMO radar has 4 transmitters and 4 receivers, ending 100 trials. The skilled person will appreciate that in the noiseless case, the dynamic range limitation is exhibited due to CDMA cross-correlation, and this is independent of the cross-section or magnitude of the targets. The minimum, average, and maximum side lobe levels (SLLs) - as used herein, the SLLs correspond to the cross-correlation residuals - are shown at 910, 920, and 930 in the case of regular “ST-CDMA”, where CDMA is applied only in the slow time (i.e., the encoding is applied across the chips of a frame, but not within the individual chirps, such that each bit is the length of one chirp). The minimum, average, and maximum SLLs according to embodiments of the disclosure are shown at 940, 950, and 960, such that the CDMA code is applied across both the fast and slow times, that is, the code bits correspond to individual chips of each chirp, and the code extends over multiple chirps. Note that due to the use of longer codes in PM-FMCW, it provides a gain of 6-10 dB relative to ST-CDMA in the initialization and first iteration, and the difference diminishes as the number of iterations increases. It appears in practice that for small to moderately sparse scenarios, each iteration provides a gain of 5-10 dB in the dynamic range that can be used to predict the number of iterations needed.

[0090] From reading the present disclosure, other variations and modifications will become apparent to the skilled person. Such variations and modifications can involve equivalent and other features, which are already known in the art of MIMO CDMA radar systems and processes and which can be used instead of or in addition to features already described herein.

[0091] Although the appended claims are directed to particular combinations of features, it will be understood that the scope of the disclosure of the present application extends to any novel one, or any novel combination, of the features disclosed herein, explicitly or implicitly, whether or not it mitigates any of the same technical problems as does the presently claimed application or it mitigates the same technical problems in a different manner from does the presently claimed application.

[0092] Features described in the context of separate embodiments can also be provided in combination in a single embodiment. Conversely, various features described in the context of a single embodiment can also be provided separately or in any suitable sub-combination. Applicants hereby give notice that new claims can be formulated to

[0093] For the avoidance of doubt, the term "comprising" does not exclude other elements or steps, the term "a" or "an" does not exclude a plurality, single processing units or other units can fulfill several functions if such units are not explicitly described as being functionally independent. Further, features described in the context of separate embodiments can also be provided in combination in a single embodiment. Conversely, various features described in the context of a single embodiment can also be provided separately or in any suitable sub-combination. The term "a" or "an" is used to introduce one or more items, i.e., it is used in the sense where the item(s) can be singular or plural. The term "plurality" is used to introduce a single item or multiple items, i.e., it is used in the sense where the item(s) can be singular or plural. The term "or" is used to link words in the alternative, e.g., the phrase "a or b" means "at least one of a or b". The terms "include", "including", "includes" and / or "contain", "containing", "contains" and / or "comprise", "comprising", "comprises" and / or "comprised of" are used synonymously to refer to open ended inclusion. The term "coupled" is used to indicate that two or more elements, which can or can not be in physical contact, act together or in connection with one another. The term "connected" is used to indicate the most direct physical or electrical connection between two elements. The terms "erect" and "vertical" are used synonymously to refer to a position where the normal to the plane of the element is perpendicular to the plane of the earth. The terms "horizontal" and "lateral" are used synonymously to refer to a position where the normal to the plane of the element is parallel to the plane of the earth. The terms "top", "bottom", "upper", "lower", "front", "back", "side", "end", "side", "front", "rear", "left", "right", "up", "down", "vertical", "horizontal", "lateral", "erect", "above", "below", "upward", "downward", "upwards", "downwards", "upwardly", "downwardly", "upwardsly", "downwardsly", "top", "bottom", "upper", "lower", "front", "back", "side", "end", "side", "front", "rear", "left", "right", "up", "down", "vertical", "horizontal", "lateral", "erect", "above", "below", "upward", "downward", "upwards", "downwards", "upwardly", "downwardly", "upwardsly", "downwardsly", "

Claims

1. A method of operating a multiple-input multiple-output, MIMO, radar system for automotive applications, characterized in that, The method comprises, transmitting a first frequency modulated continuous wave, FMCW, signal from a first transmitter; transmitting a second FMCW signal from a second transmitter; wherein the first signal and the second signal each comprise a plurality of chirps, and each signal comprises a phase modulation of the FMCW signal within each chirp by a respective code division multiple access, CDMA, phase code to produce respective first and second PM-FMCW signals; receiving, at a receiver, a composite signal comprising at least one reflection of each of the first and second PM-FMCW signals; and processing, in the digital domain, a plurality of downconverted and pre-processed sample groups of the composite signal by: applying a frequency response correction to the downconverted and pre-processed sample groups to provide corrected sample groups; applying a group delay filter to align the CDMA phase codes of the corrected sample groups to provide time-aligned sample groups; removing the respective CDMA phase codes from the time-aligned sample groups to provide respective demodulated sample groups; transforming the respective demodulated sample groups to a range-Doppler domain by applying a two-dimensional fast Fourier transform, 2D-FFT; and removing phase code cross-correlation residuals from the respective transformed demodulated sample groups by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals.

