Bistatic radar signal reflection detector

Abstract

Provided is a bistatic radar signal reflection detector. A method includes receiving a set of radar signals incident on a set of objects. The method includes selecting a GLRT detector. The method includes determining a ratio between a maximized likelihood function of the second hypothesis model and a maximized likelihood function of the first hypothesis model. The method includes determining whether a set of angles associated with the set of radar signals are available. The method includes, in response to determining that the set of angles are available, updating a set of data associated with a set of objects. The method includes, in response to determining that the set of angles is unavailable, selecting an angle estimation method, estimating the set of angles, and tracking the set of objects.

Description

Technical Field

[0001] This disclosure relates to radar signal processing, and more particularly to detecting bistatic reflections and estimating angles based on the processing of received radar signals (US Patent Classification 342). Background Technology

[0002] In automotive applications, radar is frequently used to detect obstacles, such as other vehicles or other hazards. In some scenarios and environments, the transmitted radar signal returns directly to the local radar after hitting the first object (direct-path signal), while in others, the signal returns to the local radar after reflecting off the first object and then the second (multipath signal). Because the environment illuminated by automotive radar is often congested, multipath reflections can be predominant in some cases, with the most challenging type being known as "bistatic" reflections, which affect how angles are estimated after range-Doppler processing. Traditional angle estimation methods assume direct-path reflections, leading to large angle estimation errors when the model mismatches. Special angle estimators may be required when multipath reflections occur. Bistatic reflections also affect how target tracking systems handle range-velocity-angle detection. Tracking algorithms typically assume direct-path reflections, and detections with bistatic reflections are often discarded. Therefore, determining whether range-Doppler detection includes multipath energy becomes crucial for the entire system.

[0003] By definition, for bistatic signals, the direction of arrival (DOA) is not equal to the direction of transmission (DOD), therefore bistatic reflections must be detected in the spatial (or angular) domain after range-Doppler processing. Some bistatic detectors apply linear prediction (LP) theory to synthesized uniform linear arrays (ULAs). When there are no bistatic reflections, the LP error is close to the noise power, while when bistatic reflections are present, the LP error becomes larger. Comparing the LP error to a threshold can indicate whether the signal is multipath-dependent. Other methods use angular unrolling schemes (e.g., direct matching and cross-matching) to detect bistatic reflections by testing the angular unrolling matching error between the DOA and DOD. When using such schemes, bistatic reflections are present if the direct matching error is large and the cross-matching error is smaller than the direct matching error. A third method uses a joint DOD-DOA estimation method. Multipath detection is performed for each reflection path by comparing the angular unrolling matching errors between the associated DOD and DOA.

[0004] The background description provided herein is intended to present the general context of this disclosure. The work of the currently named inventors described in this background section, and aspects of the description that may not constitute prior art at the time of filing, are neither explicitly stated nor implied to be prior art to this disclosure. Summary of the Invention

[0005] A method includes receiving a set of radar signals incident on a set of objects. The method includes determining a first hypothetical model representing a monostatic signaling model of the set of radar signals. The method includes determining a second hypothetical model representing a bistatic signaling model of the set of radar signals. The method includes selecting a selected GLRT detector from a set of generalized likelihood ratio test (GLRT) detectors based on a set of signal criteria. The method includes determining, based on the selected GLRT detector, a ratio between the maximum likelihood function of the second hypothetical model and the maximum likelihood function of the first hypothetical model. The method includes determining whether the ratio is greater than a threshold. The method includes determining whether a set of angles associated with the set of radar signals is available. The method includes updating a set of data associated with the set of objects in response to determining that the set of angles is available. The method includes selecting an angle estimation method based on the ratio in response to determining that the set of angles is unavailable. The method includes estimating the set of angles using the selected angle estimation method. The method includes tracking the set of objects based on the set of angles.

[0006] Among other features, the method includes determining that a second hypothetical model is accurate in response to determining that the ratio is greater than a threshold. Among other features, the method includes determining that a first hypothetical model is accurate in response to determining that the ratio is less than or equal to a threshold. Among other features, the method includes autonomously controlling the vehicle to avoid the group of objects.

