Phase modulated continuous wave radar systems

Abstract

A phase modulated radar system with a plurality of transmitters and a plurality of receivers. The transmitters transmit phase modulated radio signals and the receivers receive phase modulated radio signals that include signals transmitted by the transmitters and reflected from objects in an environment. The phase modulated radio signals are transmitted with a stepped carrier signal implemented using a plurality of cascading bursts for each scan, with each burst transmitted with a different carrier frequency. By cascading a plurality of consecutive bursts together over the course of a scan, the step carrier achieves a bandwidth for the phase modulated radio signals. Respective bandwidths of the individual transmission bursts are smaller than the bandwidth of the scan, and with the bandwidth of the scan including a quantity of received transmission bursts, each with a respective burst bandwidth.

Description

PHASE MODULATED CONTINUOUS WAVE RADAR SYSTEMSCROSS REFERENCE TO RELATED APPLICATION

[0001]

[0001] The present application claims the filing benefits of U.S. provisional application, Ser. No. 63 / 671 ,936, filed Jul. 16, 2024, which is hereby incorporated by reference herein in its entirety.FIELD OF THE INVENTION

[0002] The present invention is directed to radar systems, and in particular to multipleinput, multiple-output (MIMO) radar systems for vehicles and robotics.BACKGROUND OF THE INVENTION

[0003] The use of radar to determine range, velocity, and angle (elevation or azimuth) of objects in an environment is important in a number of applications including automotive radar and gesture detection. Radar systems typically transmit a radio frequency (RF) signal and listen for the reflection of the radio signal from objects in the environment. A radar system estimates the location of objects, also called targets, in the environment by correlating delayed versions of the received radio signal with the transmitted radio signal. A radar system can also estimate the velocity of the target by Doppler processing. A radar system with multiple transmitters and multiple receivers can also determine the angular position of a target. Depending on antenna scanning and / or the number of antenna / receiver channels and their geometry, different angles (e.g., azimuth or elevation) can be determined.

[0004] A radar system consists of transmitters and receivers. The transmitters generate a baseband signal, which is upconverted to a radio frequency (RF) signal that propagates according to an antenna pattern. The transmitted signal is reflected off object or targets in the environment. The received signal at each receiver is the totality of the reflected signal from all targets in the environment. The receiver down converts the received signal to baseband and compares the baseband received signal to thebaseband signal at one or more transmitters. This is used to determine the range, velocity, and angle of targets in the environment.SUMMARY OF THE INVENTION

[0005] Exemplary embodiments provide methods and a system for a phase modulated radar system that includes a carrier signal that is a stepped carrier. The stepped carrier is implemented using a plurality of cascading bursts for each scan. That is, by cascading a plurality of bursts together over the course of a scan, the step carrier can achieve the same bandwidth as a conventional wide bandwidth PMCW signal. In another embodiment, the phase modulated continuous wave transmissions are encoded with the use of a combination of code division multiplexing (CDM) and Doppler division multiplexing (DDM). As such, the phase modulated radar would include a plurality of antenna elements partitioned into code groups and then sub-partitioned into Doppler groups. For transmission of a phase modulated continuous wave signal that is coded using CDM+DDM, the radar system models the transmission output noting the number of CDM channels and the number of DDM channels.

[0006] A radar system in accordance with an embodiment of the present invention includes a transmitter pipeline and a receiver pipeline. The transmitter pipeline includes a plurality of transmitters, each transmitting transmit phase modulated radio signals. The receiver pipeline includes a plurality of receivers, each receiving modulated radio signals that include signals transmitted by the plurality of transmitters and reflected from objects in an environment. The transmitters transmit the phase modulated radio signals with a stepped carrier frequency comprising a plurality of different carrier frequencies. Each transmitter transmits the phase modulated radio signals as a plurality of transmission bursts, each burst transmitted with a different carrier frequency. Each receiver receives the phase modulated radio signals during a radar scan. The respective bandwidth of each individual transmission burst is smaller than the bandwidth of the radar scan. The bandwidth of the radar scan comprises a quantity of received transmission bursts each with a respective burst bandwidth. Thus, by cascading a plurality of bursts together over the course of a scan, the step carrier can achieve the same bandwidth as a conventional wide bandwidth PMCW signal.

[0007] A radar system in accordance with another embodiment of the present invention includes a transmitter pipeline and receiver pipeline. The transmitter pipeline includes phase modulated continuous wave transmissions. The transmitter pipeline’s phase modulation utilizes digital codes that include a combination of code division multiplexing (CDM) and Doppler division multiplexing (DDM). The phase modulated radar comprises a plurality of antenna elements partitioned into code groups and then subpartitioned into Doppler groups. For transmission of phase modulated signal that is coded using CDM+DDM, the radar system models the transmission output noting the number of CDM channels and the number of DDM channels.

[0008] In an aspect of the present invention, the number of DDM segments include one or more blank segments where no transmission takes place, the blank segments present in the scan to assist the disambiguation of Doppler ambiguities in the transmission and reception arising from DDM.

[0009] In another aspect of the present invention, the radar scan comprises a scan duration. Each transmission burst of the quantity of received transmission bursts comprises a burst duration. A sum of the burst durations of the quantity of received transmission bursts equals the scan duration. Each of the plurality of transmitters is configured to transmit the plurality of transmission bursts such that the respective carrier frequencies of the plurality of transmission bursts are incrementally stepped in frequency. The carrier frequencies of adjacent transmission bursts in time at least partially overlap.

[0010] In a further aspect of the present invention, each of the plurality of transmitters are configured to transmit each transmission burst such that adjacent transmission bursts in time do not overlap in their respective burst duration.

[0011] In another aspect of the present invention, each of the plurality of transmitters are configured to transmit each transmission burst such that two separate transmission bursts have different burst durations.

[0012] In an aspect of the present invention, each of the plurality of receivers is configured to process the received phase modulated radio signals to determine a range to a detected object according to a corresponding coarse range bin for each transmission burst. The plurality of receivers are configured to use the resulting coarse range bins for a plurality of transmission bursts to determine fine range bins for the scan duration, and wherein each coarse range bin comprises a plurality of fine range bins.

[0013] In another aspect of the present invention, each of the plurality of receivers is configured to perform Doppler processing on the fine range bins of the scan duration. The Doppler processing comprises Doppler transforming using a fast Fourier transform (FFT) across the fine range bins of the scan duration.

