DETAILED ACTION
Notice of Pre-AIA or AIA Status
The present application, filed on or after March 16, 2013, is being examined under the first inventor to file provisions of the AIA .
Priority
Acknowledgment is made of applicant’s claim for foreign priority under 35 U.S.C. 119 (a)-(d). The certified copy has been filed in parent Application No. EP22196745, filed on 09/21/2022.
Information Disclosure Statement
The information disclosure statements (IDS) submitted on 08/08/2023, 07/12/2024 and 10/14/2024 are in compliance with the provisions of 35 CFR 1.97. Accordingly, the IDS have been considered by the examiner.
Claim Objections
Claim 8 objected to because of the following informalities: ". Appropriate correction is required.
Claim Rejections - 35 USC § 103
The following is a quotation of 35 U.S.C. 103 which forms the basis for all obviousness rejections set forth in this Office action:
A patent for a claimed invention may not be obtained, notwithstanding that the claimed invention is not identically disclosed as set forth in section 102, if the differences between the claimed invention and the prior art are such that the claimed invention as a whole would have been obvious before the effective filing date of the claimed invention to a person having ordinary skill in the art to which the claimed invention pertains. Patentability shall not be negated by the manner in which the invention was made.
The factual inquiries for establishing a background for determining obviousness under 35 U.S.C. 103 are summarized as follows:
1. Determining the scope and contents of the prior art.
2. Ascertaining the differences between the prior art and the claims at issue.
3. Resolving the level of ordinary skill in the pertinent art.
4. Considering objective evidence present in the application indicating obviousness or nonobviousness.
Claims 1-7, 9, and 11-14 are rejected under 35 U.S.C. 103 as being unpatentable over Park et al. (US 2021/0333386 A1) in view of Bechter et al. (Compensation of Motion-Induced Phase Errors in TDM MIMO Radars).
Regarding Claims 1, 12, 13, and 14, Park et al. (‘386) in view of Bechter et al. teaches:
Park et al. (‘386) teaches: A method for resolving a phase ambiguity between subarrays in a virtual array of a time-division multiplexing (TDM), multiple-input multiple-output (MIMO), frequency modulated continuous-wave (FMCW), radar ([0006-0007]: “This demand is met by Multi Input Multi Output (MIMO) radar devices and methods… The MIMO radar includes transmitter circuitry which includes a plurality of transmit channels. The transmitter circuitry is configured to transmit, via a first subset of the transmit channels and during a first time interval, concurrent first frequency-modulated continuous-wave (FMCW) radar signals… transmit, via a second subset of the transmit channels and during a second time interval subsequent to the first time interval, concurrent second FMCW radar signals”).
Park et al. (‘386) teaches: wherein the TDM MIMO FMCW radar comprises an array of physical receivers including at least one row of physical receivers with a first spacing in a first direction, and further comprises a plurality of physical transmitters arranged with a second spacing in said first direction ([0004]: “MIMO (Multi Input Multi Output) is widely used to enlarge effective radar aperture size by synthesizing a virtual receiver array by combination of physically implemented multiple transmitter channels and multiple receiver channels”; [0030]: “FIG. 4 illustrates a concept of virtual array synthesis”).
Park et al. (‘386) teaches: wherein each of the subarrays in the virtual array is generated by a combination of the array of physical receivers and one of the physical transmitters ([0004]: “MIMO (Multi Input Multi Output) is widely used to enlarge effective radar aperture size by synthesizing a virtual receiver array by combination of physically implemented multiple transmitter channels and multiple receiver channels”; [0086-0087]: “After the CDM synthesis 916-1 for separating different Tx channels within a CDM subset has been performed, phase information from different CDM subsets may be merged via TDM synthesis 916-2 in order to obtain full virtual array synthesis”).
Park et al. (‘386) teaches: the method comprising: obtaining a virtual array signal of a range-Doppler bin relating to a scene with a moving object, each element of the virtual array signal corresponding to one virtual antenna element of the virtual array ([0019]: “perform a first range FFT of first receive signals… perform a first Doppler FFT of the first range signal to generate a first Doppler signal… separate range-Doppler bins associated with different transmit channels”; [0003]: “A first FFT, also commonly referred to as range FFT, yields range information. A second FFT across the range transformed samples, also commonly referred to as Doppler FFT, yields speed information. The first and second FFTs yield a so-called 2D range-Doppler map comprising range and speed (FFT) bins”).
Park et al. (‘386) teaches: compensating a velocity-induced phase shift of the virtual array signal using a phase compensation method, which introduces a phase ambiguity between the subarrays if the moving object’s velocity exceeds a threshold, thereby obtaining a compensated virtual array signal ([0087-0088]: “Between two successive TDM time slots for CDM subsets TA and TB, TDM introduces phase difference ΦG due to the time gap G and the velocity v of target… According to embodiments of the present disclosure, ΦG may be compensated to synthesize aperture. The velocity v can be measured from the outputs of the respective Doppler FFTs, but it may be ambiguous when targets’ Doppler frequency becomes larger than a maximum Doppler sampling rate or pulse repetition frequency (PRF). Therefore, compensating phase may also have an ambiguity”; [0091]: “The Doppler FFT outputs for CDM subsets TX A and TX B are separated by ΦG=(4π/λ)v due to the time gap G and the velocity v of the target”).