[0007] Among other features, selecting an angle estimation method includes selecting a first angle estimation method in response to determining that the ratio is greater than a threshold. Among other features, selecting an angle estimation method includes selecting a second angle estimation method in response to determining that the ratio is less than a threshold.

[0008] Among other features, this set of GLRT detectors is derived by the following formula: .

[0009] Among other features, this group of GLRT detectors includes a first GLRT detector, which is defined as: .

[0010] Among other features, this group of GLRT detectors includes a second GLRT detector, which is defined as: .

[0011] Among other features, this group of GLRT detectors includes a third GLRT detector, which is defined as: .

[0012] Among other features, this group of GLRT detectors includes a fourth GLRT detector, which is defined as: .

[0013] Among the other features, the first hypothesis model Defined as: Among other features, the second hypothesis model Defined as: Among the other features, It is a set of arrayed observations. Among the other features, It is the first spatial matrix of the reflection path. Among the other features, It is the second spatial matrix of the reflection path. Among the other features, It is a set of transmitted signals. Among other characteristics, This is a set of noisy data. Among other features, θ is a set of angle data containing K elements. Among other features, N is the number of radar elements. Among other features, K is the number of reflection paths. Among other features, M is the number of observations. Among other features, η is the power level associated with this set of noisy data. Among other features, γ is the threshold. Among other features, p1 is the likelihood function under the second hypothesis model. Among other features, p0 is the likelihood function under the first hypothesis model. Among other features, This represents the Frobenius norm. Among the other features, P... A It is the first projection matrix that maps the vector to the projection onto the subspace formed by A, where Among the other features, P B It is the second projection matrix that maps the vector to the projection onto the subspace formed by B, where Among the other features, Defined as Among the other features, Defined as , where I is the identity matrix.

[0014] Among other features, the method includes determining the probability that the second hypothetical model is correct. This probability is defined as:

[0015] .

[0016] Among other features, the method includes determining the probability that the first hypothesized model is correct. This probability is defined as:

[0017] .

[0018] Among other features, the first hypothetical model is based on the number of elements in the radar array receiving the set of radar signals. Among other features, the first hypothetical model is based on a first number of reflection paths. Among other features, the first hypothetical model is based on a first number of observations. Among other features, the first hypothetical model is based on the set of array observations, a first spatial matrix of the reflection path, the set of transmitted signals, and the set of noise data.

[0019] Among other features, the second hypothesis model is based on the number of elements in the radar array receiving the set of radar signals. Among other features, the second hypothesis model is based on a second number of reflection paths. Among other features, the second hypothesis model is based on a second number of observations. Among other features, the second hypothesis model is based on the set of array observations. Among other features, the second hypothesis model is based on a second spatial matrix of the reflection path. Among other features, the second hypothesis model is based on the set of transmitted signals. Among other features, the second hypothesis model is based on the set of noise data.

[0020] Among other characteristics, the set of signal criteria includes a first criterion that is met when the power level associated with a set of noise data is known; and a second criterion that is met when a set of angles associated with the spatial matrix of the reflection path is known. Among other characteristics, the set of angles includes the transmission direction and the arrival direction.

[0021] Among other features, the method includes estimating the set of angles before determining whether the ratio is greater than a threshold. Among other features, the method includes identifying a set of objects associated with bistatic reflection based on the ratio.

[0022] Among other features, the selected angle estimation method is applicable to both direct path reflection and multipath reflection, and the selected GLRT detector is either a third GLRT detector or a fourth GLRT detector.

[0023] A system includes memory hardware configured to store instructions and processor hardware configured to execute the instructions. The instructions include receiving a set of radar signals incident on a set of objects. The instructions include determining a first hypothetical model representing a monostatic signal model of the set of radar signals. The instructions include determining a second hypothetical model representing a bistatic signal model of the set of radar signals. The instructions include selecting a selected GLRT detector from a set of generalized likelihood ratio test (GLRT) detectors based on a set of signal criteria. The instructions include determining, based on the selected GLRT detector, a ratio between the maximum likelihood function of the second hypothetical model and the maximum likelihood function of the first hypothetical model. The instructions include determining whether the ratio is greater than a threshold. The instructions include determining whether a set of angles associated with the set of radar signals is available. The instructions include updating a set of data associated with the set of objects in response to determining that the set of angles is available. The instructions include selecting an angle estimation method based on the ratio in response to determining that the set of angles is unavailable. The instructions include estimating the set of angles using the selected angle estimation method. The instructions include tracking the set of objects based on the set of angles.