[0014] In yet another aspect of the present invention, each of the plurality of receivers is configured to perform Doppler processing on the coarse range bins using a Doppler transform across every pulse for each coarse range bin.

[0015] In another aspect of the present invention, each of the plurality of receivers is configured to perform Doppler processing on the received phase modulated radio signals using a fast Fourier transform (FFT) while the burst duration of every step carrier is adjusted to compensate for the changing of the frequency.BRIEF DESCRIPTION OF THE DRAWINGS

[0016] FIG. 1 is a plan view of an automobile equipped with a radar system in accordance with the present invention;

[0017] FIG. 2A and FIG. 2B are block diagrams of radar systems in accordance with the present invention;

[0018] FIG. 3 is a block diagram illustrating a radar with a plurality of receivers and a plurality of transmitters (MIMO radar) in accordance with the present invention;

[0019] FIG. 4 is a schematic of an exemplary notional bandwidth illustrating an exemplary spectral shape of the waveform in accordance with the present invention;

[0020] FIGS. 5A, 5B, and 5C illustrate exemplary waveforms of different modulation methods including phase modulation, phase modulation, and step frequency phase modulation in accordance with the present invention;

[0021] FIG. 6 is a block diagram of a data path illustrating an integration process for an exemplary step carrier PMCW in accordance with the present invention;

[0022] FIG. 7 is a diagram illustrating an exemplary indexing of fast Fourier transform (FFT) inputs to fine range bins for an exemplary DFA embodiment for a step carrier PMCW in accordance with the present invention;

[0023] FIG. 8 is a signal diagram illustrating another exemplary DFA embodiment with a scaled PRI and three exemplary pulses for a step carrier PMCW illustrating exemplary fine range ambiguity coupled with several Dopplers, marked by velocities in accordance with the present invention;

[0024] FIG. 9 is a signal diagram illustrating the signal strength for respective fine range bins for a given coarse bin in accordance with the present invention;

[0025] FIG. 10 is a block diagram illustrating a block diagram of an exemplary transmit system model with CDM and DDM indices for each transmit ante4nna element in accordance with the present invention;

[0026] FIG. 11 is a diagram illustrating an exemplary digital code processing design with used and unused or blank segments in accordance with the present invention;

[0027] FIG. 12 is a diagram illustrating exemplary design specifications illustrating an arrangement of CDM channels and DDM channels in accordance with the present invention;

[0028] FIG. 13 is a diagram illustrating an exemplary arrangement of CDM channels and DDM channels for a modified or shifted version of the arrangement of FIG. 12;

[0029] FIGS. 14A and 14B are block diagrams illustrating an exemplary transmitter and receiver structure illustrating CDM+DDM signal processing in accordance with the present invention;

[0030] FIG. 15 is a signal diagram illustrating Doppler indexing for CDM channels versus DDM channels in accordance with the present invention;

[0031] FIG. 16 is a signal diagram illustrating the results of an exemplary signal processing design in accordance with the present invention;

[0032] FIG. 17 is a signal diagram illustrating the results of an exemplary embodiment with respect to detected strong targets versus weak targets in accordance with the present invention;

[0033] FIG. 18 is a signal diagram illustrating the results of exemplary embodiment with respect to a strong target and a weak target in accordance with the present invention;

[0034] FIGS. 19A and 19B are block diagrams illustrating an exemplary embodiment removing Doppler ambiguities such as a strong target at the Doppler ambiguity of a weaker target in accordance with the present invention.DESCRIPTION OF THE PREFERRED EMBODIMENTS

[0035] Referring to the drawings and the illustrative embodiments depicted therein, wherein numbered elements in the following written description correspond to like- numbered elements in the figures, an exemplary radar system with phase modulated continuous wave signal transmission and reception comprises a carrier signal that is a stepped carrier. The stepped carrier phase modulated continuous wave comprises a plurality of cascading bursts for each scan, with each burst having a relatively small bandwidth as compared to the bandwidth of the scan. That is, by cascading a plurality of bursts together over the course of a scan, the step carrier can achieve the same bandwidth as a conventional wide bandwidth PMCW signal. In another embodiment, an exemplary phase modulated radar utilizes digital codes that include a combination of code division multiplexing (CDM) and Doppler division multiplexing (DDM). The phase modulated radar comprises a plurality of antenna elements partitioned into code groups and then sub-partitioned into Doppler groups. For transmission of phase modulated signal that is coded using CDM+DDM, the radar system models the transmission output noting the number of CDM channels and the number of DDM channels. In one embodiment, the number of DDM segments include one or more blank segments where no transmission takes place, the blank segments present in the scan to assist the disambiguation of Doppler ambiguities in the transmission and reception arising from DDM.

[0036] FIG. 1 illustrates an exemplary radar system 100 configured for use in a vehicle150. The vehicle 150 may be an automobile, truck, or bus, etc. As illustrated in FIG. 1 ,the radar system 100 may comprise one or more transmitters and one or more virtual receivers 104a-104d. Other configurations are also possible. FIG. 1 illustrates receivers / transmitters 104a-104d placed to acquire and provide data for object detection and adaptive cruise control. The radar system 100 (providing such object detection and adaptive cruise control or the like) may be part of an Advanced Driver Assistance System (ADAS) for the automobile 150.

[0037] FIG. 2A illustrates an exemplary radar system 200 with an antenna 202 that is time-shared between a transmitter 206 and a receiver 208 via a duplexer 204. As also illustrated in FIG. 2A, output from the receiver 208 is received by a control and processing module 210 that processes the output from the receiver 208 to produce display data for the display 212. As discussed herein, the control and processing module 210 is also operable to produce a radar data output that is provided to other control units. The control and processing module 210 is also operable to control the transmitter 206. FIG. 2B illustrates an alternative exemplary radar system 250 with a pair of antennas 202a, 202b, a separate antenna 202a for the transmitter 206 and another antenna 202b for the receiver 208. While pulse radar systems may use shared or separate antennas, continuous wave radars (discussed herein) will use separate antennas (for transmitting and receiving) because of their continuous operation.