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches: for each of a plurality of the subarrays, computing a frequency spectrum of those elements of the compensated virtual array signal which correspond to consecutive virtual antenna elements generated by physical receivers belonging to the same row (Bechter [equation 8, pg. 1165]: equation 8; “Because of the linearity of the DFT, (8) also holds in multi-target scenarios”; [pg. 1164]: “When a TDM MIMO scheme is applied, the active transmit element is changed after each transmission of a single chirp… The geometric location of each transmitter and each receiver forms a virtual array which has a maximum number of M·N virtual elements. A range-Doppler spectrum is found for each element of the virtual array”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the TDM-CDM MIMO radar system of Park et al. (‘386) with the Doppler frequency spectrum processing method of Bechter et al. One would have been motivated to do so because both references address the identical technical problem of motion-induced phase errors in TDM MIMO radar systems. Park explicitly recognizes that velocity-induced phase differences create ambiguity between TDM time slots ([0087-0089]) but does not disclose the specific signal processing technique to compensate for this error. Bechter provides the precise solution to this problem by teaching an adapted DFT that processes each transmitter’s chirps separately with appropriate phase compensation factors (equation 8). The motivation to combine would be to achieve accurate angle-of-arrival estimation in TDM MIMO systems when targets have high velocities, which both references identify as a critical challenge. One would have had a reasonable expectation of success in making this combination because: (1) Bechter demonstrates through both simulation and measurement that the adapted DFT processing (equation 8) successfully eliminates motion-induced phase errors in TDM MIMO systems (Bechter, Fig. 4 and Fig. 6, pg. 1166, showing identical DoA estimation for moving targets at 18 m/s and static targets); (2) both references use the same TDM MIMO architecture with multiple transmitters operating sequentially and generating virtual array elements; (3) Bechter’s method requires no additional hardware and involves only a straightforward modification to the DFT processing that Park already employs ([0003], [0012]); and (4) the physical principles are identical—both references recognize that the phase error is proportional to the Doppler frequency and the time gap between transmitters (Park [0087], Bechter equation 3).
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches: identifying, jointly for the frequency spectra of said plurality of the subarrays, an amplitude-peak frequency (Bechter [pg. 1165-1166]: “The noise-free simulation considers a target in distance 30 m, ϑ=15°, and the velocities v=0 and v=18 m/s. Fig. 4 shows the DoA estimation on the range-Doppler cell that contains the target”; [pg. 1166]: “In the range-Doppler spectrum, a dominant target peak is selected for DoA estimation”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the phase ambiguity resolution system of Park et al. (‘386) with the amplitude-peak frequency identification method of Bechter et al. One would have been motivated to do so because identifying the correct frequency bin containing the target is a prerequisite for accurate angle-of-arrival estimation after phase compensation. Park teaches performing DoA processing on range-Doppler bins ([0094]) but does not explicitly describe how to identify which frequency bin corresponds to the actual target when multiple subarrays are analyzed. Bechter solves this problem by teaching that the dominant target peak in the range-Doppler spectrum is selected across all virtual array elements (pg. 1166), which corresponds to jointly identifying the amplitude-peak frequency across the subarrays. The motivation to combine would be to ensure that the phase comparison and compensation operations are performed at the correct Doppler frequency, thereby achieving accurate target localization. One would have had a reasonable expectation of success because: (1) Bechter demonstrates successful target peak identification and subsequent accurate DoA estimation (Fig. 4, pg. 1166); (2) both references process range-Doppler data in the same manner using FFT operations; (3) peak detection in frequency spectra is a standard and well-understood signal processing technique; and (4) Bechter shows that selecting the range-Doppler cell containing the target leads to accurate angle estimation even for moving targets (pg. 1166, showing 15.2° estimation matching the actual 15° angle).
Park et al. (‘386) teaches phase comparison between subarrays but Bechter et al. more explicitly teaches: determining a residual phase shift between a pair of the subarrays within said plurality of subarrays by comparing, at the amplitude-peak frequency, the respective phases of the frequency spectra (Bechter [pg. 1166, Fig. 5]: “Measured phases at the elements of the MIMO virtual array. The element at position 10 occurs twice in the virtual aperture”; “At the position of the overlapping element there is a phase discontinuity of 1.54 rad for the conventional processing (6), due to the motion-induced phase error”; [pg. 1164, equation 3]).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine Park et al. (‘386) with the phase shift determination method of Bechter et al. One would have been motivated to do so because Park explicitly recognizes that phase differences exist between TDM time slots due to target velocity ([0087]: “TDM introduces phase difference ΦG due to the time gap G and the velocity v of target”) and states that this phase must be compensated ([0088]: “ΦG may be compensated to synthesize aperture”), but Park does not disclose how to determine the actual residual phase error that remains after initial compensation attempts. Bechter provides the solution by teaching that the phase discontinuity between subarrays (representing different transmitters) can be measured directly by comparing the phases at the overlapping or consecutive virtual array elements (Fig. 5, pg. 1166). The motivation to combine would be to accurately measure the residual phase error so that proper inverse compensation can be applied, thereby achieving full virtual array synthesis without phase distortion. One would have had a reasonable expectation of success because: (1) Bechter demonstrates that measuring the phase discontinuity at 1.54 rad successfully identifies the motion-induced error (Fig. 5, pg. 1166); (2) both references use the same mathematical framework where phase errors are additive and can be compensated by subtracting or applying inverse rotations; (3) Bechter’s Fig. 5 shows clear phase measurements across virtual array elements, demonstrating feasibility; and (4) the combination merely applies Park’s recognized need for phase measurement with Bechter’s disclosed measurement technique, both operating on the same type of TDM MIMO radar data.