[0024] Among other features, the instruction includes determining that the second hypothetical model is accurate in response to determining that the ratio is greater than a threshold. Among other features, the instruction includes determining that the first hypothetical model is accurate in response to determining that the ratio is less than or equal to a threshold.

[0025] Among other features, selecting an angle estimation method includes selecting a first angle estimation method in response to determining that the ratio is greater than a threshold. Among other features, selecting an angle estimation method includes selecting a second angle estimation method in response to determining that the ratio is less than a threshold.

[0026] Among other features, this set of GLRT detectors is derived by the following formula: .

[0027] Among other features, this group of GLRT detectors includes a first GLRT detector, which is defined as: .

[0028] Among other features, this group of GLRT detectors includes a second GLRT detector, which is defined as: .

[0029] Among other features, this group of GLRT detectors includes a third GLRT detector, which is defined as: .

[0030] Among other features, this group of GLRT detectors includes a fourth GLRT detector, which is defined as: .

[0031] Among the other features, the first hypothesis model Defined as: Among other features, the second hypothesis model Defined as: Among the other features, It is a set of arrayed observations. Among the other features, It is the first spatial matrix of the reflection path. Among the other features, It is the second spatial matrix of the reflection path. Among the other features, It is a set of transmitted signals. Among other characteristics, This is a set of noisy data. Among other features, θ is a set of angle data containing K elements. Among other features, N is the number of radar elements. Among other features, K is the number of reflection paths. Among other features, M is the number of observations. Among other features, η is the power level associated with this set of noisy data. Among other features, γ is the threshold. Among other features, p1 is the likelihood function under the second hypothesis model. Among other features, p0 is the likelihood function under the first hypothesis model. Among other features, This represents the Frobenius norm. Among the other features, P... A It is the first projection matrix that maps the vector to the projection onto the subspace formed by A, where Among the other features, P B It is the second projection matrix that maps the vector to the projection onto the subspace formed by B, where Among the other features, Defined as Among the other features, Defined as , where I is the identity matrix.

[0032] Further areas of applicability of this disclosure will become apparent from the detailed description, claims, and drawings. The detailed description and specific examples are intended for illustrative purposes only and are not intended to limit the scope of this disclosure. Attached Figure Description

[0033] This disclosure will be more fully understood from the detailed description and accompanying drawings.

[0034] Figure 1 This is a functional block diagram of an example radar system.

[0035] Figure 2 These are examples of monostatic and bistatic signals in the example scenario.

[0036] Figure 3A This is a block diagram illustrating an example of a monostatic direct reflection path.

[0037] Figures 3B to 3C This is a block diagram illustrating an example bistatic reflection path.

[0038] Figure 3D This is a block diagram of an example monostatic multipath reflection path.

[0039] Figure 4 This is a flowchart of an example method for determining whether a signal is monostatic or bistatic.

[0040] Figure 5 This is a flowchart of an example method for selecting a detector using the generalized likelihood ratio test (GLRT).

[0041] Figure 6 This is a flowchart of an example method for determining whether a signal is monostatic or bistatic after angle estimation.

[0042] In the accompanying drawings, reference numerals may be used repeatedly to identify similar and / or identical elements. Detailed Implementation

[0043] introduction

[0044] This disclosure provides a method for determining whether a radar signal reflection is monostatic or bistatic. In the proposed method, the ratio between the maximum likelihood of the bistatic hypothesis and the maximum likelihood of the monostatic hypothesis is found using the generalized likelihood ratio test (GLRT) framework. The proposed detector does not require special array design and can theoretically handle any combination of direct-path and multipath reflections within the same range-Doppler cell. Due to bistatic reflections with two reciprocal paths (such as...), Figure 2 Signals 204 and 208 (shown in the diagram) represent the most common case, and therefore the GLRT detector is derived based on this model. The GLRT detector can also be extended to other multipath models.

