Method for robust radar detection and digitally modulated radar

Abstract

The method disclosed herein includes the steps of: generating (104) a radar signal in the digital domain along at least one transmission path, the radar signal comprising M periodic repetitions of a code sequence of length Lc multiplied by an asymptotic phase rotation, where Lc and M are integers, K is an integer or non-integer, and n is a discrete-time index corresponding to the code rate. The method further includes the step of: generating (105) a processing input signal in the digital domain along at least one reception path based on a digitized reflected signal corresponding to the radar signal, by multiplying the digitized reflected signal by an asymptotic phase rotation. In this context, K is defined such that the ratio is a non-integer, and M is defined such that the ratio is an integer. Alternatively, in the case where the code sequence is a Zadoff code sequence, K is defined such that the ratio is a non-integer, and M is defined such that the ratio is an integer.

Description

Technical Field

[0001] This disclosure relates to radar signal generation and reception, particularly in digitally modulated radar, to promote robustness to non-ideals associated with radar transceivers. Background Technology

[0002] Digitally modulated radars are becoming increasingly prominent due to their inherent ability to be programmed for different applications and their capacity to perform subsequent processing via digital signal processing techniques. Theoretically, these radars can combine high sensitivity (i.e., the ability to detect small targets) with high resolution and high-definition velocity (i.e., the ability to estimate target relative velocities without ambiguity). However, non-ideals associated with the radar transceiver significantly degrade radar performance. For example, DC offset in the received signal, nonlinearity in the baseband signal, and any mismatch between the I and Q signal components can make it possible to produce artifacts such as radar range profiles and range sidelobes or ghosting in the range Doppler image.

[0003] In traditional radar systems, the DC offset needs to be estimated to attenuate or compensate for it, a complex process requiring additional hardware and / or software support. Furthermore, some code sequences (e.g., Zadoff code sequences) are highly sensitive to third-order harmonic distortion introduced by nonlinearities in the baseband signal. Additionally, techniques for IQ imbalance robustness are only known for OFDM radars, not for all digitally modulated radars, such as phase-modulated continuous wave (PMCW) radars.

[0004] For example, document EP 3 627 787 A1 proposes a radar detection technique robust to IQ imbalance. Specifically, EP 3627 787 A1 discloses the generation of OFDM radar signals robust to IQ imbalance based on specific symmetry properties in the subcarrier domain. Summary of the Invention

[0005] One objective is to provide a method and digitally modulated radar that facilitate robust radar detection against non-ideals of conventional front-ends, addressing the aforementioned limitations. Therefore, embodiments of this application aim to provide an improved radar detection scheme and digitally modulated radar.

[0006] These and other objectives are achieved through embodiments of this application, as described in the appended independent claims. Advantageous implementations of these embodiments are further defined in the dependent claims.

[0007] According to a first aspect of this application, a method for promoting robust radar detection is provided. The method includes the steps of: generating a radar signal in a digital domain along at least one transmission path, the radar signal comprising multiplication by an asymptotic phase rotation. The method involves M periodic repetitions of a code sequence of length Lc, where Lc and M are integers, K is an integer or non-integer, and n is a discrete-time index corresponding to the code rate. The method further includes the step of: based on the digitized reflected signal corresponding to the radar signal, multiplying the digitized reflected signal by an asymptotic phase rotation... In the digital domain, a processing input signal is generated along at least one receiving path. In this context, K is defined such that the ratio... It is a non-integer, and M is defined such that the ratio It is an integer. As a supplement or alternative, in the case where the code sequence is a Zadoff code sequence, K is defined such that the ratio... It is a non-integer, and M is defined such that the ratio It is an integer.

[0008] Therefore, this disclosure proposes a simplified technique for generating radar signals robust to front-end non-ideals. This disclosure also proposes a simplified process for radar detection based on said radar signals, wherein the signal generation and said signal processing are not limited by the type of radar. Furthermore, this disclosure proposes signal processing criteria for achieving robust radar detection that is not limited by the type of modulation scheme used in digitally modulated radar.

[0009] Therefore, the constraints that must be maintained are limited only to factors such as the code sequence length Lc and variables such as K and M, which are defined or chosen regardless of the type of code sequence and / or modulation scheme. As a supplement or alternative, this disclosure provides signal processing criteria for achieving robust radar detection that can be implemented for a specific code sequence (i.e., a Zadoff code sequence).

[0010] In one implementation of the first aspect, the method further includes the step of defining K such that the ratio It is a non-integer and M is defined such that the ratio It is an integer. Therefore, this disclosure also proposes signal processing criteria for achieving robust radar detection to address additional non-ideals at the front end, which are not limited by the type of modulation scheme used in digitally modulated radar.

[0011] In one implementation of the first aspect, if the code sequence corresponds to a binary code sequence, for example, the modulation corresponds to binary phase shift keying (BPSK), then the method further includes the step of defining K as an integer or non-integer given by the following formula:

[0012]

[0013] Where d is an integer.

[0014] Therefore, this disclosure also proposes modulated-by-modulation (BPSK) implementations of the method of the invention to address additional non-ideals at the front end. Specifically, this disclosure further addresses power amplifier nonlinearity and proposes a radar signal generation and processing scheme that stabilizes said nonlinearity. Power amplifier nonlinearity is caused, in particular, by nonlinear distortion introduced in the front-end amplification stage (e.g., within the power amplifier), and further introduces unwanted artifacts (e.g., range sidelobes or ghosting) into the range image.

[0015] In one implementation of the first aspect, the method further includes the step of performing digital processing on the input signal in relation to a code sequence to generate a series of range profiles. Additionally, the method includes the step of accumulating M consecutive range profiles. For example, the digital processing, range processing, or processing algorithm can be defined based on the type of radar. For PMCW radar, range processing can be based on a correlation algorithm. For OFDM radar, the processing algorithm can be an FFT-based algorithm (frequency domain processing).

[0016] Generally, radar signals consist of periodic repetitions of code sequences, and reflected signals typically consist of a linear combination of delayed versions of the code sequences. This delay is proportional to the target range position. To estimate this delay, the radar performs range processing on the received reflected signals to produce a range profile (e.g., in the case of a PMCW radar, related to the complex conjugate of the transmitted code sequences). The range profile contains a peak value at the location of each target, which is proportional to the power reflected back from the target. Furthermore, accumulating multiple coherent range profiles (e.g., M range profiles) advantageously improves, for example, the signal-to-noise ratio.

[0017] In one implementation of the first aspect, the method further includes the step of generating a radar signal comprising multiple periodic repetitions of a code sequence having a number of M periodic repetitions. Specifically, the radar signal may comprise N periodic repetitions of the code sequence having M periodic repetitions, where N is an integer. Additionally, the method includes the step of processing N range images to generate a range-Doppler map. In other words, the number of repetitions N defines the number of samples used for Doppler processing to estimate the target velocity.

[0018] In one implementation of the first aspect, the method further includes the step of generating multiple radar signals along multiple transmission paths in the digital domain, each of the radar signals comprising the same code sequence length, the same number of periodic repetitions of the code sequence, and non-identical asymptotic phase rotations relative to the multiple radar signals, wherein each of the transmission paths is associated with a dedicated radar signal of the multiple radar signals. In this aspect, the orthogonality between the radar signals of each of two transmission paths in the corresponding multiple transmission paths is satisfied by the following:

[0019]

[0020] Where K p It is an integer or non-integer for the first transmission path, and K q It is an integer or non-integer for the second launch path.