Park et al. (‘386) in view of Bechter et al. teaches: applying an inverse of the residual phase shift to the compensated virtual array signal (Park [0093]: “For example, if NrnM=2, phase information of the detected peaks of the second virtual array may be rotated by 0° and 180° when combining it with the phase information of the first virtual array. The amplitude of angle spectrum at target angle is higher when it is compensated correctly”;
Park et al. (‘386) does not explicitly teach applying phase compensation, but Bechter teaches: (Bechter [pg. 1165, equation 7] - which represents applying phase compensation to compensate for the time offset; [pg. 1165, equation 8] - shows the generalized compensation for transmitter Txm; [pg. 1166]: “The phase-corrected range-Doppler processing (8) does not show such a severe deviation. Instead, the phase values at the position of the overlapping elements are nearly identical”; [page 1166, Fig. 5]: shows that after applying the adapted DFT with phase correction, the phase discontinuity is eliminated).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine Park et al.’s teaching of phase compensation through rotation with Bechter et al.’s specific method of applying inverse phase shifts. One would have been motivated to do so because Park explicitly recognizes the need for phase compensation between TDM subarrays ([0087-0088], but does not disclose the precise mathematical implementation of how to apply this compensation during the signal processing stage. Bechter provides the solution by teaching that the inverse phase shift is applied during the Doppler FFT processing itself, through the modified exponential term in equation 8, which applies the exact negative (inverse) of the motion-induced phase error. Park teaches that phase rotation achieves compensation, while Bechter teaches the specific implementation of this rotation through the adapted DFT formulation. The motivation would be to implement Park’s recognized phase compensation approach using an efficient method that corrects the phase error at the signal processing stage without requiring post-processing corrections or additional hardware.
Regarding Claim 2, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches: wherein identifying the amplitude-peak frequency includes determining a frequency of a main amplitude peak in a sum of the two frequency spectra’s respective amplitude parts (Bechter [pg. 1166]: “In the range-Doppler spectrum, a dominant target peak is selected for DoA estimation”; “Fig. 4 shows the DoA estimation on the range-Doppler cell that contains the target”; [pg. 1165-1166]: “The noise-free simulation considers a target in distance 30 m, ϑ=15°, and the velocities v=0 and v=18 m/s. Fig. 4 shows the DoA estimation on the range-Doppler cell that contains the target”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to determine the amplitude-peak frequency by summing the amplitude parts of the frequency spectra from different subarrays. One would have been motivated to do so because summing the spectra improves signal-to-noise ratio and makes the target peak more prominent, which is particularly important when the target signal may be weak or when there is noise in individual subarrays. This technique would maximize detection reliability by coherently combining the energy from multiple subarrays while allowing incoherent noise to partially cancel. One would have had a reasonable expectation of success because: (1) Bechter demonstrates successful dominant peak identification across the virtual array (pg. 1166); (2) summing amplitude spectra is a well-known signal processing technique for improving detection performance; (3) the frequency of the target peak is identical across all subarrays (they all see the same Doppler shift from the same target), so the peaks will align when summed; and (4) Bechter’s results show clear, identifiable peaks in the DoA spectrum (Fig. 4, pg. 1166), demonstrating that the target energy is sufficiently concentrated to be detected.
Regarding Claim 3, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Park et al. and Bechter et al. in combination teach: wherein identifying the amplitude-peak frequency includes determining a frequency which corresponds to a main or non-main amplitude peak in each of the respective frequency spectra’s amplitude parts (Bechter [pg. 1166]: “In the range-Doppler spectrum, a dominant target peak is selected for DoA estimation”; Park [0085]: “The peak signals whose values are above a threshold level, called CFAR threshold, may be taken as the possible target candidates”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to identify not only main peaks but also non-main amplitude peaks in the frequency spectra. One would have been motivated to do so because in multi-target scenarios or in the presence of clutter, the highest amplitude peak in one subarray may not correspond to the same target as the highest peak in another subarray. Additionally, there may be multiple targets of interest at different velocities, each creating peaks at different frequencies. Park explicitly teaches using CFAR thresholding to identify multiple peak candidates ([0085]), not just the single highest peak, demonstrating that identifying non-main peaks is necessary for proper target detection. The motivation to combine Park’s multi-peak detection with Bechter’s frequency spectrum processing would be to handle realistic scenarios with multiple targets or with targets that may not produce the highest amplitude return in every subarray. One would have had a reasonable expectation of success because: (1) CFAR processing is a standard and proven technique for identifying multiple peaks above a noise threshold; (2) Park explicitly teaches that multiple peak candidates are identified ([0085]); (3) Bechter’s adapted DFT method (equation 8) works for multi-target scenarios, as explicitly stated (pg. 1165: “Because of the linearity of the DFT, (8) also holds in multi-target scenarios”); and (4) identifying multiple peaks simply involves applying the same peak detection algorithm repeatedly to find all local maxima above the threshold, which is well-established in the art.