[0045] Example radar system

[0046] Figure 1 This is a high-level block diagram of radar system 108. Radar system 108 may be mounted to and / or integrated within vehicle 112. Radar system 108 is configured to detect one or more objects in the vicinity of vehicle 112. In various embodiments, radar system 108 may be a forward-looking radar system.

[0047] In various embodiments, radar system 108 may be mounted on the top, bottom, front, rear, left, or right side of vehicle 112. In various embodiments, radar system 108 includes multiple radar subsystems. For example, radar system 108 may include a first front-mounted radar subsystem positioned near the left side of vehicle 112 and a second front-mounted radar subsystem positioned near the right side of vehicle 112. In various embodiments, the location of radar system 108 may be selected to provide a specific field of view that covers an area of ​​interest where one or more objects may be present. For example, the field of view may include a 360-degree field of view, one or more 180-degree fields of view, and / or one or more 90-degree fields of view.

[0048] In various implementations, vehicle 112 may include one or more systems that use data provided by radar system 108. For example, vehicle 112 may include a driver assistance system and / or an autonomous driving system. The driver assistance system may use data provided by radar system 108 to monitor one or more blind spots of vehicle 112 and / or alert the driver of vehicle 112 to potential collision risks with objects. The autonomous driving system may use data provided by radar system 108 to drive vehicle 112, avoid collisions with objects, perform emergency braking, change lanes, and / or adjust the speed of vehicle 112, etc.

[0049] In various embodiments, radar system 108 may include at least one antenna array 136 and at least one transceiver 140. In various embodiments, radar system 108 may include processor hardware 144 and memory hardware 148. Memory hardware 148 may include radar software 152. In various embodiments, radar software 152 may be configured to analyze radar signals, detect one or more objects, and / or determine one or more characteristics of the objects (e.g., position and / or velocity). In various embodiments, the radar software is implemented entirely or partially in hardware.

[0050] Monostatic and bistatic reflection

[0051] Figure 2 Examples of monostatic and bistatic reflections are provided. Vehicle 112 includes radar system 108. Radar system 108 generates a monostatic signal 212 that is reflected directly from target vehicle 220 to vehicle 112. Under different conditions, the signal may not be reflected directly back to vehicle 112. As an example, a multipath (in this case, bistatic) signal 204 is reflected from reflective surface 224 to target vehicle 220 and then back to vehicle 112. As another example, signal 208 is reflected from target vehicle 220 to reflective surface 224 and then back to vehicle 112.

[0052] Figure 3A This is an example of monostatic direct path reflection. Figure 3A In this context, DOD equals DOA. Signal R1 leaves the local radar 304, hits target 312a, and is reflected directly back to the local radar 304 without being reflected from the reflective surface 308. Figure 3B This is an example of an asymmetric bistatic reflection path (DOD is not equal to DOA). Signal R1 leaves the local radar 304, hits target 312b and is guided to reflector surface 308, and then returns to the local radar 304. Figure 3C This is the second example of an asymmetric bistatic reflection path (DOD is not equal to DOA). Signal R3 leaves the local radar 304, hits the reflector surface 308 and is guided to the target 312c, and then returns to the local radar 304. Figure 3D This is an example of monostatic multipath reflection (DOD equals DOA, but the reflection is not direct). The radar signal is reflected from reflector 308 towards target 312d. Then the signal is reflected from target 312d back to reflector 308 and then back to the local radar 304.

[0053] First GLRT detector

[0054] The signal model of bistatic reflection can be written as: ,in These are array observations. It is the spatial matrix of the reflection path with unknown θ. It is an unknown transmission signal. It is spatially and temporally whitened zero-mean Gaussian noise with unknown power η.

[0055] The number of elements in a Multiple-Input Multiple-Output (MIMO) composite array is denoted by N, the number of observation snapshots by M, and the number of reflection paths by K (when only direct path reflections exist, the number of targets equals the number of reflection paths). As an example, K = 2 and .like Figure 2 As shown, signal 204 has a DOD of θ1 and a DOA of θ2, while signal 208 has a DOD of θ2 and a DOA of θ1. The space matrix is ​​a subspace spanned by the guidance vectors of two reciprocal bibase paths: The steering vector of signal 204 The steering vector of signal 208 and These represent the transmit and receive subarray manifolds, respectively, and This represents the Kronecker product operation.