[0021] Additionally, the method includes the step of generating multiple processing input signals along multiple receiving paths in the digital domain, each of which is associated with a process input signal generated based on a digitized reflection signal corresponding to a specific dedicated radar signal in the multiple transmission paths.

[0022] Therefore, this disclosure also proposes a multiple-input multiple-output (MIMO) scheme for digitally modulated radar. This disclosure also proposes a solution that effectively allows a receiver to distinguish one or more signals from each of the transmitters.

[0023] According to a second aspect of this application, a digitally modulated radar for promoting robust radar detection is provided. The digitally modulated radar includes at least one code generation unit configured to generate radar signals in the digital domain along at least one transmission path, the radar signals including multiplication by an asymptotic phase rotation. The number of periodic repetitions of a code sequence of length Lc is M, where Lc and M are integers, K is an integer or non-integer, and n is a discrete-time index corresponding to the code rate.

[0024] The digitally modulated radar also includes at least one processing unit configured to rotate the digitized reflected signal corresponding to the radar signal with an asymptotic phase in the digital domain. The multiplication is used to generate a processed input signal along at least one receiving path based on the digitized reflected signal. In this context, the at least one code generation unit is also configured to define K such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer. As a supplement or alternative, when the code sequence is a Zadoff code sequence, at least one code generation unit is also configured to define K such that the ratio... It is a non-integer, and M is defined such that the ratio It is an integer.

[0025] In one implementation of the second aspect, the at least one code generation unit is further configured to define K such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer.

[0026] In one implementation of the second aspect, where the code sequence corresponds to a binary code sequence, at least one code generation unit is further configured to define K as an integer or non-integer given by the following formula:

[0027]

[0028] Where d is an integer.

[0029] In one implementation of the second aspect, the at least one processing unit is further configured to perform digital processing on the processing input signal in relation to a code sequence to generate a series of range images. Thereafter, the at least one processing unit is also configured to accumulate M consecutive range images.

[0030] In one implementation of the second aspect, the at least one code generation unit is configured to define Lc with respect to the code rate relative to the radar's maximum unambiguous range. Alternatively, the at least one code generation unit also defines Lc with respect to a given radar bandwidth relative to the radar's maximum unambiguous range. In this respect, the radar's maximum unambiguous range is defined as the maximum range that the radar can cover without ambiguity. Preferably, the radar bandwidth is defined based on a preferred range resolution (i.e., a measure of the minimum spacing distance of its targets). Furthermore, the maximum unambiguous range depends on the code sequence, particularly the code sequence duration, which is a function of the code rate or the radar bandwidth.

[0031] In one implementation of the second aspect, the digitally modulated radar includes multiple transmission paths, each transmission path including the at least one code generation unit. Furthermore, for each transmission path, the corresponding code generation unit is configured to generate a dedicated radar signal in the digital domain, the dedicated radar signal having the same code sequence length, the same number of periodic repetitions of the code sequence, and non-identical asymptotic phase rotations relative to the multiple transmission paths. In this context, the orthogonality between the radar signals of each of two transmission paths in the corresponding multiple transmission paths is satisfied by the following:

[0032]

[0033] Where K p It is an integer or non-integer for the first transmission path, and K q It is an integer or non-integer for the second launch path.

[0034] Furthermore, the digitally modulated radar includes multiple receiving paths, each of which includes the at least one processing unit. Additionally, for each receiving path, the corresponding processing unit is configured to generate a processed input signal in the digital domain based on digitized reflected signals corresponding to the respective dedicated radar signals of the multiple transmitting paths.

[0035] It should be noted that the digital modulation radar according to the second aspect corresponds to the method and its various implementations according to the first aspect. Accordingly, the digital modulation radar of the second aspect achieves the same advantages and effects as the method and its corresponding implementations of the first aspect, and vice versa. Attached Figure Description

[0036] The above aspects and implementations will be explained in the following description of specific embodiments with reference to the accompanying drawings, in which:

[0037] Figure 1 An exemplary embodiment of a method according to one aspect of this disclosure is shown;

[0038] Figure 2 An exemplary transmission frame for acquiring a radar data cube is shown;

[0039] Figure 3A An ideal M-sequence distance image with two targets is shown;

[0040] Figure 3B An ideal APS range image with two targets is shown;

[0041] Figure 4 A first exemplary embodiment of a radar transceiver according to one aspect of this disclosure is shown;

[0042] Figure 5A It shows in detail Figure 4 The code generation unit;

[0043] Figure 5B It shows in detail Figure 4 Processing unit;

[0044] Figure 6A An exemplary transmitting side of a second exemplary embodiment of a radar transceiver according to one aspect of the present disclosure is shown;

[0045] Figure 6B An exemplary receiving side of a second exemplary embodiment of a radar transceiver according to one aspect of this disclosure is shown;

[0046] Figure 7 Detailed illustration Figure 6B An exemplary receiving path on an exemplary receiving side;

[0047] Figure 8A The accumulated M-sequence range profile with a sequence length of 511 and π / 2-BPSK modulation is shown.

[0048] Figure 8B The accumulated range profile of an M-sequence with a sequence length of 511 and π / 3-BPSK modulation is shown; and

[0049] Figure 9 The distance image from a Zadoff code sequence with three targets is shown in the presence of front-end non-ideality. Detailed Implementation

[0050] Reference will now be made in detail to various embodiments of the present disclosure, examples of which are illustrated in the accompanying drawings. However, various modifications may be made to the following embodiments of the present disclosure, and the scope of the present disclosure is not limited to these embodiments. Reference numerals for similar entities in different embodiments have been partially omitted.

[0051] exist Figure 1 The diagram illustrates an exemplary embodiment of the method 100 according to a first aspect of this disclosure. In a first step 101, a code sequence having a code sequence length Lc is selected. In a second step 102, a value of parameter K is selected such that the ratio Lc / 2·K is a non-integer. Alternatively, in the second step 102, for the Zadoff code sequence, a value of parameter K is selected such that the ratio 2·Lc / K is a non-integer. In a third step 103, a value of parameter M is selected such that the ratio Lc·M / 2·K is an integer. Alternatively, in the second step 103, for the Zadoff code sequence, a value of parameter M is selected such that the ratio 2·Lc·M / K is an integer.

[0052] In the fourth step 104, in the digital domain, the M periodic repetitions of the code sequence (preferably N periodic repetitions of the M periodic repetitions of the code sequence) are rotated with an asymptotic phase. The radar signal is generated by multiplication. In step 105, the digitized reflected signal corresponding to the radar signal is multiplied with an asymptotic phase rotation in the digital domain. The digitized reflected signal is multiplied to generate a processed input signal. Finally, in step 106, distance processing is performed on the processed input signal.

[0053] In other words, radar signals are generated by multiplying the periodic repetition of a code sequence with an asymptotic phase rotation. The processed input signal is generated by multiplying the digitized reflected signal of the radar signal (hereinafter referred to as the reflected signal for simplicity) with an asymptotic phase rotation having the opposite rotation angle. Thus, for two asymptotic phase rotations at a given time, the discrete-time index remains the same. The aforementioned criteria for parameters Lc, M, and K may need to be satisfied individually based on the desired code sequence, and may also need to be satisfied in combination specifically for Zadoff code sequences.

[0054] Range processing typically involves the generation of range profiles. In this regard, the method comprises the following sequential steps: performing range processing on the input signal related to a code sequence to generate a series of range profiles, and further accumulating M consecutive range profiles to improve the SNR.