Regarding Claim 4, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches: wherein determining the residual phase shift includes computing a difference between the respective phases of the frequency spectra at the amplitude-peak frequency (Bechter [pg. 1166]: “At the position of the overlapping element there is a phase discontinuity of 1.54 rad for the conventional processing (6), due to the motion-induced phase error”; [equation 3, pg. 1164]: “For M transmitters the phase relation at the mth transmitter Txm is Δ˜ϕTxm = kd sin ϑ + 2πfDTr (m−1)/M” which inherently teaches computing phase differences between transmitters).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to compute the residual phase shift as a difference between the respective phases. One would have been motivated to do so because phase errors are inherently differential—what matters for array synthesis is the relative phase between subarrays, not the absolute phase of any individual subarray. Bechter demonstrates this principle by measuring the phase discontinuity (i.e., difference) of 1.54 rad at the overlapping element between subarrays (pg. 1166), and equation 3 shows that the phase error depends on the transmitter index (m), implying that different transmitters have different phases that must be compared. The motivation would be to quantify the exact phase mismatch between subarrays so that the appropriate inverse compensation can be applied. One would have had a reasonable expectation of success because: (1) Bechter successfully measures phase differences between transmitters and demonstrates that this measurement identifies the motion-induced error (Fig. 5, pg. 1166); (2) phase difference computation is a fundamental operation in complex signal processing—simply subtracting the argument (angle) of two complex numbers; (3) both references work in the complex domain where phase is readily accessible from the FFT outputs; and (4) Bechter shows that after identifying and compensating the phase difference, the phase values at overlapping elements become “nearly identical” (pg. 1166), proving that the difference computation and compensation works correctly.
Regarding Claim 5, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 4.
Bechter et al. teaches: wherein: the ratio of the first and second spacings is such that the virtual antenna elements of the virtual array are equidistant in the first direction (Bechter [pg. 1165-1166]: “The simulated radar uses the parameters in Tab. I and a TDM MIMO array with 2 transmitters in a distance of 5λ and 10 receivers with λ/2-spacing. It forms a 20 elements uniform linear virtual array with element spacing λ/2”; “The virtual array is a uniform linear array with element spacing 0.545λ”).
Park et al. (‘386) in teaches: determining the residual phase shift further includes rounding the difference between the respective phases of the frequency spectra to a multiple of 2π/M, where M is the number of physical transmitters (Park [0089]: “For example, if P=2G, the phase compensation offset is either 0° or 180°, if P=3G, the phase compensation offset is 0°, 120°, -120°, etc.” which teaches discrete phase compensation values that are multiples of 360°/M (i.e., 2π/M radians); [0093]: “When combining the NTDM virtual arrays, the number of phase compensation candidates are NTDM. For example, if NTDM=2, the phase compensation offset is either 0° or 180°, if NTDM=3, the phase compensation offset is 0°, 120°, -120°, etc.”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to round the computed phase difference to the nearest multiple of 2π/M. One would have been motivated to do so because, as Park explicitly teaches, the phase ambiguity in TDM systems with M transmitters results in discrete possible compensation values separated by 2π/M intervals (360°/M degrees) ([0089], [0093]). Park clearly states that for M=2 transmitters, the compensation is either 0° or 180° (i.e., 0 or π radians), and for M=3, it is 0°, 120°, or -120° (i.e., 0, 2π/3, or -2π/3 radians). This quantized nature arises from the fundamental physics of the TDM operation where each transmitter is offset by a time delay of one M-th of the total TDM period. When combined with Bechter’s teaching of computing the continuous phase difference (claim 4), rounding to the nearest discrete value would resolve the ambiguity by selecting the physically correct compensation value from the limited set of possibilities. The motivation would be to correctly handle velocity ambiguity cases where the measured Doppler frequency exceeds the unambiguous range, which both references identify as problematic (Park [0088], Bechter pg. 1). One would have had a reasonable expectation of success because: (1) Park explicitly teaches the discrete set of compensation values and demonstrates that selecting the correct one yields the highest amplitude in the angle spectrum ([0093]); (2) rounding a continuous value to the nearest discrete option is a straightforward mathematical operation; (3) the quantization interval (2π/M) is known from the system parameters (number of transmitters M), making the rounding operation deterministic; and (4) Park demonstrates that this approach successfully resolves phase ambiguity ([0093-0094], describing angle spectrum amplitude being higher when compensated correctly).
Regarding Claim 6, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches: wherein: the respective phases of the frequency spectra are compared, at the amplitude-peak frequency, for a plurality of pairs of subarrays which are uniformly spaced in the first direction (Bechter [pg. 1165-1166]: “It forms a 20 elements uniform linear virtual array with element spacing λ/2”; [pg. 1166, Fig. 5]: “Measured phases at the elements of the MIMO virtual array” showing phase comparison across the uniformly-spaced array).
Park et al. (‘386) does not explicitly teach, but it would have been obvious to compute: the residual phase shift is determined as a mean over all said pairs of the subarrays. One would have been motivated to do so because averaging phase measurements across multiple pairs of subarrays reduces measurement noise and improves phase shift estimation accuracy, which is a fundamental principle of signal processing. When multiple independent measurements of the same quantity are available, averaging them reduces the variance of the estimate by a factor of N (the number of measurements), thereby improving accuracy. This is important in radar systems where noise, clutter, and multipath effects often corrupt individual measurements. One is motivated to achieve the most accurate possible phase shift estimate by leveraging all available data from the multiple subarray pairs, rather than relying on a single potentially noisy pair. One would have had a reasonable expectation of success because: averaging is a well-established noise reduction technique with proven theoretical foundation; Bechter demonstrates that phase can be successfully measured across the virtual array elements, when the subarrays are uniformly spaced, they all measure the same underlying phase shift (plus noise), so averaging will converge to the true value, and the resulting averaged phase shift would have lower uncertainty and would lead to more accurate angle-of-arrival estimation.
Regarding Claim 7, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1, wherein computing the frequency spectrum for a subarray includes performing a Fast Fourier Transform, FFT ([0003]: “For frequency-modulated continuous-wave (FMCW) radar systems, for example, it is known to obtain information on range, speed, and angles by performing multiple Fast Fourier Transforms (FFTs) on samples of radar mixer outputs. A first FFT, also commonly referred to as range FFT, yields range information. A second FFT across the range transformed samples, also commonly referred to as Doppler FFT, yields speed information”; [0019]: “perform a first range FFT… perform a first Doppler FFT”).