[0056] Based on the above symbols, the corresponding part of the signal model with direct path reflection can be written as follows: ,in Therefore, a binary hypothesis test can be formulated as follows:

[0057]

[0058] GLRT is then given by the following formula

[0059]

[0060] in and They are and The likelihood function is given by γ, where γ is the threshold. Specifically,

[0061] ,as well as

[0062] .

[0063] In some implementations, the unknown angle and / or unknown noise power are pre-estimated so that they can be excluded from the list of unknown variables. The maximum likelihood estimation (MLE) results for the unknowns are expressed as... The test statistic can be further written as

[0064]

[0065] in

[0066] ,as well as

[0067] .

[0068] In the above equation, H represents the Hermitian operator. Let Frobenius norm be represented. The two MLE problems mentioned above regarding unknown angles can be solved using nonlinear least squares (NLS). Taking the first derivative of the likelihood function with respect to η and setting the function to zero, we obtain... Down ,exist Down Based on the test statistic, the first GLRT detector is given by the following formula:

[0069] .

[0070] Therefore, the first GLRT detector can be interpreted as comparing the estimated noise power under two assumptions. When in When the estimated noise power is much larger, It may be true. If noise power estimates represented by Δ0 and Δ1 are used (e.g., or The results will be more accurate. If the minimum residual projected onto the direct path subspace is much larger than the minimum residual projected onto the bistatic path subspace, it means that the bistatic reflection model can better fit the observations, and It may be true.

[0071] Second GLRT detector

[0072] In some cases, the noise power η may be known a priori through certain estimation processes in the range-Doppler processing stage, which leads to the second GLRT detector:

[0073] .

[0074] Without strictly pursuing sign accuracy, the second GLRT detector can be simplified to...

[0075]

[0076] in

[0077]

[0078] and

[0079] .

[0080] The second GLRT detector can also be interpreted as a test of model fit. If more energy than the direct path subspace can be projected from the observations onto the bistatic path subspace, then the bistatic reflection model... It can better fit the observed values.

[0081] Third GLRT detector

[0082] In some cases, the angle θ can also be known a priori without needing to predetermine the reflection model. For example, in some implementations, a general algorithm is used to estimate the angle by leveraging the fact that the DOD set is the same as the DOA set, without distinguishing the reflection model. Therefore, the third GLRT detector can be written as

[0083] .

[0084] Similarly, without strictly pursuing sign accuracy, the third GLRT detector can be simplified to

[0085] .

[0086] Compared to the second GLRT detector, the third GLRT detector does not involve angle estimation for each model. The third GLRT detector also examines model fit with known angles.

[0087] Fourth GLRT detector

[0088] Another possible scenario is that the angle is known but the noise power is unknown (e.g., when the estimated noise power is unreliable). The GLRT detector is given by the following equation.

[0089] .

[0090] By representing the MLE result of the unknown as The test statistic can be written as

[0091]

[0092] in and Similar to the first detector, this detector can be interpreted as comparing the estimated noise power under two assumptions. When in When the estimated noise power is much larger, This is likely true. Based on noise power estimation, the fourth GLRT detector is...

[0093] .

[0094] Other considerations

[0095] Although the expressions for the aforementioned GLRT detectors differ, they can be unified to finding a better model to fit the observations. For example, a second GLRT detector can be correlated with the first GLRT detector using a monotonically increasing function, such as... .

[0096] Based on the insights gained above, other multipath reflection models can be extended by changing the spatial matrix involved. For example, to distinguish between a hybrid reflection model with one direct path reflection and one bistatic reflection and a direct path reflection model, one can use... replace When the comparison models are not mutually exclusive, the threshold used for decision-making should be adjusted.

[0097] The detector mentioned earlier is a binary detector (in other words, the decision is "true" or "false"). Alternative methods can provide probabilities instead of simple binary analysis. Soft decision methods based on GLRT detectors can be used to provide probabilistic features. The probability of events where Model B is correct is given by the following formula.

[0098]

[0099] The probability of an event that model A is correct is given by the following formula.