[0055] exist Figure 2 The diagram illustrates an exemplary transmission frame 200 for acquiring a radar data cube. Transmission frame 200 is shown as a typical transmission frame for a single-input single-output (SISO) digital modulation radar (e.g., a SISO-PMCW radar or a SISO-OFDM radar). The code sequence length Lc defines the number of samples or chips 201 in the code sequence 202 (as shown herein as "s"). In other words, the code sequence length Lc defines the number of range bins used for range processing operations. Thus, the so-called pulse repetition frequency of a digital modulation radar can be defined, for example, as a sequence of length Lc that repeats itself at a chip rate fc.

[0056] The parameter M defines the number of times the accumulated identical code sequence "s" is repeated 203, similar to conventional pseudo-random code sequences such as Maximum Length Sequences (MLS). In other words, the number of repetitions M defines the number of range images to be coherently accumulated. The parameter N defines the number of times the M sequences 203 are repeated 204, i.e., the number of samples used for Doppler processing to estimate target velocity. Therefore, the code sequence "s" is repeated M·N times in transmission frame 200 to acquire a radar data cube.

[0057] During phase detection, the transmitted signal is reflected from all targets or obstacles in front of the radar. This produces a linear combination of delayed versions of the code sequence “s”. This delay is proportional to the target range position. To estimate this delay, the received signal is digitally processed, for example, correlated with the transmitted code sequence “s” in a PMCW radar (e.g., for binary code sequences), or correlated with the complex conjugate of the transmitted code sequence “s” (e.g., for complex code sequences). This produces a range profile containing a peak at the location of each target, the peak being proportional to the power reflected back from said target.

[0058] exist Figure 3A and 3B The image shown is an exemplary range profile for an ideal radar transceiver. Specifically, Figure 3A An ideal range profile with two targets for an M-sequence (MLS) is shown. The horizontal axis represents the distance to the targets in meters, and the vertical axis represents the reflected power amplitude in decibels. Here, the two targets are indicated by circles 301 and 302 at their respective peaks. On the other hand, Figure 3B An ideal range image with two targets is shown for a near-perfect sequence (APS). The horizontal axis represents the distance to the targets in meters, and the vertical axis represents the reflected power amplitude in decibels. Here, the two targets are indicated by circles 303 and 304 at the corresponding peaks.

[0059] Ideally, the resulting range image should contain zeros within the range interval, and these zeros should not correspond to any target. In practice, some non-null values ​​may appear to originate from code sequence characteristics. These values ​​are called range sidelobes. For example, in the case of PMCW radar, the M-sequence always produces range sidelobes, which are more pronounced than... Figure 3A The correlation peak shown is -20log10(Lc)dB lower, where Lc is the code sequence length or the number of samples in the code sequence. On the other hand, APS does not produce any sidelobes, such as Figure 3B As shown. An ideal code sequence should have the following properties:

[0060]

[0061] Where b c (n) is a code sequence in the digital field, D is the cyclic shift in the code sequence corresponding to the target distance, and ()* is the complex conjugate operation.

[0062] code sequence b c The number of samples Lc in (n) defines the number of distance intervals in the range image. A high Lc value is needed to cover long distances. The code sequence is continuously repeated to generate the signal b(n), for example, as in... Figure 2 The M·N repetitions are shown. For example, in the case of PMCW radar, each of these repetitions produces a range image defined by the following formula:

[0063]

[0064] Where u(n) is the signal received at the ADC output, m=0,…,M-1, and k=0,…,N-1.

[0065] Generally, the choice of the sequence "s" depends on the type of digitally modulated radar, such as a phase-modulated continuous wave (PMCW) radar or an orthogonal frequency division modulated (OFDM) radar. However, OFDM radar uses the same frame structure. The only difference from PMCW is the range processing operation.

[0066] In the case of PMCW radar, range processing is performed by means of correlation. Therefore, code sequence selection depends on the periodic autocorrelation characteristics of the code sequence, as defined in equations (1) and (2). Here, code sequences that satisfy the above autocorrelation characteristics are selected. In this regard, the APS and M sequences are exemplified in this disclosure, which perfectly satisfy the above autocorrelation characteristics. Specifically, the APS produces two non-zero values ​​in the range image instead of just one, and the M sequence gives an amplitude of 1 instead of 0 when Δ≠0.

[0067] In the case of OFDM radar, range processing is performed through a process similar to OFDM channel equalization in wireless communication. This operation extracts the range image and removes the code sequence itself. Therefore, each sequence of the Lc complex sample can be used, and thus there are no constraints on the values ​​of the complex sample.

[0068] As mentioned above, M consecutive range images are accumulated to improve the signal-to-noise ratio (SNR), while Doppler processing is achieved through a slow discrete Fourier transform (DFT), i.e., N repetitions of the M code sequences. Multiple-input multiple-output (MIMO) techniques can also be applied if multiple antennas are used.

[0069] Generally speaking, M consecutive distance images r of Lc distance intervals m,k (D) can be accumulated to produce a range image r with improved SNR. k (D)

[0070]

[0071] Where D is the distance range index.

[0072] Therefore, r m,k (D) is a distance interval value, where the distance image is for all distance intervals r for all values ​​D = 0, 1, ..., Lc-1. m,k The set of (D). Similarly, r k (D) is a distance interval value, where the distance image is for all distance intervals r for all values ​​D = 0, 1, ..., Lc-1. k The set of (D).

[0073] Each group of M code sequences is repeated N times. These repetitions are used to compute the Doppler profile for each distance interval using the Discrete Fourier Transform (DFT):

[0074]

[0075] Where l is the Doppler interval index.

[0076] R(D,L) is called a range-Doppler map and is the value of a range-Doppler cell. A range-Doppler map is the set of all range-Doppler cells R(D,l) for all values ​​D = 0, 1, ..., Lc-1 and l = 0, 1, ..., N-1. Therefore, for each range interval D, it contains a Doppler image defined by index l. Each index l is a Doppler interval corresponding to the Doppler frequency shift attributed to the relative velocity of the target.

[0077] To observe high-speed targets, radar must be able to detect high Doppler frequency shifts. The latter is limited by parameters Lc, M, and Tc, where Tc is the chip rate given by the following formula:

[0078]

[0079] This value is also known as fuzzy Doppler. The product LcMTc gives the duration of M code repetitions.

[0080] For an ideal transceiver, the response to a point target produces spikes in the range image or range-Doppler plot, which may be limited by the theoretical sidelobes of the code sequence. However, for a non-ideal transceiver, the ideal response is degraded, which typically manifests itself as the appearance of ghosted targets or increased sidelobes. Typical sources of non-ideality are power amplifier nonlinearity, IQ mismatch, receiver DC offset, or baseband nonlinearity.

[0081] The sources of these non-ideals can be illustrated with reference to conventional radar transceivers. For example, IQ mismatch or imbalance can occur in the mixers present on both the transmitter and receiver sides. Power amplifier nonlinearity can occur in the power amplifier on the transmitter side. Baseband nonlinearity can occur after and / or before the domain conversions (digital to analog and analog to digital) of the various signals on both the transmitter and receiver sides. This baseband nonlinearity can correspond to second-order harmonic distortion and / or third-order harmonic distortion.

[0082] However, the receiver DC offset has several sources. Since this is an additive process, the effects of all DC offset sources can be combined, for example, in the coefficient γ, which will be described later in this disclosure.