Regarding Claim 9, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1, wherein the physical transmitters are used sequentially according to a transmission schedule, and said pair of subarrays are consecutive with respect to the transmission schedule ([0007-0008]: “The transmitter circuitry is configured to transmit, via a first subset of the transmit channels and during a first time interval, concurrent first frequency-modulated continuous-wave (FMCW) radar signals… transmit, via a second subset of the transmit channels and during a second time interval subsequent to the first time interval, concurrent second FMCW radar signals”; [0031]: “FIG. 5A shows a Time Division Multiple Access (TDMA) MIMO radar transmission method”.
Regarding Claim 11, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but it would have been obvious to extend the method to: further comprising determining a residual phase shift for all remaining subarrays of the virtual array and applying inverses thereof.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to determine and compensate phase shifts for all subarrays rather than just a subset. One would have been motivated to do so because the goal of phase ambiguity resolution is to synthesize the complete virtual array with proper phase alignment across all elements, thereby maximizing angular resolution and estimation accuracy. Park explicitly teaches that the objective is to combine all TDM virtual arrays to achieve full virtual array synthesis with maximum aperture ([0057]: “phase information from different CDM subsets may be merged via TDM synthesis 916-2 in order to obtain full virtual array synthesis”; [0093]: “the N rnM virtual arrays may be combined to one virtual array having N rXN R elements”). If phase compensation were applied to only some subarrays while leaving others uncompensated, those uncompensated subarrays would retain phase errors that would degrade the synthesized array performance. One would have had a reasonable expectation of success because: (1) Park explicitly teaches combining all virtual arrays and demonstrates that proper phase compensation yields higher amplitude in the angle spectrum ([0093]); (2) Bechter demonstrates that the phase compensation method successfully eliminates errors across the entire virtual array (Fig. 5, pg. 1166, showing corrected phases across all 18 virtual elements); (3) applying the same compensation method to all subarrays is a straightforward repetition of the process already described; (4) the compensation is based on the known TDM timing structure, so the required phase shift for each subarray can be calculated from the transmitter sequence; and (5) both references demonstrate that complete phase correction leads to successful angle-of-arrival estimation.
Regarding Claim 12, the claim is substantially the same as claim 1 and thus, the same cited sections and rationale as corresponding method claim 1 is applied.
Park et al. (‘386) in view of Bechter et al. teaches: and computing the angle or arrival on the basis of the processed virtual array signal (Park [0093-0094]: “The fine angle (θfine) between θcoarse-θth to θcoarse+θth can be estimated by taking the Digital Beam Forming (DBF) vector containing the maximum value. The amplitude of angle spectrum at target angle is higher when it is compensated correctly”; “The angular spectrum may be determined by performing DoA processing of the selected range-Doppler bins along a synthesized overall virtual receive channel domain”; [0013]: “determine a first angular spectrum associated with the first number of range-Doppler-maps by performing Direction-of-Arrival (DoA) processing”).
Regarding Claim 13, the claim is substantially the same as claim 1 and thus, the same cited sections and rationale as corresponding method claim 1 is applied.
Park et al. (‘386) teaches: the signal processing device comprising processing circuitry configured to resolve, in a virtual array signal comprising at least one range-Doppler bin, a phase ambiguity between the subarrays of the virtual array by performing a method comprising [all the method steps of claim 1] (Park [0100]: “Functions of various elements shown in the figures, including any functional blocks labeled as “means”… may be implemented in the form of dedicated hardware, such as “a signal provider”, “a signal processing unit”, “a processor”, “a controller”, etc. as well as hardware capable of executing software in association with appropriate software. When provided by a processor, the functions may be provided by a single dedicated processor, by a single shared processor, or by a plurality of individual processors, some of which or all of which may be shared”; [0100]: “digital signal processor (DSP) hardware, network processor, application specific integrated circuit (ASIC), field programmable gate array (FPGA), read only memory (ROM) for storing software, random access memory (RAM), and non-volatile storage”; [0094]: “for fine angle estimation, the receiver circuitry 620 may be further configured to combine first angular information… by applying a number of different phase offset candidates”).
Regarding Claim 14, the claim is substantially the same as claim 1 and thus, the same cited sections and rationale as corresponding method claim 1 is applied.
Park et al. (‘386) in view of Bechter et al. teaches: A non-transitory computer-readable storage medium having stored thereon instructions for implementing a method, when executed on a device having processing capabilities [all the method steps of claim 1] (Park [0101]: “a flow chart, a flow diagram, a state transition diagram, a pseudo code, and the like may represent various processes, operations or steps, which may, for instance, be substantially represented in computer readable medium and so executed by a computer or processor”; [0100]: “digital signal processor (DSP) hardware, network processor, application specific integrated circuit (ASIC), field programmable gate array (FPGA), read only memory (ROM) for storing software, random access memory (RAM), and non-volatile storage”).
Claims 8 and 10 are rejected under 35 U.S.C. 103 as being unpatentable over Park et al. (US 2021/0333386 A1) in view of Bechter et al. (Compensation of Motion-Induced Phase Errors in TDM MIMO Radars) and further in view of Chen et al. (US 2020/0233076 A1).