2. The method of claim 1, wherein, Removing phase code cross-correlation residuals from each respective transformed demodulated signal comprises a residual cancellation iteration comprising: applying a detection mask to the respective transformed demodulated signal, thereby isolating high-probability target points; applying a 2D inverse FFT, 2D-IFFT, resulting in an isolated target signal; estimating respective estimated cross-correlation components of each other transformed demodulated signal from the isolated target signal and combining them to produce an estimated residual signal; and applying a 2D FFT to the estimated residual signal to produce a transformed estimated residual signal; and subtracting the transformed estimated residual from the respective transformed demodulated signal.

3. The method of claim 2, wherein, Estimating respective estimated cross-correlation components in each of the other transformed demodulated signals from the transformed reduced target signal comprises: multiplying the other transformed demodulated signals by pre-computed cross-correlation coefficients between the other signals and the signal.

4. The method of claim 2, wherein, Further comprising another residual cancellation iteration.

5. A CDMA automotive MIMO radar system, characterized by Comprising: a plurality of transmitters each configured to transmit a radar frame comprising a plurality of chirps having frequency modulation therein, wherein the frame from each transmitter comprises a code division multiple access, CDMA, encoding according to which each of the plurality of chirps comprises a phase modulation within; a plurality of receivers; a respective RF front-end processor for each of the plurality of receivers and configured to provide a group of downconverted and pre-processed digital samples to a digital signal processor; and a digital signal processor configured to: applying a frequency response correction to the downconverted and pre-processed sample groups to provide corrected sample groups; applying a group delay filter to align the CDMA phase codes of the corrected sample groups to provide time-aligned sample groups; removing the respective CDMA phase codes from the time-aligned sample groups to provide respective demodulated sample groups; transforming the respective demodulated sample groups into a range-Doppler domain by applying a two-dimensional fast Fourier transform (2D-FFT); and removing phase code cross-correlation residuals from the respective transformed demodulated sample groups by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals.

6. The CDMA automotive MIMO radar of claim 5, wherein, the processor is configured to remove phase code cross-correlation residuals from each respective transformed demodulated signal, including performing a residual cancellation iteration by: applying a detection mask to the respective transformed demodulated signal, thereby isolating high-probability target points; applying a 2D inverse FFT (2D-IFFT) to obtain an isolated target signal; estimating respective estimated cross-correlation components of each other transformed demodulated signal from the isolated target signal and combining them to produce an estimated residual signal; and applying a 2D FFT to the estimated residual signal to produce a transformed estimated residual signal; and subtracting the transformed estimated residual from the respective transformed demodulated signal.

7. The CDMA automotive MIMO radar of claim 5, wherein the processor is configured to perform a plurality of residual cancellation iterations.

8. A method of operating a multiple-input multiple-output, MIMO, radar system for automotive applications, characterized in that, the method includes, transmitting a first frequency modulated continuous wave (FMCW) signal from a first transmitter; transmitting a second FMCW signal from a second transmitter; wherein the first and second signals each include a plurality of chirps, and each signal includes phase modulating the FMCW signal within each chirp by a respective code division multiple access (CDMA) phase code to produce respective first and second PM-FMCW signals; receiving a composite signal at a receiver, the composite signal including at least one reflection of the first and second PM-FMCW signals; processing a plurality of downconverted and pre-processed sample groups of the composite signal in a digital domain by: applying a frequency response correction to the downconverted and pre-processed sample groups to provide corrected samples; applying a group delay filter to align the CDMA phase codes of the corrected samples to provide time-aligned sample groups; removing the respective CDMA phase codes from the time-aligned sample groups to provide respective demodulated sample groups; transforming the respective demodulated sample groups into a range-Doppler domain by applying a two-dimensional fast Fourier transform (2D-FFT); and removing phase code cross-correlation residuals from the respective transformed demodulated sample groups by computing estimated residuals and subtracting the estimated residuals from the respective transformed demodulated signals.

9. The method of claim 8, wherein Removing the respective CDMA phase code includes multiplying a group of frequency response corrected samples that are group corrected with a conjugate of the PM-FMCW signal.

10. The method of claim 9, wherein The first and second signals each include a plurality of repeated frames of chirps, and each chirp within a frame has a different CDMA phase code.

Images