[0083] Furthermore, using multiple antennas allows for the advantageous combination of Multiple-Input Multiple-Output (MIMO) techniques. A significant challenge of MIMO is allowing the receiver to distinguish signals from each transmitter. For example, in the case of time-domain MIMO, each transmitter must transmit the same signal in a different time slot before moving to the next slower time point. Therefore, increasing the number of transmitters degrades Doppler ambiguity. On the other hand, in the case of Doppler-domain MIMO, each transmitter appears in a different part of the Doppler image. This means that in the presence of N... tx In the case of multiple transmitters, each target will appear N in the Doppler image. tx This makes detection more complex and also shows that it degrades the ambiguity Doppler frequency shift. Generally, the ambiguity Doppler in MIMO radar can be expressed as:

[0084]

[0085] Where N tx It refers to the number of transmitters.

[0086] exist Figure 4 The image shows a first exemplary embodiment of a digitally modulated radar 400 according to a second aspect of this disclosure. Specifically, the digitally modulated radar 400 is explained according to a single-input single-output (SISO) implementation. The digitally modulated radar 400 includes a code generation unit (CGU) 401 that generates a radar signal 402 in the digital domain. A digital-to-analog converter (DAC) 403 downstream of the CGU 401 converts the digital radar signal 402 in the analog domain, thereby generating an analog radar signal 404. A low-pass filter (LPF) 405 downstream of the DAC 403 performs baseband filtering on the analog radar signal 404, i.e., filters out higher frequency components, thereby generating a filtered radar signal 406.

[0087] Downstream of LPF 405, mixer 407 modulates the filtered radar signal 406 with a carrier sine wave, preferably generated by a local oscillator (not shown), thereby generating an RF radar signal 408. An amplifier, particularly a power amplifier (PA) 409, amplifies the RF radar signal 408, thereby generating a radar transmission signal 410 transmitted via transmitter antenna 411. Therefore, the transmission path for transmitting radar signal 402 may include DAC 403, LPF 405, mixer 407, PA 409, and optionally CGU 401 and transmitter antenna 411.

[0088] The radar transmitted signal 410 is typically reflected from all targets, and the resulting echo, or reflected signal, or radar received signal 430, is received by the receiving antenna. Generally, only a portion of the transmitted signal 410 is reflected back to the receiving antenna 431. An amplifier, particularly a low-noise amplifier (LNA) 429, amplifies the radar signal 430, thereby generating an amplified radar received signal 428. A mixer 427 downstream of the LNA 429 demodulates the amplified radar received signal 428 using a carrier signal generated by a local oscillator, i.e., a direct conversion of the amplified radar received signal 428, thereby generating a baseband signal 426.

[0089] A low-pass filter (LPF) 425 downstream of mixer 427 filters the baseband signal 426, thereby generating a filtered baseband signal 424. An analog-to-digital converter (ADC) 423 downstream of LPF 425 converts the filtered baseband signal 424 into the digital domain, thereby generating a reflected signal 422 in the digital domain corresponding to the radar signal 402 generated by CGU 401 in the digital domain. The digitally modulated radar 400 also includes a processing unit 421 downstream of ADC 423, which generates a processed input signal in the digital domain based on the reflected signal 422 to perform range and Doppler processing. Therefore, the receiving path or acquisition path for receiving the reflected signal 422 may include LNA 429, mixer 427, LPF 425, ADC 423, and optionally processing unit 421 and receiving antenna 431.

[0090] It is understood that the transmitting antenna 411 and the receiving antenna 431 can be implemented as dedicated antennas in an antenna array for the respective transmitting and receiving paths of the digital modulation radar 400. It is also conceivable that the transmitting antenna 411 and the receiving antenna 431 can be implemented in a single antenna array and that this antenna array can operate in an interchangeable manner, for example, by means of a circulator, for the transmitting and receiving paths of the digital modulation radar 400.

[0091] CGU 401 is also configured to communicate with processing unit 421 to provide information related to the selected code sequence used for range processing. The communication signal is shown as dashed line 412. It is conceivable that CGU 401 and processing unit 421 could be implemented as a single entity, for example, as a baseband processing unit. Furthermore, additional means for generating and / or processing radar signals, such as memory or storage for storing code sequences, control commands, etc., and interfaces such as user interfaces, are not explicitly shown, but are evident in the above implementation.

[0092] As previously mentioned, due to the non-ideal nature of mixers 407 and 427, the I and Q components may not be perfectly orthogonal, and therefore, amplitude and / or phase mismatch may occur between the I and Q components. Furthermore, the nonlinearity of PA 409 further degrades radar performance by producing range sidelobes or ghosting in the range image. The analog radar signal 404 and the filtered baseband signal may suffer from baseband nonlinearity in terms of second-order and / or third-order harmonic distortion. The effects of LNA 429, mixer 427, LPF 425, and ADC 423 can additionally cause receiver DC offset that can be applied to the reflected signal 422.

[0093] exist Figure 5A and Figure 5B In the diagram, the generation of radar signal 402 in CGU 401 and the generation of processing input signal 524 in processing unit 421 are shown in detail. Figure 5A The CGU 401 includes a signal generator 501 that generates chip signals or chips based on a predefined code sequence. This code sequence may correspond to, for example... Figure 2 The transmission frame shown comprises a sequence “s” of Lc chips, which is repeated M times, and finally, the M repetitions of “s” are further repeated N times. In other words, the sequence “s” is repeated M·N times in a single transmission frame.

[0094] Here, the code sequence is shown as b c (n), where n defines the discrete-time index corresponding to the code rate. In conventional digital modulation radar, the code sequence b c (n) is repeatedly transmitted as a radar signal, i.e., the repeated code sequence b. c (n) is directly used for the successive conversion, filtering, and amplification stages. However, in order to generate a radar signal robust to the non-ideals of the transceiver, the code generation unit 401 also includes b c The periodic repetition of (n) (shown here as b(n)) multiplied by an asymptotic phase rotation The multiplier 503. Therefore, radar signal 402 can be expressed in digital code sequence as:

[0095]

[0096] Where b(n) is b c The periodic repetition of (n).

[0097] On the other hand, the processing unit 421, in such a way Figure 4 The described baseband-converted received reflected signal 422 (shown herein as u(n)). Processing unit 421 includes a progressive phase rotation of the reflected signal u(n). The multiplier 523 performs the multiplication. This generates a processing input signal 524 (shown here as v(n)), which is used for subsequent processing. Therefore, the processing input signal v(n) can be expressed as:

[0098]

[0099] It can be seen that the asymptotic phase rotation of the received signal has a rotation angle opposite to that of the asymptotic phase rotation of the transmitted signal. Therefore, multiplier 523 effectively removes the complex exponent.

[0100] Processing unit 421 further includes processing block 525, which, for example, uses code sequence information 412 fed from CGU 401 (especially signal generator 501) related to code sequence b. c(n) performs range processing on the input signal 524 to generate a radar range image. The range processing may be based on conventional correlation algorithms (e.g., for PMCW radar) or frequency domain range processing (e.g., for OFDM radar). Processing block 525 also accumulates M coherent range images, i.e., sums M coherent range intervals, to remove range sidelobes from non-ideals, which will be described in detail in subsequent sections. The accumulation of M coherent images further improves, for example, the SNR. Processing unit 421 also includes a Discrete Fourier Transform (DFT) block 527, which performs a DFT on N samples to produce a Doppler image.

[0101] It should be noted that the processing unit 421 can be implemented by hardware, software, or any combination thereof. The processing unit 421 may include one or more application-specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field-programmable gate arrays (FPGAs), processors, controllers, microcontrollers, microprocessors, etc.

[0102] The parameters Lc, K, and M and their effects on the reduction of range sidelobes (especially due to IQ imbalance) are described in detail below.

[0103] The reflected signal 422 can be described in the time domain as follows:

[0104] u(t)=αy(t)+βy * (t) (9)

[0105] Where y(t) is the reflected signal without IQ imbalance.