[0100] .

[0101] When the angle is known or estimated in advance, the maximization step in the above equation can be removed.

[0102] Angle estimation framework

[0103] Sometimes angle (or phase) estimation and bistatic reflection detection are not clearly decoupled. For example, in some methods, angle (or phase) estimation must be performed on all possibilities before determining whether multipath reflection exists. Once determined, the corresponding estimate is selected as the final result. In other words, estimation is a prerequisite for bistatic detection, but the estimation is not finalized until bistatic detection is complete. The first and second GLRT detectors belong to this category.

[0104] Some methods separate bistatic reflection detection from angle or phase estimation. Detection can be performed without estimating any angle or phase by utilizing the linear prediction (LP) characteristics of a uniform linear array (ULA) formed by a MIMO radar system. Based on the detection results, an appropriate estimator is selected to find the angle. If the angle is first estimated using methods applicable to both direct-path and multi-path reflections, then a third or fourth GLRT detector is used to identify the detection with bistatic reflections.

[0105] flow chart

[0106] Figure 4 This is a flowchart of an example method for determining whether a signal reflection is monostatic or bistatic. Control begins at 404, after receiving the input radar signal. At 408, control selects a generalized likelihood ratio test (GLRT) detector (e.g., using a reference detector). Figure 5 (Methods described). In 412, the first assumption of control maximization. The likelihood function is given. At 416, the control maximizes the second hypothesis. The likelihood function is calculated. At 420, the control uses a selected GLRT detector to generate a ratio between the first and second hypotheses. At 424, the control determines whether this ratio is greater than a specified threshold. In some implementations, the threshold is determined by simulation to produce the most accurate results. If the ratio is greater than the threshold, the control moves to 428 and determines the second hypothesis. If true, then if the ratio is not greater than the threshold, control shifts to 432 and the first hypothesis is determined. True. Control ends after 428 or 432.

[0107] In some implementations, after 428 or 432, control performs additional actions instead of ending. For example, if a reference is used... Figure 4 The described method determines monostatic or bistatic reflections before angle estimation occurs, and then controls the selection of the angle estimation algorithm based on whether monostatic or bistatic determination is used. As another example, if a reference is used... Figure 4 The described method, after angle estimation occurs, controls the use of monostatic or bistatic determinations to adjust the angle data (e.g., by ignoring phantom targets or adjusting the distance to the target). In some implementations, the control sends the monostatic or bistatic determinations to another system or module to adjust the angle and / or target data.

[0108] Figure 5 This is a flowchart of an example method for selecting a generalized likelihood ratio test (GLRT) detector. Control begins at 504 and determines whether the noise power level is known. If the noise power is known, control moves to 508. If the noise power is unknown, control moves to 512. At 508, control determines whether the angle (e.g., DOA or DOD) is known. If the angle is known, control moves to 516 and selects a third GLRT detector. If the angle is unknown, control moves to 520 and selects a second GLRT detector.

[0109] At 512, control determines whether the angle (e.g., DOA or DOD) is known. If the angle is known, control moves to 524 and selects the fourth GLRT detector. If the angle is unknown, control moves to 528 and selects the first GLRT detector.

[0110] Figure 6This is a flowchart of an example method for determining the angle and whether a signal is bistatic or monostatic. Control begins at 604 and determines the angle (e.g., by a method applicable to direct path and multipath reflections). At 608, control selects a GLRT detector (e.g., a third or fourth GLRT detector). At 612, control determines whether monostatic or bistatic reflection is present. At 616, control sends the monostatic / bistatic determination and control terminates.

[0111] Terms and Conditions

[0112] Various exemplary embodiments of the present invention are described in the following clauses.

[0113] Clause 1: A method comprising:

[0114] Receive a set of radar signals incident on a set of objects;

[0115] Determine the first hypothetical model representing the monostatic signal model of this group of radar signals;

[0116] Determine a second hypothetical model representing the bistatic signal model of this group of radar signals;

[0117] Based on a set of signal criteria, select a GLRT detector from a set of generalized likelihood ratio test (GLRT) detectors;

[0118] Based on the selected GLRT detector, the ratio between the maximum likelihood function of the second hypothesis model and the maximum likelihood function of the first hypothesis model is determined;

[0119] Determine if the ratio is greater than the threshold;

[0120] Determine if a set of angles associated with this set of radar signals is available;

[0121] In response to determining that the set of angles is available, update a set of data associated with a set of objects; and

[0122] In response to determining that this set of angles is unavailable:

[0123] Based on this ratio, an angle estimation method is selected;

[0124] Estimate the set of angles using the selected angle estimation method; and

[0125] Track this group of objects based on this set of angles.