Regarding Claim 8, Park et al. (‘386) in view of Bechter et al. teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Chen et al. (‘076) teaches: wherein: the array of physical receivers has a plurality of rows in the first direction (Chen [0037]: “The transmitters and receivers may be arranged in any suitable manner including, for example, a one-dimensional array as shown in the illustrated example of FIG. 3 or in a two-dimensional array”; [0073]: “A uniform rectangular MIMO array may be fully described by four parameters including column (azimuth) spacing (dx), the row (elevation) spacing (dz), the number of columns (M), and the number of rows (N)”; [0076]: “the radar system 2100 includes any suitable number of transmitters 2102 and any suitable number of receivers 2104 arranged in any suitable manner in an antenna array”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the TDM MIMO radar system with phase ambiguity resolution of Park et al. (‘386) and Bechter et al. with the two-dimensional antenna array configuration of Chen et al. (‘076). One would have been motivated to do so because Chen explicitly teaches that two-dimensional arrays with multiple rows are advantageous for MIMO radar systems to achieve two-dimensional angle-of-arrival estimation ([0037], [0073]). Park also recognizes the need for spatial information in multiple dimensions ([0003]: “A third FFT involving phase information of signals of different antenna elements of an (virtual) antenna array can yield additional spatial or angular information”), and combining Park’s TDM-CDM approach with Chen’s 2D array structure would enable full 2D DoA estimation while maintaining the benefits of both TDM and CDM multiplexing. One would have had a reasonable expectation of success because: Chen demonstrates that uniform rectangular arrays with multiple rows work effectively in MIMO radar systems ([0073], [0085-0086]), the phase compensation methods taught by Park and Bechter apply to any receiver in the array regardless of whether receivers are arranged in one dimension or two dimensions—the TDM phase error depends on transmitter timing, not receiver spatial arrangement, all three references work with the same type of FMCW MIMO radar systems, ensuring compatibility, and Chen explicitly teaches processing methods that work with 2D arrays ([0073-0074], equations 36-41).
Park et al. (‘386) does not explicitly teach, but Chen et al. (‘076) teaches: the steps of computing a frequency spectrum and identifying an amplitude-peak frequency are performed for all rows (Chen [0073-0074]: “the signals received at each receiver corresponding to the different transmitters are arranged within a matrix corresponding to the rows and columns of a virtual uniform rectangular MIMO array”; [0086]: “the AOA analyzer 2126 calculates the angle of arrive (e.g., the azimuth and elevation) of targets detected by the receivers. In some examples, the AOA analyzer 2126 calculates the AOA based on an FFT analysis of the virtual array matrix generated by the virtual array generator 2124”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to perform the frequency spectrum computation and amplitude-peak frequency identification for all rows in Chen’s 2D array structure when combined with the processing methods of Park and Bechter. One would have been motivated to do so because Chen teaches that achieving accurate 2D angle-of-arrival estimation requires processing data from all rows of the array. Park teaches performing FFT operations on received signals, and Bechter teaches computing frequency spectra for each subarray. Combining theses teachings, one would naturally apply the frequency spectrum computation taught by Bechter to all rows of receivers taught by Chen in order to achieve the 2D DoA capability that Park identifies as desirable.
Park et al. (‘386) does not explicitly teach, but it would have been obvious to extend the processing to: and the residual phase shift is determined as a mean over all rows.
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to determine the residual phase shift as a mean over all rows. One would have been motivated to do so because when the physical receiver array has multiple rows extending in the first direction, each row provides an independent measurement of the same motion-induced phase error when processing along the first direction. The phase error caused by TDM operation depends on the transmitter switching time and target velocity, not on the secondary direction position of the receivers (i.e. x/y axis). Therefore, each row measuring along the x-direction experiences the same underlying phase error between subarrays, plus independent measurement noise. Averaging the phase shift measurement s from all rows would reduce the noise variance and improve estimation accuracy. The motivation would be to achieve more robust and reliable phase shift estimation by leveraging the redundance measurements available from multiple rows, thereby improving the overall phase compensation accuracy and subsequent angle-of-arrival estimation. Park et al. explicitly teaches that two-dimensional angle-of-arrival estimation (azimuth and elevation) is a goal of MIMO radar systems and achieving 2D DoA requires processing antenna arrays in multiple dimensions. Additionally, automotive radar applications, which both references address, require both azimuth and elevation angle estimation to fully localize targets in 3D space. Extending Bechter’s phase compensation method to multiple rows would enable full 2D processing while maintaining the motion-induced phase error correction.
Regarding Claim 10, Park et al. (‘386) in view of Bechter et al. and further in view of Chen et al. (‘076) teaches: The method of claim 1.
Park et al. (‘386) does not explicitly teach, but Chen et al. (‘076) teaches: wherein the array of physical receivers includes at least one column of physical receivers with a third spacing in a second direction (Chen [0037]: “The transmitters and receivers may be arranged in any suitable manner including, for example, a one-dimensional array as shown in the illustrated example of FIG. 3 or in a two-dimensional array”; [0073]: “A uniform rectangular MIMO array may be fully described by four parameters including column (azimuth) spacing (dx), the row (elevation) spacing (dz), the number of columns (M), and the number of rows (N)”; [0073]: “Based on these parameters, the position of an antenna element in the pth column and the qth row of an array is given by” Equation 36).