[0106] Therefore, in the ideal case, when there is no amplitude or phase mismatch between the I and Q branches, α = 1 and β = 0.

[0107] Therefore, for all real values ​​of b(n), IQ mismatch on the transmit and / or receive paths will produce range sidelobes, such as:

[0108]

[0109] Where Δ is the propagation delay.

[0110] In equation (10), the first term It is a delayed version of the code sequence multiplied by a constant phase rotation, which depends on the propagation delay Δ. Therefore, for this first term, the autocorrelation property of the code sequence is unaffected.

[0111] However, the second item It is also a delayed version of the code sequence multiplied by a constant and asymptotic phase rotation. Additionally, this term includes the asymptotic phase rotation. This asymptotic phase rotation disrupts the correlation properties of the code sequence, leading to range sidelobes. However, under the selection criterion of K, this asymptotic phase rotation will be mitigated during radar digital processing.

[0112] For example, in the case of a PMCW radar with a real code sequence b(n), the first three range profiles can be expressed as:

[0113]

[0114]

[0115]

[0116] Where b c (n+D) is the code sequence b c Circular shift in (n).

[0117] Each distance like r k (D) Affected by the distance sidelobe. This can be described as:

[0118]

[0119] Where r sl,k (D) contains information from r k (D) distance sidelobe.

[0120] Equation (12) can be rewritten as:

[0121]

[0122] Because b(n) is a b containing Lc samples c The periodic repetition of (n).

[0123] As can be seen, all range images will have the same range sidelobes as follows:

[0124]

[0125] The range sidelobe is related to the phase rotation, which depends on the range image index. Multiply. Therefore, equation (13) can be rewritten in simplified form as:

[0126]

[0127] From equation (15), it can be deduced that if Lc is not a multiple of K, the range sidelobes in the coherent range image will be phase-dependent and thus decay during the accumulation period.

[0128] Furthermore, it can be deduced from equation (15) that if Lc·M / K is an integer, then the phase of the range sidelobe at index D in all M range images is equally distributed on the complex circle. Therefore, the accumulation of M consecutive range images eliminates the range sidelobe, as shown in the following equation:

[0129]

[0130] This can also be extended to complex code sequences and OFDM radar, as they transmit complex digitally modulated signals. Especially for complex code sequences, ghosting can be observed... This is similar to Equation (10), but with the complex operator ()*, i.e., the complex conjugate of the code sequence. However, they are affected by the same phase rotation due to K. Therefore, the solution proposed here can also be used to effectively compensate for the range sidelobes caused by the IQ imbalance (on the transmitter and receiver sides) using complex code sequences.

[0131] The following sections describe in detail the parameters K and their effects on range sidelobe reduction (especially due to power amplifier nonlinearity). The following techniques are effective for binary code sequences, such as binary phase-modulated radars like PMCW radar.

[0132] The code sequence b(n) is considered as a binary code sequence, which is multiplied in the digital domain by a complex exponent and an asymptotic phase rotation, as described above. This produces the radar signal s(n), as shown in equation (7). This radar signal is then converted in the analog domain and modulated into a carrier sine wave.

[0133] Generally, the transmission is bandwidth-limited, which can be modeled using the impulse response h(n) via the LPF 405. The signal is then amplified by the PA 409. To reduce the power consumption of the transmitted radar signal, the PA 409 is preferably operated close to its saturation point. However, this introduces nonlinear distortion in the transmitted signal, which degrades radar performance.

[0134] The proposed technique creates waveforms capable of eliminating the largest sources of distance sidelobes. For this reason, the following two assumptions are made:

[0135] 1) The nonlinear model of a power amplifier can be approximated by a cubic model:

[0136] y(nT c )=a1x(nT c )+a3x(nT c )|x(nT c )| 2 (17)

[0137] This assumption is actually satisfied when there are no even nonidealities in the baseband power amplifier model and the third-order nonlinearity is actually dominant.

[0138] 2) Only three taps in the LPF discrete impulse response are significant:

[0139]

[0140] Since h(n) is a low-pass filter, this assumption is realistic.

[0141] The input signal v(n) is processed by a series of terms that are linear combinations of binary code sequences b(n), which depend on the target position and contain useful information. Other terms depend on nonlinear combinations of b(n). These nonlinear combinations will produce range sidelobes or ghosting in the range image. However, for each case, all nonlinear combinations are multiplied by:

[0142]

[0143] For a value of K where z(K) = 0, no range sidelobes will be produced. At least, the sidelobes and ghosting generated by the three most significant taps of the impulse response h(n) and the third-order nonlinearity are eliminated due to the waveform characteristics.

[0144] According to equation (19), the parameter K that eliminates the nonlinear combination of equation (19) can be expressed in simplified form as:

[0145]

[0146] Where d is an integer. Therefore, possible ideal values ​​of K can be expressed as: ±3, ±1.5, etc.

[0147] The selection criteria for parameters Lc, K, and M used to achieve receiver DC offset robustness are described in detail below.

[0148] In the presence of DC offset, the reflected signal 422 at ADC 423 can be described as:

[0149] u(n)=u0(n)+γ (21)

[0150] Where u0(n) is the ideal received signal without DC offset, and γ is a complex number that defines the cumulative DC offset in the receive path.

[0151] To remove the DC offset, parameters K, Lc, and M must be selected to satisfy the following two criteria:

[0152] 1) Lc is not a multiple of 2K, and

[0153] 2) It is an integer.

[0154] Attenuation is possible because the π / K modulation operation is also applied to the DC offset as follows:

[0155]

[0156] Therefore, the phase of the DC offset will vary with distance from the image. The K value, selected using the above criteria, can be removed during coherent accumulation.

[0157] The selection criteria for parameters Lc, K, and M used to achieve baseband nonlinear robustness are described in detail below.

[0158] When nonlinearity exists in the baseband, all types of code sequences will produce range sidelobes and / or ghosting targets. This is especially true for Zadoff code sequences, which are highly sensitive to third-order harmonic distortion (HD3) in the baseband.

[0159] Zadoff code sequences are well-known complex code sequences, defined as follows:

[0160]

[0161] Ideally, they do not produce range sidelobes. However, range sidelobes occur in the presence of front-end nonidealities. In particular, they are highly sensitive to third-order nonlinearity in the baseband.

[0162] Based on the effect of third-order harmonic distortion on Zadoff code sequences, it seems possible to attenuate the range sidelobes generated by this nonlinearity. To achieve this, the parameters K, Lc, and M must be chosen to satisfy the following two criteria:

[0163] 1) 2Lc is not a multiple of K, and

[0164] 2) It is an integer.

[0165] However, this solution only works for Zadoff code sequences, as other code sequences do not exhibit such properties.

[0166] exist Figure 6A and Figure 6B The following illustrates a second exemplary embodiment of digital modulation radars 600A and 600B according to a second aspect of this disclosure. Specifically, the digital modulation radars 600A and 600B are explained according to a multiple-input multiple-output (MIMO) implementation. In this context, Figure 6A The explanation focused on the 600A multi-antenna transmitter side of the MIMO radar, and... Figure 6B The 600B multi-antenna receiver side of the MIMO radar was explained.

[0167] refer to Figure 6A The transmitter 600A includes multiple transmission paths 1-Ntx. Each of the multiple transmission paths includes code generation units (CGUs) 6011-601. Ntx It generates corresponding radar signals 6021-602 in the digital domain. Ntx However, it is also conceivable that CGU 6011-601 Ntx It can be implemented as a single unit, which is operatively coupled to each of multiple transmission paths and generates corresponding radar signals 6021-602 sequentially or simultaneously for the respective transmission paths. Ntx The radar signals 6021-602 Ntx They are the same in terms of code sequence length and the number of periodic repetitions of the code sequence, however, they include asymptotic phase rotations that are different from each other.