[0038] There are several types of signals used in radar systems. One type of radar signal is known as a frequency modulated continuous waveform (FMCW). In this type of system, the transmitter of the radar system sends a continuous signal in which the frequency of the signal varies across some range of frequencies. This is sometimes called a chirp radar system. At the receiver a matched filter can be used to process the received signal. The output of the matched filter is a so-called “pulse-compressed” signal with a pulse duration inversely proportional to the bandwidth used in the chirp signal. Mixing (multiplying) the reflected wave from a target with a replica of the transmitted signal results in a CW signal with a frequency that represents the distance between the radar transmitter / receiver and the target. By sweeping up in frequency and then down in frequency the Doppler frequency can also be determined.

[0039] The transmitted signal from each radar transmitter consists of a baseband signal which is upconverted to an RF signal by an RF upconverter followed by an antenna. The received signal at each radar receiver antenna is downconverted by an RF downconverter to a complex baseband signal. The baseband transmitted radio signals and the reflected radio signals after downconversion in the receiver are provided to the processor. As an example, a baseband signal used for transmission might consist of repeated sequences of random or pseudo-random binary values for one transmitter, e.g., (-1 , -1 , -1 , -1 , 1 , 1 , 1 , -1 , 1 , 1 , -1 , -1 , 1 , -1 , 1), although any sequence, including non-binary sequences and non-periodic sequences could be used and different sequences would be used for different transmitters. The use of truly random number generators and pseudo random number generators to produce the values used to phase modulate the radio signal before transmission is described in detail in U.S. Patent Application, Ser. No. 15 / 204,003, filed July 7, 2016, which is hereby incorporated by reference herein in its entirety.

[0040] The radar sensing system of the present invention may utilize aspects of the radar systems described in U.S. Pat. Nos. 10,261 ,179; 9,971 ,020; 9,954,955; 9,945,935; 9,869,762; 9,846,228; 9,806,914; 9,791 ,564; 9,791 ,551 ; 9,772,397; 9,753,121 ; 9,689,967; 9,599,702; 9,575,160, and / or 9,689,967, and / or U.S. Publication Nos. US-2017-0309997; and / or U.S. patent applications, Ser. No. 17 / 148,004, filed Jan. 13, 2021 , Ser. No. 16 / 674,543, filed Nov. 5, 2019, Ser. No. 16 / 259,474, filed Jan. 28, 2019, Ser. No. 16 / 220,121 , filed Dec. 14, 2018, Ser. No. 15 / 496,038, filed Apr. 25, 2017, Ser. No. 15 / 689,273, filed Aug. 29, 2017, Ser. No. 15 / 893,021 , filed Feb. 9, 2018, and / or Ser. No. 15 / 892,865, filed Feb. 9, 2018, and / or U.S. provisional applications, Ser. No. 62 / 960,220, filed Jan. 13, 2020, and Ser. No. 62 / 816,941 , filed Mar. 12, 2019, which are all hereby incorporated by reference herein in their entireties.Stepped Carrier Phase Modulated Continuous Wave Signal Use:

[0041] In conventional principles of radar operation, there is a known equation that relates the bandwidth used and the achievable range resolution. Usually the range resolution is measured by the range bin size (RBS) and the equation becomes: RBS = — (1).Where c is the speed of light and B is the bandwidth.

[0042] Note that in Equation (1), to achieve a certain RBS, the only available design parameter is the bandwidth. Interestingly, Equation (1) does not prohibit lengthy scan durations to achieve the total bandwidth.

[0043] As an example, in conventional frequency modulated continuous wave (FMCW) signals, the amount of time allotted to the FMCW to sweep across frequencies dictates its occupied bandwidth. Thus, for a given FMCW slope, the more time allotted, the more bandwidth was used by the waveform, therefore, the smaller RBS is achieved.

[0044] Following the technique of FMCW sweeping frequencies, the occupied carrier frequencies of a PMCW signal can also be time varying. From that insight, an exemplary stepped carrier PMCW signal is disclosed.Exemplary Theory of Operation:

[0045] Referring to FIG. 4, given an arbitrary PMCW spectra, the definition of a notional bandwidth B is necessary to establish its stepping operation. Unlike the true spectral shape of the waveform, an exemplary notional bandwidth B can be used to coarsely establish boundaries for each burst of PMCW transmissions.

[0046] With bandwidth B and burst duration T, the step carrier PMCW works by transmitting each burst at a different carrier frequency. With reference to FIGS. 5A, 5B, and 5C, the notion of carrier stepping (FIG. 5C) is illustrated in comparison to a conventional FMCW (FIG. 5A) and PMCW (FIG. 5B). Unlike conventional PMCW, the step carrier PMCW has a small bandwidth per burst duration. In total per scan, by cascading many bursts together, the step carrier can achieve the same bandwidth as the wide bandwidth PMCW counterpart.

[0047] As illustrated in FIG. 5C, the exemplary carrier PMCW bursts show no overlap of notional bandwidth, in fact, there is frequency overlap with its true bandwidth. In addition, the bandwidth of multiple bursts can exhibit frequency overlap, but not timeoverlap. Also note that the illustration of FIG. 5C shows similar burst durations of T, but this T is not required to be uniform across the step carriers.

[0048] On the reception of the stepped carrier PMCW signal, the common technique of time and spatial integration can serve as abstract algorithms to increase target signal- to-noise ratio (SNR) before detection. Equation (2), illustrated below serves as an abstract coherent integration algorithm. The outer integration for spatial domain accounts for all of the antenna elements in a commonly termed virtual array. Next, summation of the dynamics of target phases from its velocity occurs over each pulse repetition interval (PRI) of Tn, where the PRI may not be the same. Finally, there is an inner integration of the phase modulated code over each PRI starting at each range bin Rk.Note that the hat accent notation (i.e. , a caret-shaped symbol placed above a variable to give it special meaning) is the estimated unknown variable.

[0049] The abstract algorithms in Equation (2) provide a more complete model of the waveform characteristics, they are difficult to manipulate to further signify algorithm processing. A phasor decomposition model is easier to use for further development. The phasors are written for after range bin estimation, or range integration, as: 7hM)zfl(v)e7A. / < (3)From left to right in Equation (3), the first phasor is the result of time integration of the range autocorrelation function at the estimated range bin. This phasor is the resultant of the range integration over a pulse duration. Usually, the burst durations per (each) step are equal to the pulse duration. Again, the pulse durations are typically equal for every pulse in the scan, but this is not always the case.