Park et al. (‘386) does not explicitly teach, but Chen et al. (‘076) teaches: and the physical transmitters are arranged with a fourth spacing in said second direction (Chen [0073]: describing the complete 2D array structure with spacing in both dimensions for both transmitters and receivers as part of the virtual uniform rectangular MIMO array; [0076]: “the radar system 2100 includes any suitable number of transmitters 2102 and any suitable number of receivers 2104 arranged in any suitable manner in an antenna array”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to combine the TDM MIMO radar system with phase ambiguity resolution of Park et al. (‘386) and Bechter et al. with the two-dimensional antenna array configuration having columns in a second direction with transmitters also arranged with spacing in that second direction, as taught by Chen et al. (‘076). One would have been motivated to do so because both Park and Chen recognize that two-dimensional spatial information is necessary for complete angle-of-arrival estimation in both azimuth and elevation. Park explicitly states that processing phase information of antenna array elements can “yield additional spatial or angular information-so-called Direction-of-Arrival (DoA) information” ([0003]), and Chen teaches that achieving full 2D DoA (both azimuth and elevation angles) requires a 2D array structure with spacing in two perpendicular directions ([0073], [0086]). The motivation would be to achieve complete 2D angle estimation capability (azimuth and elevation) while maintaining Park’s TDM-CDM multiplexing advantages and Bechter’s phase compensation benefits, especially since automotive radar applications, which all three references address, require both azimuth and elevation information to fully localize targets in 3D space for safe vehicle operation. One would have had a reasonable expectation of success because: Chen explicitly demonstrates that uniform rectangular 2D arrays work effectively in MIMO radar systems, with successful 2D angle-of-arrival estimation in both azimuth and elevation ([0073-0086]), the phase compensation methods taught by Park and Bechter apply to any receiver in the array regardless of the dimensional arrangement—the TDM phase error depends on transmitter timing and target velocity, not on the specific spatial configuration, all three references work with the same fundamental type of FMCW MIMO radar systems, ensuring signal compatibility, Chen’s processing methods using 2D-FFT operations ([0074], [0086]) are compatible with Park’s FFT-based processing ([0003], [0012]) and Bechter’s DFT-based compensation (equation 8, page 2), and Park explicitly contemplates virtual array synthesis from multiple transmitters and receivers arranged in arrays ([0004], [0030], Fig. 4]), and Chen’s 2D physical array is a natural extension that provides the additional dimension needed for full spatial coverage.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches processing methodology for the first direction and it would have been obvious to apply the same processing to the second direction: the method further comprising: for each of a second plurality of the subarrays, computing a frequency spectrum of those elements of the compensated virtual array signal which correspond to consecutive virtual antenna elements generated by physical receivers belonging to the same column (Bechter teaches computing frequency spectrum for subarrays along one dimension using equation 8, pg. 1165 for processing along rows)
Park et al. (‘386) and Bechter et al. do not explicitly teach, but Chen et al. (‘076) teaches the 2D array structure with columns in the second direction [0073]).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to compute frequency spectra for elements along columns (in the second direction) using the same methodology that Bechter teaches for computing frequency spectra along rows (in the first direction). One would have been motivated to do so because Chen teaches that 2D angle-of-arrival estimation requires processing in both spatial dimensions ([0073-0074], [0086]: describing elevation and azimuth calculation), and the motion-induced phase errors that Park and Bechter address occur due to TDM timing delays that affect all receivers regardless of their position in either the x or y direction. The phase error depends on the transmitter switching sequence and target velocity (Park [0088], Bechter equation 3), not on which spatial direction is being analyzed. Therefore, the same frequency spectrum computation method that compensates for TDM phase errors in the first direction would necessarily also compensate for the same phase errors when analyzing the second direction. The motivation would be to achieve accurate angle-of-arrival estimation in elevation (second direction) with proper phase compensation, just as Bechter achieves for azimuth (first direction). One would have had a reasonable expectation of success because: the mathematical FFT/DFT operations taught by Bechter (equation 8) are dimension-agnostic—they operate identically on data from a row or from a column, Bechter demonstrates successful elimination of motion-induced phase errors using the adapted DFT in one dimension (Figs. 4 and 6), Chen demonstrates that 2D-FFT processing works successfully for MIMO arrays ([0074], Equation 38), and the same FFT-based approach naturally incorporates Bechter’s phase compensation methodology, the TDM phase error is separable by dimension - compensating the error along columns is independent of compensating along rows, allowing the same technique to be applied in each dimension separately, and all three references process virtual array signals using FFT-based methods, demonstrating compatibility.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches peak identification for one direction and it would have been obvious to apply this to the second direction: identifying, jointly for the frequency spectra of said second plurality of the subarrays, a second amplitude-peak frequency (Bechter [pg. 1166]: “In the range-Doppler spectrum, a dominant target peak is selected for DoA estimation”; [pg. 1165]: “Fig. 4 shows the DoA estimation on the range-Doppler cell that contains the target” showing peak identification).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to identify the amplitude-peak frequency jointly for the frequency spectra computed along columns (second direction), using the same peak identification methodology that Bechter teaches for the first direction. One would have been motivated to do so because identifying the correct Doppler frequency bin containing the target is a necessary prerequisite for accurate phase comparison and angle-of-arrival estimation in any spatial direction. Bechter teaches selecting the dominant target peak for DoA processing, and this same target detection step would be required when processing along columns to ensure that phase measurements are made at the correct frequency corresponding to the actual target. The target’s Doppler frequency is the same regardless of which spatial direction is being analyzed—the target has one velocity, which produces one Doppler shift that appears in the frequency spectra of all subarrays whether arranged in rows or columns. The motivation would be to ensure accurate target identification in the elevation dimension before performing phase comparison operations. One would have had a reasonable expectation of success because: Bechter demonstrates successful target peak identification and subsequent accurate DoA estimation (Fig. 4), the Doppler frequency and amplitude peak are intrinsic properties of the target, not dependent on the spatial direction of analysis, peak detection in frequency spectra is a standard signal processing technique that works identically for any spatial dimension, and Chen demonstrates successful 2D processing that inherently requires identifying targets in both dimensions ([0086]).