[0168] Each of the multiple launch paths also includes the corresponding CGU 6011-601. Ntx Downstream digital-to-analog converters (DACs) 6031-603 Ntx And convert the corresponding digital radar signals 6021-602 in the analog domain. Ntx This generates analog radar signals 6041-604 for the corresponding transmission paths. Ntx Each of the multiple transmit paths also includes the corresponding DAC 6031-603. Ntx Downstream low-pass filter (LPF) 6051-605 Ntx And execute the corresponding simulated radar signals 6041-604 Ntx The baseband filtering generates filtered radar signals 6061-606 for the corresponding transmission paths. Ntx .

[0169] Each of the multiple launch paths also includes the corresponding LPF 6051-605 Ntx Downstream mixers 6071-607 Ntx And modulates the corresponding filtered radar signal 6061-606 using, for example, a carrier sine wave generated by a local oscillator (not shown). Ntx This generates RF radar signals 6081-608 for the corresponding transmission path. Ntx Each of the multiple transmission paths also includes a power amplifier (PA) 6091-609. Ntx And amplify the corresponding RF radar signal 6081-608 Ntx This generates radar transmission signals 6101-610 for the corresponding transmission path. Ntx The signal is transmitted via the corresponding transmitting antenna 6111-611 NtxTransmission. Therefore, each of the multiple transmission paths may include at least one of a DAC, LPF, mixer, PA, and optional CGU and transmit antenna to generate and / or transmit radar signals.

[0170] according to Figure 6B The receiving side 600B includes multiple receiving or receiving paths 1-Nrx. Generally, radar transmitted signals from the transmitting side 600A are reflected from all targets, and the resulting echoes or reflected signals or radar received signals are received by the receiving side 600B. The receiving side 600B includes multiple receiving paths 1-Nrx. Each of the multiple receiving paths includes antennas 6311-631. Nrx And receive radar transmission signals 6101-610. Ntx The corresponding radar received signals 6301-630 Nrx In fact, the transmitted signal was 6101-610. Ntx Only a portion of it was reflected back to the receiving antenna 6311-631 Nrx .

[0171] Each of the multiple receiver paths also includes a low-noise amplifier (LNA) 6291-629 Nrx And amplify the corresponding radar received signal 6301-630 Nrx This generates amplified radar receiving signals 6281-628 for the corresponding receiving path. Nrx Each of the multiple receive paths also includes the corresponding LNA 6291-629. Nrx Downstream mixers 6271-627 Nrx The corresponding amplified radar received signal (6281-628) is demodulated using a carrier signal generated by a local oscillator. Nrx This generates baseband signals 6261-626 for the corresponding receiving path. Nrx .

[0172] Each of the multiple receive paths also includes the corresponding mixers 6271-627. Nrx Downstream low-pass filter (LPF) 6251-625 Nrx And for the corresponding baseband signals 6261-626 Nrx Filtering is performed to generate filtered baseband signals 6241-624 for the corresponding receiving paths. Nrx Each of the multiple receive paths also includes the corresponding LPF 6251-625. Nrx Downstream analog-to-digital converters (ADCs) 6231-623 Nrx And the corresponding filtered baseband signals 6241-624 NrxConvert to the digital domain, thereby generating reflected signals 6221-622 in the digital domain for the corresponding receiving path. Nrx .

[0173] Each of the multiple receive paths also includes the corresponding ADC 6231-623 Nrx Downstream processing units 6211-621 Nrx And according to the corresponding reflected signals 6221-622 Nrx For each receiving path, a processing input signal is generated in the digital domain to perform range and Doppler processing. Therefore, each receiving or receiving path for receiving reflected signals may include at least one of an LNA, a mixer, an LPF, an ADC, and optionally a processing unit and a receiving antenna.

[0174] It should be understood that the transmitting antenna 6111-611 Ntx and receiving antenna 6311-631 Nrx Dedicated antennas for corresponding transmit and receive paths can be implemented as antenna arrays. Transmit antenna 6111-611 is also conceivable. Ntx and receiving antenna 6311-631 Nrx It can be implemented in a single antenna array and the antenna array can be operated in an interchangeable manner, for example by means of a circulator, for corresponding transmit and receive paths.

[0175] CGU 6011-601 of the 600A transmitter side Ntx It can be used with the corresponding processing units 6211-621 of the receiving side 600B. Nrx Communication is used to provide information about the selected code sequence, for example, for distance processing. It can be envisioned that CGU 6011-601... Ntx Processing units 6211-621 Nrx It can be implemented as a single entity, for example, as a baseband processing unit. Furthermore, additional means for generating and / or processing radar signals, such as a memory or storage for storing code sequences, control commands, etc., and interfaces such as a user interface, are not explicitly shown, but are evident in the above implementation.

[0176] Figure 7 An exemplary receive path for the receiver side 600B is shown in detail. Typically, CGU 6011-601... Ntx Each corresponds to Figure 4 CGU 401 and the preceding ones Figure 5A The operations described. Therefore, the operations are not repeated, especially those described in CGU 6011-601. NtxA radar signal is generated. However, it should be noted that for K, there may be several possible values ​​that satisfy the criteria listed above, and each of the multiple transmission paths can generate a radar signal with a different value of K, such that the transmitted signals are rotated in different asymptotic phases relative to each other. Thus, each of the multiple reception paths should be able to separate the signal from each transmission path during coherent accumulation.

[0177] To achieve this, processing units 6211-621 along the corresponding receiving path Nrx Each component comprises multiple processing sections, preferably the number of processing sections equal to the number of transmission paths in the transmitter 600A. Each processing section includes at least one multiplier 7231-723. Ntx Then at least one processing block 7251-725 Ntx and at least one DFT block 7271-727 Ntx Multiplier 7231-723 Ntx Each of them multiplies the reflected signal (e.g., the reflected signal 6221 along the first receiving path of the receiving side 600B) in the digital domain with an asymptotic phase rotation having a K value corresponding to the K value selected at the respective transmitting path.

[0178] For example, the first transmission path of the transmitter 600A, namely, the CGU 6011 of the transmitter 600A, can select K1, and can be achieved by combining the periodic repetition of the code sequence with asymptotic phase rotation. The radar signal is generated by multiplication. In this process, the multiplier 7231 of the first processing section of the first processing unit 6211 along the first receiving path multiplies the reflected signal 6221 with an asymptotically phase-rotated multiplier. Multiply them and generate the first processing input signal 7241.

[0179] The processing block 7251 of the first processing section of the first processing unit 6211 along the first receiving path performs range processing on the first processed input signal 7241 to generate a radar range image. The DFT block 7271 of the first processing section of the first processing unit 6211 along the first receiving path further performs DFT on N samples to generate a Doppler image.

[0180] Similarly, the second transmission path of the transmitter 600A, namely the CGU 6012 of the transmitter 600A, can be K2, and can be achieved by combining the periodic repetition of the code sequence with asymptotic phase rotation. The radar signal is generated by multiplication. In this process, the multiplier 7232 of the second processing section of the second processing unit 6211 along the first receiving path multiplies the reflected signal 6221 with an asymptotically phase-rotated multiplier. Multiply them and generate the second processing input signal 7242.