[0050] Next is the Doppler or pulse phasor before Doppler processing. In this phasor, the pulse index n and the carrier index k are not assumed to be the same. For many scans, however, n = k is a simple waveform configuration. Following Doppler is the angle phasor, before beamforming, where the index v is for the virtual antennaelements. Of these three phasors above, the target reflection phase and the range delay phase that were present in Equation (2) are ignored. This is because they are not important for target detection in this embodiment.

[0051] There is a final term that should not be ignored, the phase term for the carriers during the stepping process. This phasor, 0c k, can be considered as a residue from the wcr(t) phases as the carrier steps in frequency. Depending on implementation, 0c kmay jump in a predictable manner that could be determined through a calibration process. For the remaining descriptions of these exemplary embodiments, the 0c kcalibration is assumed to be perfect and that residual phase effects from this phase are ignored.

[0052] Using the phasor model in Equation (3), the integration process for each phasor term continues as separable processes. A data path diagram representing the integration process for an exemplary step carrier PMCW is illustrated in FIG. 6.

[0053] In this exemplary design, each pulse, index k, is transmitted with a PMCW code and modulated at a specific carrier, fc k. The entire scan would contain many pulses which is some integer multiple of the number of stepped subbands, NB. At the receiver, after demodulation of the known carrier frequency for each pulse, the correlation module, CIU, outputs a sequence of code correlations per pulse. After correlating all pulses, the entire CIU output data set is referred to as RDC1 . Following RDC1 , there are three specific exemplary algorithms for stepped carrier PMCW processing embodiments: a Range Focus Algorithm (RFA), a Doppler Focus Algorithm, and a Fine Range Disambiguation. The Angle Processing algorithm and the Target Activation may use conventional techniques.

[0054] In addition to conventional techniques, the final algorithm of Fine Range Disambiguation (FRD) is needed after Target Activation. FRD is needed because RFA outputs fine range bins that are ambiguous with the target velocity. Thus, FRD is after Target Activation, such that the velocity can be estimated, which is then used to disambiguate the velocity with the fine range bin locations.Range Focusing Algorithm (RFA):

[0055] In the range phasor, RZZinb) is called the coarse range bin estimation and it is output by the correlation block, CIU, every pulse interval. Since the bandwidth used per pulse is only B, smaller than the total scan bandwidth, the range bin sizes after the CIU are larger than the desired size. Hence, the coarse range bins, index by b, are rough estimates of the true target delay, T.

[0056] Supposing that the step carriers are linearly stepped by B freguency spacing per pulse for the NB pulses, after which the step returns back to the first carrier. That is: fc,n = fc,o + (n % NB)BWhere n % NBis a modulo operation on n. Moreover, NB is also the number of uniform steps taken to aggregate the total scan bandwidth.

[0057] Then RFA operates an NB size Inverse fast Fourier transform (IFFT) for the NB pulses at a given coarse range index, b. The RFA outputs are written as:Sb(m) = IFFTWfi[ 7?ZZJl(b)] (4)In eguation (4), the pulse index k is understood to be modulo NB, such that the fast Fourier transform (FFT) outputs NB number of fine range bins, index m. Due to this FFT, for the NB pulses, RFA outputs NB additional fine range bins corresponding to the coarse range bin b.Doppler Focusing Algorithm (DFA):

[0058] Doppler Focusing Algorithm follows RFA with three processing approaches. One is to process the Doppler transforming using conventional FFT across fine range bins, m. Another is to process across coarse range bins, b, using a special transformation. A third option is to also use a conventional FFT, except the pulse duration of every step carrier is slightly adjusted to compensate for the changing of the freguency. Only one of these configurations should be executed per scan per the datapath selection in FIG. 6.

[0059] An exemplary first DFA may be simpler to describe. Using a conventional FFT, every Sb(m) is sampled at index m and fed into the FFT. Across the entire scan, thereare Np= (number-of-pulses I NB) which can be configured to yield an integer. To be more precise, the fine range bin outputs can be written as:Sb(m,p)The p index would now represent the multiple of NB pulses across the scan, where p = 0 ... NP / NB-1 . Therefore, the DFA of fine range bins would use conventional FFT of length Np, and executed multiple times across every coarse range bin b, and each fine range bin index m. More precisely, suppose DFA1 is the output of the DFA using fine range bins:DFAlb m(d) = FFTp[ Sb(m, p) ] (5)Where b index in set of coarse range bins, m = 0 ... NB-1 , p = 0 ... NP / NB-1 . The d is the Doppler bin index. It is noted that the pulse index as the FFT index is counting in increments of NB pulses. That is, if p’ is the index to pulses, p = NB p’. This is because the fine range processing distinguishes the NB pulses as NB fine range bins. FIG. 7 illustrates an example of how the fine range bins are indexed before input to the DFATs FFT.

[0060] The example in FIG. 7 shows DFA1 when m = 1. Due to RFA, the NB pulses would produce NB fine range bins given by Sb(m,p). One affect of this algorithm is that the maximum unambiguous Doppler has been reduced by a factor of NB because DFA1 increments by NB pulses.DFA2:

[0061] Unlike the first DFA, the exemplary second option process Doppler transforms across every pulse for each coarse range bin. At the output of the second DFA, call that DFA2,Where the Doppler index d = 0 ... Np-1 , coarse range bin b = {set of coarse range bins}.

[0062] In this second DFA process, the Doppler estimation for each coarse bin b is applicable for every fine range bin m within the given b. That is, for a given output of DFA2bo, that entire Doppler vector applies to Sbo(m), for m = 0 ... NB-1 at bo.DFA3:

[0063] For the third option of DFA, the output is an exemplary DFA3 and can be made such that:DFA3b(d) = FFTn[ Rzz(n, 6)] (7)This exemplary option is interesting in that it uses the same conventional FFT as the first exemplary option but operates on the CIU outputs directly, not the RFA outputs. To do that, the third option requires every PRI to be adjusted such that each PRI is adjusted slightly according to the carrier frequency ratio to the previous frequency. That is:Where To is the desired virtual PRI that’s never transmitted. Inverting To would translate to the conventional maximum Doppler specification. Similarly, fo is the base carrier that’s never transmitted.