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches phase shift determination for one direction and it would have been obvious to apply this to the second direction: determining a second residual phase shift between a second pair of the subarrays within said plurality of subarrays by comparing, at the amplitude-peak frequency, the respective phases of the frequency spectra (Bechter [pg. 1166, Fig. 5]: “Measured phases at the elements of the MIMO virtual array”; “At the position of the overlapping element there is a phase discontinuity of 1.54 rad for the conventional processing (6), due to the motion-induced phase error”; [equation 3, page 1164]: showing phase relationships between transmitters).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to determine a second residual phase shift between subarrays when processing along columns (second direction) by comparing phases at the amplitude-peak frequency, using the same phase comparison methodology that Bechter teaches for the first direction. One would have been motivated to do so because Park explicitly recognizes that phase differences exist between TDM time slots due to target velocity ([0087]: “TDM introduces phase difference ΦG due to the time gap G and the velocity v of target”), and this same phase error affects measurements in all spatial directions since it arises from the temporal switching between transmitters, not from spatial geometry. Bechter teaches measuring the phase discontinuity between subarrays by comparing phases, and this same measurement technique would identify the residual phase error when analyzing along columns. The TDM time gap creates the same phase error regardless of whether one is analyzing azimuth (along rows) or elevation (along columns) because the error depends only on the time delay and target velocity. The motivation would be to accurately measure the residual phase error in the elevation dimension so that proper compensation can be applied to achieve accurate 2D angle estimation. One would have had a reasonable expectation of success because: Bechter demonstrates successful phase discontinuity measurement of 1.54 rad, the mathematical operation of comparing phases is identical regardless of spatial direction—it involves computing the phase difference between complex numbers from the FFT output, the phase error mechanism taught by Park and Bechter applies equally to all spatial dimensions since it depends on transmitter timing, not receiver positions, and Chen’s 2D processing framework ([0073-0074]) provides the structure for performing independent phase measurements in both dimensions.
Park et al. (‘386) does not explicitly teach, but Bechter et al. teaches applying inverse phase compensation for one direction and it would have been obvious to apply this to the second direction: and applying an inverse of the second residual phase shift to the compensated virtual array signal (Bechter [pg. 1165, equations 7-8]: teaches application of phase compensation through the modified DFT; [pg. 1166]: “The phase-corrected range-Doppler processing (8) does not show such a severe deviation. Instead, the phase values at the position of the overlapping elements are nearly identical”).
It would have been obvious to a person of ordinary skill in the art before the effective filing date of the claimed invention to apply the inverse of the second residual phase shift (determined for the column/elevation direction) to compensate the virtual array signal in the second dimension. One would have been motivated to do so because Park teaches that phase compensation is necessary to synthesize the complete virtual array ([0088]: “ΦG may be compensated to synthesize aperture”; [0093]: describes phase rotation to combine virtual arrays), and achieving accurate 2D DoA requires proper phase alignment in both spatial dimensions. Bechter demonstrates that applying the inverse phase shift eliminates the motion-induced error (pg. 1166, Fig. 5 showing nearly identical phases after correction), and this same compensation principle applies when processing the second dimension. The mathematical principle is identical: if the TDM operation introduces a phase error of +φ in the elevation dimension, applying an inverse phase of -φ cancels that error. The motivation would be to achieve complete phase error correction in both azimuth and elevation, enabling accurate 2D angle-of-arrival estimation with the full angular resolution that the virtual array synthesis provides. One would have had a reasonable expectation of success because: Bechter demonstrates successful phase error elimination through inverse compensation (Fig. 4 showing identical DoA for v=18 m/s and v=0, pg. 1166), Park teaches that proper phase compensation results in higher amplitude in the angle spectrum ([0093]), and the same result would be achieved in the second dimension, Chen demonstrates successful 2D angle estimation ([0086]), showing that 2D processing is achievable, and the TDM phase error mechanism is the same in both dimensions, so the same inverse compensation technique that works in the first dimension would work in the second dimension.
Conclusion
Any inquiry concerning this communication or earlier communications from the examiner should be directed to REMASH R GUYAH whose telephone number is (571)270-0115. The examiner can normally be reached M-F 7:30-4:30.
Examiner interviews are available via telephone, in-person, and video conferencing using a USPTO supplied web-based collaboration tool. To schedule an interview, applicant is encouraged to use the USPTO Automated Interview Request (AIR) at http://www.uspto.gov/interviewpractice.
If attempts to reach the examiner by telephone are unsuccessful, the examiner’s supervisor, Vladimir Magloire can be reached at (571) 270-5144. The fax phone number for the organization where this application or proceeding is assigned is 571-273-8300.
Information regarding the status of published or unpublished applications may be obtained from Patent Center. Unpublished application information in Patent Center is available to registered users. To file and manage patent submissions in Patent Center, visit: https://patentcenter.uspto.gov. Visit https://www.uspto.gov/patents/apply/patent-center for more information about Patent Center and https://www.uspto.gov/patents/docx for information about filing in DOCX format. For additional questions, contact the Electronic Business Center (EBC) at 866-217-9197 (toll-free). If you would like assistance from a USPTO Customer Service Representative, call 800-786-9199 (IN USA OR CANADA) or 571-272-1000.
/REMASH R GUYAH/Examiner, Art Unit 3648
/VLADIMIR MAGLOIRE/Supervisory Patent Examiner, Art Unit 3648
Prosecution Insights