[0181] The processing block 7252 of the second processing section of the first processing unit 6211 along the first receiving path performs range processing on the second processing input signal 7242 to generate a radar range image. The DFT block 7272 of the second processing section of the first processing unit 6211 along the first receiving path further performs DFT on N samples to generate a Doppler image.

[0182] For all processing sections of the first processing unit 6211 along the first receiving path, the above-described signal processing scheme is performed (preferably simultaneously). Furthermore, processing units 6211-621 along each receiving path of the receiving side 600B are also processed similarly. Nrx Each of them performs the above signal processing scheme sequentially or simultaneously.

[0183] However, the above signal processing scheme is applicable if the relationship between different values ​​of K meets the following criteria:

[0184]

[0185] Among them, K p K is an integer or non-integer for the first transmission path among multiple transmission paths. q It is an integer or non-integer for the second transmission path among the plurality of transmission paths.

[0186] In other words, if the corresponding parameter K of two different transmission paths (e.g., defined by indices p and q) among multiple transmission paths... p and K q If the conditions described in equation (24) are satisfied, then they are orthogonal.

[0187] Generally, this solution can orthogonalize up to M transmit paths or transmitters. However, in practice, the number of orthogonalizable transmit paths, Ntx, is less than M. Therefore, if fewer than Ntx transmitters are required, another MIMO mode (e.g., time-domain MIMO) can be combined, because the π / K modulation scheme already provides MIMO capability if the cumulative number M is large enough. However, if a large number of transmit paths or transmitters are required, the value of M can be increased accordingly, or the π / K modulation scheme can be combined with another MIMO mode.

[0188] In the following sections, different implementations of MIMO schemes are presented based on the types of code sequences or modulation schemes that can be used in different DMRs.

[0189] A) Zadoff code sequence

[0190] Perform the following steps to obtain a suitable K value to achieve IQ imbalance robustness, DC offset robustness and baseband HD3 robustness, while providing MIMO capability in PMCW radar.

[0191] Step 1: Select up to M integers A p , making

[0192]

[0193] in

[0194] Step 2: To benefit from MIMO capabilities, the following constraints must be met.

[0195]

[0196] Step 3: Choose parameter K such that:

[0197]

[0198] B) Binary code sequence

[0199] Perform the following steps to obtain a suitable K value to achieve IQ imbalance robustness, DC offset robustness, and power amplifier nonlinear robustness, while providing MIMO capability in PMCW radar.

[0200] Step 1: Calculate up to M integers A p , making

[0201]

[0202] Where d is an integer.

[0203] Step 2: Only keep A that meets the following conditions p value:

[0204]

[0205] in

[0206] Step 3: To benefit from MIMO capabilities, the following constraints must be met.

[0207]

[0208] Step 4: Select parameter K according to equation (27).

[0209] C) Other code sequences

[0210] Perform the following steps to obtain a suitable K value to achieve IQ imbalance robustness, DC offset robustness, and MIMO capability in all types of DMR.

[0211] Step 1: Select up to M integers A p , making

[0212]

[0213] in

[0214] Step 2: To benefit from MIMO capabilities, the following constraints must be met.

[0215]

[0216] Step 3: Select parameter K according to equation (27).

[0217] Some exemplary combinations of parameters Lc, K, and M based on the type of code sequence used are shown in the table below:

[0218]

[0219] For example, for a 511-bit binary code sequence (BPSK) of length Lc, six consecutive range profiles (i.e., M) need to be accumulated, and for each transmit path or transmitter, the value of K needs to be defined as 3 or 1.5. This specific combination allows for the generation of radar signals robust to IQ imbalance, DC offset robustness, and power amplifier nonlinearity. When using Zadoff code sequences, a 544-bit Zadoff code sequence of length Lc requires the accumulation of seven consecutive range profiles, where the value of K for each transmit path or transmitter needs to be defined as 14 or 3.0464.

[0220] It should also be noted that the number of orthogonalizable transmit paths or transmitters is always less than M. This is because it is difficult to compute M values ​​of K that simultaneously satisfy the MIMO criterion and all other criteria. Furthermore, for Zadoff code sequences, seven coherent range profiles are chosen, i.e., M = 7, for MIMO implementations with two orthogonal transmit paths or transmitters. However, in SISO implementations, several solutions with M = 3 are available.

[0221] For code sequences other than BPSK and Zadoff code sequences, at least five consecutive range images are selected for two orthogonal transmission paths or transmitters, i.e., M = 5. However, if three orthogonal transmission paths or transmitters are implemented, at least seven consecutive range images need to be accumulated, i.e., M = 7.

[0222] In other words, for code sequences, besides BPSK and Zadoff code sequences, three orthogonal transmitters with M = 7 code sequence repetitions can be implemented. However, for Zadoff code sequences, the number of orthogonal transmitters is limited to two transmitters with the same number of code repetitions. This is because the latter also benefits from HD3 robustness, which adds more constraints to the possible values ​​of K.

[0223] exist Figure 8A and Figure 8B The figure shows range profiles for M-sequence code sequences with a code sequence length of 511 for different modulation techniques. In particular, for M-sequences, the range sidelobes arise from the limitations imposed on the code sequence itself, as it is known in the art that M-sequences produce ghosted targets when subjected to amplifier nonlinearity.

[0224] Figure 8A The accumulated range image with two targets 801 and 802 is shown, specifically for an M-sequence code sequence with a code sequence length of 511, modulated using π / 2 binary phase shift keying (π / 2-BPSK) known in the art. The range image is simulated for a SISO implementation. The horizontal axis represents the distance to the targets 801 and 802 in meters, and the vertical axis represents the reflected power amplitude in decibels. Here, the two targets are indicated by circles 801 and 802 at their respective peaks. The amplitude of the range sidelobes can be approximated as approximately -55 dB, however, due to amplifier nonlinearity, ghosting targets are produced, as indicated at peak 803.

[0225] Figure 8B The cumulative range profiles of two targets 801 and 802, based on the proposed solution, are shown, particularly for an M-sequence code sequence with a code sequence length of 511. The range profiles are simulated for a SISO implementation. Here, the values ​​of K and M are chosen as 3, and binary phase-shift keying (π / 3-BPSK) modulation is used. The horizontal axis represents the distances of the targets 801 and 802 in meters, and the vertical axis represents the reflected power amplitude in decibels. Here, the two targets are indicated by circles 801 and 802 at their respective peaks. The amplitude of the range sidelobes can be approximated as approximately -50 dB, and there are no ghost targets, indicating stability against amplifier nonlinearity. Therefore, the proposed solution (e.g., π / 3-BPSK modulation technique illustrated herein) is superior in terms of ghost target suppression compared to techniques known in the art.

[0226] Figure 9The range profiles from a Zadoff code sequence with three targets 901, 902, and 903 are shown in the presence of front-end non-ideals. The range profiles are simulated for a SISO implementation. The horizontal axis represents the distances to the targets 901, 902, and 903 in meters, and the vertical axis represents the reflected power amplitude in decibels. According to the solution proposed in this disclosure, the bright lines correspond to the range profiles from conventional Zadoff code sequences, and the dark lines correspond to the range profiles from Zadoff code sequences with π / 3 modulation.

[0227] As can be seen, for the range profile of a conventional Zadoff code sequence, front-end non-ideality produces significant range sidelobes. Therefore, the third target 903 is invisible at approximately 12 meters. However, for π / K modulation, i.e., for π / 3-Zadoff according to the solution presented in this disclosure, the range sidelobes are much lower, and the third target 903 is clearly visible at approximately 12 meters. Therefore, the proposed solution (e.g., π / 3-Zadoff as illustrated herein) is superior in terms of sidelobe suppression compared to techniques known in the art.