[0064] FIG. 7 illustrates a timing diagram of the PRI using three (3) pulses and a stepped carrier on each pulse. The n is the pulse index where fc n+1and Tn+i are the corresponding carrier frequency and the pulse duration, respectively. As discussed herein, the algorithm of equation (8) is used to set the pulse duration for nthpulse.Angle Processing:

[0065] As discussed herein, the angle processing in the exemplary step carrier PMCW can be performed with conventional methods. The conventional angle processing algorithms can also be used in the step carrier PMCW.Target Activation:

[0066] An exemplary Target Activation processor, or also called “target detection,” for an exemplary step carrier PMCW can be similar to conventional Constant False Alarm Rate (CFAR) detection techniques. As discussed herein, the exemplary step carrier PMCW embodiment can utilize conventional techniques of target activation.Fine Range Disambiguation:

[0067] After RFA, a target at a particular fine range bin is coupled with its non-zero velocity. The coupling effect is usually referred to as range and Doppler ambiguity, typical of FFT-based algorithms. FIG. 8 illustrates examples of the fine range ambiguity coupled with several Dopplers, marked by velocities. In each example, the actual position was the target inside the fine range bin and the Doppler frequency was modulated on the target without range walk effects. Even without the range walk effects, it can be seen (in FIG. 8) that the target peaks in range can be confused with another peak due to Doppler. Therefore, it’s necessary to disambiguate range and Doppler parameters to estimate the true target.

[0068] Depending on the selected DFA option, each disambiguation procedure is a little different. For example, for the first algorithm, the DFA1 outputs are Doppler estimates from the fine range bin, via RFA. Due to RFA, the estimated fine range bin already compensated for the Doppler that’s intra-block of every NB pulses. Therefore, the DFA1 output is actually the Doppler rotation every NB pulses. A more precise way to express this is as follows:Let M = Quotient of — , let m = the activated fine range bin of the target. NBLet mO be the final disambiguated fine range bin output for the target in question.'fo2vT ! Af2vT m0= m —Where T is the pulse duration or PRI, c is speed of light, fO is the base carrier frequency, f is the step frequency size, and v is the estimated target velocity. For DFA1 , v is calculated as: c d V = -2Tf0(10)Where d is the Doppler bin index yielded by DFA1 . Then:

[0069] In most cases, the step frequency size is much smaller than the base carrier frequency, thus, equation (11) is usually a real number that must be rounded to the closest integer for the estimated range bin.

[0070] When using the exemplary DFA2 and DFA3, the Doppler estimates use coarse range bin inputs. They also output Doppler bin index d and use equation (11 ) to disambiguate the fine range bin (see FIG. 9). The important difference is that DFA3 has a virtual pulse duration, To, which replaces the T variable in equations (9) and (10).Non-linear steps:

[0071] As discussed previously in the theory of operation section, an exemplary system simplification was made where the step index k was set to equal the burst index n. When k = n, the steps occur in linear steps of time, but this characteristic doesn’t have to be a requirement. In an exemplary non-linear stepping scheme, the waveform can transmit bursts with any step index, even randomly. For proper step frequency demodulation, however, the transmit step pattern has to be known to the receiver.

[0072] With the known stepping patterns, the exemplary RFA and DFA procedures may be modified.RFA under non-linear stepping patterns:

[0073] With linear stepping patterns, the RFA uses an exemplary inverse fast Fourier transform (IFFT) to transform coarse range bins into fine range bins. That algorithm works under the constraint of the number of bursts, NB, is the same number of uniform frequency steps taken to aggregate burst bandwidth of B into total scan bandwidth, NB X B. As long as these constraints are followed for non-linear stepping, such as random stepping, the same RFA procedure with IFFT can also be used to get fine range bins with the exception of re-indexing the CIU outputs.

[0074] In one exemplary embodiment, to see how the RFA outputs are re-indexed, the RFA’s transform coefficients are expanded as follows:In equation (12), the Fourier transform kernel uses the same output index as IFFT except the CIU outputs are indexed by a burst function s(m,n). The function s(m,n) contains the known stepping pattern where if s(m,n) = m, m is the linear stepping pattern. Note the exemplary trivial pattern of s(0,n) = 0.

[0075] As illustrated below, in one exemplary embodiment, if the patterns in s(m,n) are exactly reversed from the linear time index, then, N-m = s(m,n), where if m = N, this is the same as m = 0 due to the wrap-around effect of the Fourier kernel. Note that m is also called the fine range bin index, but its easier to regard this index as the frequency index which coherently sums the phase rotations at exactly that index. The table (I) below explicitly lists this example for m = 1 , 2, and 7 in correspondence to n.DFA under non-linear stepping patterns.

[0076] Since there were three different DFA procedures given earlier, each one may be treated a bit differently when stepping patterns are non-linear. For the exemplary DFA1 , since its inputs are the fine range bins, as long as the RFA was processed as described using the modified RFA technique, the exemplary DFA1 proceeds in the same manner as given in the DFA1 section discussed above.

[0077] For the exemplary DFA2, equation (6) can also be used without modification. It is important to note that the Doppler transform kernel has an inherent abstraction of the stepping frequency by burst index n, as fcn. This means that the exemplary DFA2 inherently requires that the stepping pattern be known and sequenced by the burst.

[0078] For the exemplary DFA3, equation (7) can also be used without modification as long as equation (8) is satisfied. Regardless of step patterns, burst durations in the exemplary DFA3 are adjusted according to the ratio of step frequencies to the base carrier. As equation (8) illustrates, the step pattern is embedded in the step frequency / C,n-MIMO Multiplexing with Code and Doppler Division:

[0079] In phase modulated radars, digital codes are used to isolate different transmitters of the antenna array. For the most straight forward realization of digital coding the transmitters is to assign a unique code to each transmitter. By the term “unique code,”it means that between codes have low cross correlation such that conventional beamforming of antenna arrays provides sufficient peak to side lobe ratios (PSLR). This coding technique is called Code Division Multiplexing or CDM.

[0080] Besides CDM, another popular multiplexing technique, often employed in frequency modulated radars, is Doppler Division Multiplexing, or DDM. In the straightforward realization of DDM, each transmitter is assigned a unique Doppler frequency such that the antenna array elements can be separated using different Dopplers.

[0081] In one exemplary embodiment, CDM and DDM are combined in a hybrid technique for phase modulated radars, herein referred to as “CDM+DDM.” In this proposal, all transmit antenna elements are partitioned into code groups and then subpartitioned into Doppler groups.

[0082] For transmission of CDM+DDM, the system parameters denote N as the number of CDM channels, and M as the number of DDM channels. A model of transmitter output x(t), in a simple form to only accommodate CDM+DDM analysis, is as follows:where fcis the carrier frequency cn(t) is the unique code index n which exhibit low cross correlation between unique codes, for n = 0 ... N-1 fnmis the DDM frequency, n denotes the CDM channel, and m is the DDM frequency index for m = 0 ... M-1 . It is possible to simplify the doppler transmission component to sampled phases at the pulse rate like this: gj2Ttcf>nm(£Tp)rectwhere Tpis the pulse repetition interval (PRI), and I is the PRI counter. Thus, 4>nm(ZTp) is the sampled doppler phase rotation at PRI due to fnm. The p is the transmitter index for each of the transmit antenna array elements. The p can be either CDM index major or DDM index major. For default CDM major index, p = mN + n.

[0083] FIG. 10 illustrates a block diagram of the transmit system model with the CDM and DDM indices for each transmit antenna element.

[0084] When DDM is incorporated into FMCW, the typical approach is to divide up the unambiguous doppler frequency span into smaller segments with the DDM carrier index placed at the boundaries of those segments. Moreover, some of those segment allocations are assigned with null transmissions such that only a subset of segments are actually used for DDM. This design is illustrated in FIG. 11 . The num_ddm is the number of used DDMs for transmission. As illustrated, there may be multiple blank segments that have no transmission. The blank segments are present to assist the disambiguation of doppler ambiguities arising from DDM.

[0085] There is an underlining assumption in FIG. 11 that target peaks located at different DDM indices are orthogonal to others. That means that targets are not spanning multiple indices due to targets not smearing across Doppler and that Doppler resolution, determined by the number of pulses in a scan, is high enough. With that assumption, the DDM index is calculated as fnmTp.

[0086] While blank segments in DDM design embodiments help disambiguation, CDM+DDM, however, can deal with Doppler ambiguities using an exemplary design approach utilizing non-uniform DDM segments. Specifically, the exemplary approach is an analytical procedure on a code and Doppler dimensionality grid called the Hybrid Multiplex Map, HMM. Exemplary HHM embodiments will be discussed herein.Hybrid Multiplex Map:

[0087] The hybrid multiplex map, HMM, is a graphical representation of the exemplary design specifications given to fnm. An example realization of HMM is illustrated in FIG. 12. As illustrated in FIG. 12, CDM channels are illustrated as vertically stacked and are the unique code division channels that create CDM. In this example, four CDM channels are created using four unique codes. It is assumed that the cross correlations between unique codes are sufficiently low to create clear separation between the CDM channels.

[0088] The DDM channels are arranged horizontally in FIG. 12, with each small rectangle representing the Doppler index location of the DDM assignment for each CDM channel. From this illustration (FIG. 12), it is clear that each rectangle has fnmof different values at every CDM and DDM channel.

[0089] By design, the transmitters can choose the fnmwith the system parameters of PRI, and number of PRI, to form a whole number for CDM+DDM index as follows: CDM+DDM index = fnmTpNpri(15)

[0090] On the HMM grid, the CDM+DDM index is called h(n,m) and the distance between consecutive doppler indices is measured byA / i(n, m) = h(n, m) — h(n, m — 1) (16)

[0091] By definition of the index, h(n,m), is only valid for non-negative n or m. For any n or m < 0, h = 0.

[0092] Given an exemplary design of the indices h(n,m), the collection of a specification is called an HMM object. This exemplary HMM object can be shifted in the DDM dimension using the notation HMM(k), where k is the amount of shift in number of DDM index bins.

[0093] An example of a shifted HMM is illustrated in FIG. 13. The white rectangles are the shifted versions of the original unshifted HMM of shaded rectangles. Every shifted rectangle is shifted by the same amount indicated by the number k in HMM(k). HMM(0) would indicate the unshifted map. Similar to aliasing effects, as shifts increase indefinitely, the DDM index that exceed the number of unambiguous Doppler bins, usually this is also equal to Npri, would wrap-around back to 0. This wrap-around effect is the same as taking the modulo operation on the HMM shifts.

[0094] The HMM(k) is a useful map to analyze situations when the shifted indices that overlap with the unshifted versions. From the previous example, it can be seen that at some shift amount, the top row of CDM channel would eventually cause the first DDM index, h (0, 0), to overlap with h(0, 1 ). That occurs at k = A / i(0, 1) . When an overlap occurs, this is counted as one hit.

[0095] In one embodiment, an exemplary analysis of how well an HMM is designed, is to count the hits than any HMM(k) has on HMM(O). Essentially, the count of hits of HMM(k) for any k, is the basic influence on the magnitude of the beamform images resulting from the Doppler ambiguities.Beamforminq in CDM + PPM:

[0096] To understand how Doppler ambiguities in HMM(k) impacts beamforming, the virtual receiver processing of CDM+DDM is examined. FIGS. 14A and 14B overview the exemplary major processing steps needed at the receiver for a transmitter that uses CDM+DDM.

[0097] Although most of the exemplary processing steps are similar, there are a few differences. First is the number of reference codes used by the range correlator unit called the CIU. In contrast to pure CDM systems, the number of reference codes is less than the number of transmitters in the system. As a result of re-using codes across some transmitters, the first data object called RDC1 does not have all of the virtual receiver, VRX, elements needed for beamforming.

[0098] These VRX elements are fully populated after a new exemplary processing block referred to as the Doppler to VRX ( or “Dop. to VRX”) mapper. This block follows the Doppler transformation block, called the FEU. What the Dop. to VRX mapper does is to extract the VRX elements that are encoded into the DDM index location.

[0099] Due to this extraction of VRX elements from Doppler, it is possible that another Doppler bin may contain energy from another target return of the same range bin, may be the same angle of arrival, and different velocity. These other possible Doppler energies are called ambiguities in the data object, RDC2. After beamforming by the RAU block, the final data set, RDC3, would contain angular images.