[0228] Furthermore, the π / 3-Zadoff range image also shows a correlation peak of 904 at approximately 75 m. This is also attributed to the ghosting target due to front-end non-ideals. Unfortunately, the proposed π / K modulation implementation alone cannot remove it. However, this can be overcome by combining π / K modulation with other solutions known in the art, without much difficulty or effort.

[0229] Therefore, the solution proposed in this disclosure addresses the front-end non-idealities in digitally modulated radars such as PMCW and OFDM radars, namely IQ imbalance, power amplifier nonlinearity, receiver DC offset, and especially third-order harmonic distortion in the baseband, while providing MIMO capability. Although the solution proposed in this disclosure is described for continuous transmission (CW) radars such as PMCW and OFDM radars, the proposed method and underlying technology can also be used for pulse radars with intra-pulse modulation.

[0230] Generally, in these radars, a near-ideal ambiguity function can be achieved through careful selection of waveform sequences and algorithms used to construct the radar data cube. However, transceiver non-ideals (such as nonlinearity, phase noise, IQ imbalance, etc.) can degrade the ambiguity function, manifesting as range sidelobes and / or ghosted targets. Range sidelobe degradation leads to reduced sensitivity, while ghosted targets cause false alarms; therefore, both must be avoided as much as possible.

[0231] It is important to note that in the specification and claims, the word "comprising" does not exclude other elements or steps, and the indefinite articles "a" or "an" do not exclude a plural. A single element or other unit may perform the function of several entities or items recited in the claims. Furthermore, disclosure regarding any aspect is also related to other aspects of that disclosure.

[0232] It should be noted that this application claims the benefit of European Patent Application No. 20208013.1, filed on November 17, 2020, the subject matter of which is incorporated herein by reference in its entirety, particularly with respect to the mathematical model used in this application.

[0233] Although the invention has been described and illustrated with respect to one or more embodiments, equivalent changes and modifications will occur to those skilled in the art upon reading and understanding this specification and the accompanying drawings. Furthermore, while a particular feature of the invention may be disclosed only with respect to one of several implementations, such feature may be combined with one or more other features of other implementations, as may be necessary and advantageous for any given or particular application.

Claims

1. A method (100) for promoting robust radar detection in digitally modulated radar, comprising the following steps: In digital modulation radar, the number of periodic repetitions of the code sequence, including the code sequence length Lc, is multiplied by an asymptotic phase rotation. To generate (104) radar signals along at least one transmission path in the digital domain, where Lc and M are integers, K is an integer or non-integer, and n is a discrete-time index corresponding to the code rate; as well as In digitally modulated radar, based on the digitized reflected signal corresponding to the radar signal, the digitized reflected signal is multiplied by an asymptotic phase rotation after baseband conversion. In the digital domain, a processed input signal (105) is generated along at least one receiving path, wherein for two asymptotic phase rotations at a given time, the discrete-time index n remains the same. Where K is defined as such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer, and / or In the case where the code sequence is a Zadoff code sequence, K is defined such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer.

2. The method according to claim 1, Its features are, The method further includes the following step: defining K such that the ratio It is a non-integer and M is defined such that the ratio It is an integer.

3. The method according to claim 1 or 2, characterized in that, In the case where the code sequence corresponds to a binary code sequence, the method further includes the step of defining K as an integer or non-integer given by the following formula: Where d is an integer.

4. The method according to claim 1 or 2, characterized in that, The method further includes the following steps: performing digital processing on the processed input signal in relation to the code sequence to generate a series of distance profiles.

5. The method according to claim 4, Its features are, The method further includes the step of accumulating M consecutive distance images.

6. The method according to claim 1 or 2, characterized in that, The method further includes the step of generating multiple radar signals along multiple transmission paths in the digital domain, each of the radar signals comprising the same code sequence length, the same number of periodic repetitions of the code sequence, and non-identical asymptotic phase rotations relative to the multiple radar signals, wherein each of the transmission paths is associated with a dedicated radar signal of the multiple radar signals. The orthogonality between the radar signals of each of the two transmission paths in the corresponding plurality of transmission paths is satisfied by the following: Where K p It is an integer or non-integer for the first transmission path, and K q It is an integer or non-integer for the second launch path.

7. The method according to claim 6, Its features are, The method further includes the step of generating multiple processing input signals in the digital domain along multiple receiving paths, each of the receiving paths being associated with a process input signal generated based on a digitized reflected signal corresponding to a specific dedicated radar signal in the multiple transmitting paths.

8. A digitally modulated radar (400, 600A, 600B) for promoting robust radar detection, comprising: At least one code generation unit (401) is configured to multiply the number of periodic repetitions of the code sequence, including the code sequence length Lc, by an asymptotic phase rotation. To generate radar signals along at least one transmission path in the digital domain (402), where Lc and M are integers, K is an integer or non-integer, and n is a discrete-time index corresponding to the code rate; as well as At least one processing unit (421) is configured to multiply the digitized reflected signal (422) corresponding to the radar signal (402) by an asymptotic phase rotation after baseband conversion. The digitized reflected signal (422) is used to generate a processing input signal (524) in the digital domain along at least one receiving path, wherein for two asymptotic phase rotations at a given time, the discrete-time index n remains the same. The at least one code generation unit (401) is further configured to define K such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer, or In the case where the code sequence is a Zadoff code sequence, the at least one code generation unit (401) is further configured to define K such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer.

9. The digital modulation radar according to claim 8, Its features are, The at least one code generation unit (401) is also configured to define K such that the ratio It is a non-integer, and M is defined such that the ratio It is an integer.

10. The digital modulation radar according to claim 8 or 9, Its features are, In the case where the code sequence corresponds to a binary code sequence, the at least one code generation unit (401) is further configured to define K as an integer or non-integer given by the following formula: Where d is an integer.

11. The digital modulation radar according to claim 8 or 9, Its features are, The at least one processing unit (421) is also configured to perform digital processing on the processing input signal (524) relative to the code sequence to generate a series of range images.

12. The digital modulation radar according to claim 11, Its features are, The at least one processing unit (421) is also configured to accumulate M consecutive distance images.

13. The digital modulation radar according to claim 8 or 9, Its features are, The at least one code generation unit (401) is configured to define Lc with respect to the code rate relative to the maximum unambiguous range of the radar.

14. The digital modulation radar according to claim 8 or 9, Its features are, The digital modulation radar also includes multiple transmission paths (1-Ntx), each of which includes at least one code generation unit (6011-601). Ntx ),as well as For each of the transmission paths, the corresponding code generation unit (6011-601) Ntx ) is configured to generate dedicated radar signals (6021-602) in the digital domain. Ntx The dedicated radar signals have the same code sequence length, the same number of periodic repetitions of the code sequence, and non-identical asymptotic phase rotations relative to the plurality of transmission paths, and The orthogonality between the radar signals of each of the two transmission paths in the corresponding plurality of transmission paths is satisfied by the following: Where K p It is an integer or non-integer for the first transmission path, and K q It is an integer or non-integer for the second launch path.

15. The digital modulation radar according to claim 14, Its features are, The digital modulation radar also includes multiple receiving paths (1-Nrx), each of which includes at least one processing unit (6211-621). Nrx ),as well as For each of the receiving paths, the corresponding processing unit (6211-621) Nrx ) is configured to respond to the corresponding dedicated radar signals (6021-602) of the plurality of transmission paths (1-Ntx). Ntx The corresponding digitally reflected signal (6221-622) Nrx Generate and process input signals in the digital domain (7241-724) Ntx ).

Images