[0100] Another definition is needed to analyze the location and magnitude of the angular images, the number of hits per receive channel. The number of hits is a function of HMM and the shift number k. For any HMM design, it counts every location overlap of HMM(k) to HMM(0).num. hits (HMM, k) = HMM(k) hits on HMM(0) (17)

[0101] Since the angular images are direct consequence of hits, it is clear that the optimization problem of preventing or minimizing images in CDM+DDM design is: argmin(num of hits(HMM, k) ) (18)HMM

[0102] One example of achieving this optimization is to define the minimum hit of < 1 per receive channel, for all k. To do that, produce an exemplary HMM with the following specifications:1) for m = 0, h(n, 0) = 0 (19)2) for m > 0, A / i(n, m) = mN + n (20)

[0103] As long as the number of available Doppler index > 2 max(A / i(n, m)), which implies that number of pulses in the scan meets this constraint, the maximum of number of hits is not greater than 1 , per receive channel, for any k. This example design of an HMM embodiment is illustrated in FIG. 15. The number of hits for all k is illustrated in FIG. 16.

[0104] Even with a well-designed HMM embodiment, as per the example, the singular hits will nonetheless create interference patterns to beamforming. To see how interference patterns arise from beamforming, define r as the vector of VRX, also known as a skewer, and seas the steering vector at angle of arrival 0. Then the typical beamform output would be: se*r (21)

[0105] Let r0be the desired VRX vector without doppler ambiguity, then all of the ambiguous VRX components are additional terms in the final beamform, which is:Equation 22

[0106] On the right-hand side of the above equation (equation 22) are the interferences, if non-zero, of the Doppler ambiguities that align with the VRX location of r0, i.e. hits. Notice that index k of the summation starts with 1 , which is the first DDM ambiguity of agiven CDM channel n. dkis the Doppler to VRX mapper function that maps the VRX index in the DDM dimension to the Doppler index designed by HMM, or any other DDM approach.

[0107] Under the design example of the exemplary HMM embodiment, most of the r terms on the right side are zeros, ignoring noise. Given the hit 1 HMM example, every receiver channel m has at most one non-zero r on the right side. That non-zero r isn’t the same VRX component of r0, it’s actually one of the VRX elements from another target. For this reason, the right-hand side summation is called inter-target interference (ITI).

[0108] A numerical example on ITI analysis can be done using a uniform linear array (ULA) for an antenna array with 12 transmitter and 8 receiver elements. Of the 12 transmitters, 4 CDM channels and 3 DDM channels can yield a simple HMM embodiment with hits of no greater than 1 . Through this exemplary system, suppose two targets are in the radar returns with high SNRs of 60 and 40 dB, respectively.

[0109] Both of these targets can be assumed to be at the same range and same AoA, 10 degrees, to combine into some receive skewer r. Then Equation 22 can be used to analyze ITI effects of the stronger target onto the weaker target. The analyze result is illustrated in FIG. 17.

[0110] The solid line is plot of the left-hand side of Equation 22 for just the weak target and displays a clear target peak with expected angular sidelobes of an ULA. After adding the strong target with only 1 hit ambiguity, which can occur at some velocity of a strong target where the exact Doppler ambiguity is not material to Equation 22 analysis, the resulting beamform result from the right-hand side and the left-hand side produce the dashed line response. Although the peak of the weak target at 10 degrees is still the highest peak, there are many high sidelobes across angles. The occasional sidelobe peaks in angle is the result of the interference pattern generated by the righthand side of Equation 22 depending on the sebeing applied.

[0111] From this simple example, there exist algorithms that could minimize effects of the ITI of Equation 22 while preserving as much power as possible for the left-hand side of Equation 22. One such algorithm is the zero-forcing algorithm.Zero-Forcing Algorithm for ITI:

[0112] In the previous section, a simple example of two targets where one target is much stronger than the other, the stronger target can produce noticeable ITI that can interfere with detection of the weaker target. Due to the design of the exemplary HMM embodiment, the zero-forcing algorithm works well with reducing effects of ITI without degrading the desired beamforming energy of the weaker target.

[0113] The exemplary algorithm of zero-forcing is simple, find the hit location of ITI on the right-hand side of Equation 22 and replace those r elements with 0. This is possible only after the detection of the strong target first. With the a priori knowledge of the exemplary HMM embodiment, the Doppler ambiguity of the strong target at dkreveals the location of ITI with respect to the weak target.

[0114] Continuing with the previous numerical example, suppose that the ITI of the strong target is removed by replacing the ITI locations with 0 values, the beamform outcome of the weak target is almost recovered to the original without ITI, see FIG. 18. After zero-forcing, the main peak of the weak target is reduced by less than 1 dB while the angular sidelobes have been reduced by almost 20 dB.

[0115] Even after zero-forcing, however, the sidelobes are still much higher than without ITI. In fact, the residue of the occasional peaks is still present in the angular response. These residual spikes in the angular dimension are the direct result of replacing values with zeros, creating discontinuities in the input skewer samples.

[0116] Where zero-forcing is limited in application there are scenarios with too many strong targets at the same range and within the Doppler ambiguities of the weak target. The more strong targets there are, the more elements in the skewer that have to be erased, which degrades the main power of the weak target and elevates the angular sidelobes from those extra discontinuities.

[0117] To summarize, zero-forcing is a simple technique that can remove almost all of ITI in simple scenarios like one very strong target at the Doppler ambiguity of a weaker target. An example implementation of zero-forcing is illustrated in FIGS. 19A and 19B. With a priori knowledge of an exemplary HMM embodiment, ambiguity locations can be zeroed-out after activation of strong targets. Once the zero-force erasure of the ambiguities from strong targets execute another step of beamform and activation, the weak targets should be able to be detected if SNR is sufficient. The looping of zeroforce erasure should stop after the second round of activations because further zeroforcing of the weaker targets would create too many zero-values in the skewer to make the subsequent beamforming valueless.

[0118] Thus, the exemplary embodiments discussed herein provide a radar system with phase modulated continuous wave signal transmission and reception a variety of improvements. For example, the phase modulated continuation wave signal can include a carrier signal that is a stepped carrier. The stepped carrier phase modulated continuous wave comprises a plurality of cascading bursts for each scan, with each burst having a relatively small bandwidth as compared to the bandwidth of the scan. That is, by cascading a plurality of bursts together over the course of a scan, the step carrier can achieve the same bandwidth as a conventional wide bandwidth PMCW signal. In another embodiment, an exemplary phase modulated radar utilizes a plurality of digital codes: a combination of code division multiplexing (CDM) and Doppler division multiplexing (DDM). The phase modulated radar comprises a plurality of antenna elements partitioned into code groups and then sub-partitioned into Doppler groups. For transmission of phase modulated signal that is coded using CDM+DDM, the radar system models the transmission output noting the number of CDM channels and the number of DDM channels. In one embodiment, the number of DDM segments include one or more blank segments where no transmission takes place, the blank segments present in the scan to assist the disambiguation of Doppler ambiguities in the transmission and reception arising from DDM.

[0119] Changes and modifications in the specifically described embodiments can be carried out without departing from the principles of the present invention which isintended to be limited only by the scope of the appended claims, as interpreted according to the principles of patent law including the doctrine of equivalents.

Claims

CLAIMS:1 . A radar system comprising: a transmitter pipeline comprising a plurality of transmitters, each configured to transmit phase modulated radio signals; and a receiver pipeline comprising a plurality of receivers, each configured to receive phase modulated radio signals that include signals transmitted by the plurality of transmitters and reflected from objects in an environment; wherein each transmitter is configured to transmit the phase modulated radio signals with a stepped carrier frequency comprising a plurality of different carrier frequencies, wherein each transmitter is configured to transmit the phase modulated radio signals as a plurality of consecutive transmission bursts, each burst transmitted with a different carrier frequency; and wherein each receiver is configured to receive the phase modulated radio signals during a radar scan, wherein a respective bandwidth of each individual transmission burst is smaller than a bandwidth of the radar scan, and wherein the bandwidth of the radar scan comprises a quantity of received transmission bursts each with a respective burst bandwidth.

2. The radar system of claim 1 , wherein the radar scan comprises a scan duration, and wherein each transmission burst of the quantity of received transmission bursts comprises a burst duration, and wherein a sum of the burst durations of the quantity of received transmission bursts equals the scan duration.

3. The radar system of claim 2, wherein each of the plurality of transmitters is configured to transmit the plurality of consecutive transmission bursts such that the respective carrier frequencies of the plurality of transmission bursts are incrementally stepped in frequency.

4. The radar system of claim 3, wherein the carrier frequencies of adjacent transmission bursts in time at least partially overlap.

5. The radar system of claim 2, wherein each of the plurality of transmitters are configured to transmit each consecutive transmission burst such that adjacent transmission bursts in time do not overlap in their respective burst duration.

6. The radar system of claim 2, wherein each of the plurality of transmitters are configured to transmit each consecutive transmission burst such that two separate transmission bursts have different burst durations.

7. The radar system of claim 2, wherein each of the plurality of receivers is configured to process the received phase modulated radio signals to determine a range to a detected object according to a corresponding coarse range bin for each individual transmission burst, wherein the plurality of receivers are configured to use the resulting coarse range bins for a plurality of corresponding transmission bursts to determine fine range bins for the scan duration, and wherein each coarse range bin comprises a plurality of fine range bins.

8. The radar system of claim 7, wherein each of the plurality of receivers is configured to perform Doppler processing on the fine range bins of the scan duration, wherein the Doppler processing comprises a fast Fourier transform (FFT) performed across the fine range bins of the scan duration.

9. The radar system of claim 7, wherein each of the plurality of receivers is configured to perform Doppler processing on the coarse range bins using a Doppler transform across every pulse for each coarse range bin.

10. The radar system of claim 7, wherein each of the plurality of receivers is configured to perform Doppler processing on the received phase modulated radio signals using a fast Fourier transform (FFT) while the burst duration of every step carrier is adjusted to compensate for the changing of the frequency.

11. A method for processing radio signals for a radar system comprising a transmitter pipeline comprising a plurality of transmitters and a receiver pipeline comprising a plurality of receivers, the method comprising:transmitting with the plurality of transmitters phase modulated radio signals; receiving with the plurality of receivers phase modulated radio signals that include signals transmitted by the plurality of transmitters and reflected from objects in an environment; wherein the transmitted phase modulated radio signals are transmitted with a stepped carrier frequency comprising a plurality of different carrier frequencies, wherein the phase modulated radio signals are transmitted as a plurality of consecutive transmission bursts, each burst transmitted with a different carrier frequency; and wherein received phase modulated radio signals are received during a radar scan, wherein a respective bandwidth of each individual transmission burst is smaller than a bandwidth of the radar scan, and wherein the bandwidth of the radar scan comprises a quantity of received transmission bursts each with a respective burst bandwidth.

12. The method of claim 11 , wherein the radar scan comprises a scan duration, and wherein each transmission burst of the quantity of received transmission bursts comprises a burst duration, and wherein a sum of the burst durations of the quantity of received transmission bursts equals the scan duration.

13. The method of claim 12, wherein the plurality of transmission bursts are transmitted such that the respective carrier frequencies of the plurality of consecutive transmission bursts are incrementally stepped in frequency.

14. The method of claim 13, wherein the carrier frequencies of adjacent transmission bursts in time at least partially overlap.

15. The method of claim 12, wherein each transmission burst is transmitted such that adjacent transmission bursts in time do not overlap in their respective burst duration.

16. The method of claim 12, wherein each transmission burst is transmitted such that two separate transmission bursts have different burst durations.

17. The method of claim 12 further comprising processing with the plurality of receivers the received phase modulated radio signals to determine a range to a detected object according to a corresponding coarse range bin for each respective transmission burst, wherein the resulting coarse range bins are used for a plurality of respective transmission bursts to determine fine range bins for the scan duration, and wherein each coarse range bin comprises a plurality of fine range bins.

18. The method of claim 17 further comprising performing Doppler processing on the fine range bins of the scan duration, wherein the Doppler processing comprises performing a fast Fourier transform (FFT) across the fine range bins of the scan duration.

19. The method of claim 17 further comprising performing Doppler processing on the coarse range bins using a Doppler transform across every pulse for each coarse range bin.

20. The method of claim 17 further comprising performing Doppler processing on the received phase modulated radio signals using a fast Fourier transform (FFT) while the burst duration of every step carrier is adjusted to compensate for the changing of the frequency.

Images