[0126] Clause 2: The method described in Clause 1 further includes:

[0127] In response to determining that the ratio is greater than a threshold, it is determined that the second hypothesis model is accurate; and

[0128] In response to determining that the ratio is less than or equal to the threshold, it is determined that the first hypothesis model is accurate.

[0129] Clause 3: The method according to any one of Clauses 1 to 2 further includes autonomously controlling the vehicle to avoid the group of objects.

[0130] Clause 4: The method according to any one of Clauses 1 to 3, wherein selecting the angle estimation method includes:

[0131] In response to determining that the ratio is greater than a threshold, the first angle estimation method is selected; and

[0132] In response to the determination that the ratio is less than the threshold, the second angle estimation method is selected.

[0133] Clause 5: The method according to any one of Clauses 1 to 4, wherein the set of GLRT detectors is derived by the following formula:

[0134] .

[0135] Clause 6: The method according to any one of Clauses 1 to 5, wherein the set of GLRT detectors comprises:

[0136] The first GLRT detector is defined as follows: ,

[0137] The second GLRT detector is defined as follows: ,

[0138] The third GLRT detector is defined as follows: ,as well as

[0139] The fourth GLRT detector is defined as follows: .

[0140] Clause 7: The method described in accordance with Clause 6, wherein:

[0141] The selected angle estimation method is applicable to both direct path reflection and multipath reflection, and

[0142] The selected GLRT detector is either the third GLRT detector or the fourth GLRT detector.

[0143] Clause 8: The method described in Clause 6, wherein:

[0144] First hypothesis model Defined as: ,

[0145] Second Hypothesis Model Defined as:

[0146] It is a set of array observations.

[0147] It is the first spatial matrix of the reflection path.

[0148] It is the second space matrix of the reflection path.

[0149] It is a set of transmitted signals.

[0150] It is a set of noisy data.

[0151] θ is a set of angle data containing K elements.

[0152] N is the number of radar elements.

[0153] K is the number of reflection paths.

[0154] M is the number of observations.

[0155] η is the power level associated with this set of noise data.

[0156] γ is the threshold.

[0157] p1 is the likelihood function under the second hypothesis model.

[0158] p0 is the likelihood function under the first hypothesis model.

[0159] Describing the Frobenius norm,

[0160] P A It is the first projection matrix that maps the vector to the projection onto the subspace formed by A, where, ,

[0161] P B It is the second projection matrix that maps the vector to the projection onto the subspace formed by B, where, ,

[0162] Defined as ,and

[0163] Defined as , where I is the identity matrix.

[0164] Clause 9: The method according to any one of Clauses 1 to 8 further comprises:

[0165] Determine the probability that the second hypothesis model is correct.

[0166] The probability is defined as follows:

[0167] .

[0168] Clause 10: The method according to any one of Clauses 1 to 9 further comprises:

[0169] Determine the probability that the first hypothesis model is correct.

[0170] The probability is defined as follows:

[0171] .

[0172] Clause 11: The method according to any one of Clauses 1 to 10, wherein:

[0173] The first hypothesis model is based on:

[0174] The number of elements in the radar array that receive this group of radar signals.

[0175] The first number of reflection paths,

[0176] The first number of observations,

[0177] This set of array observations,

[0178] The first spatial matrix of the reflection path,

[0179] The group of transmitted signals, and

[0180] This set of noise data; and

[0181] The second hypothesis model is based on:

[0182] The number of elements in the radar array that receive this group of radar signals.

[0183] The second number of reflection paths,

[0184] The second number of observations,

[0185] This set of array observations,

[0186] The second spatial matrix of the reflection path,

[0187] The group of transmitted signals, and

[0188] This set of noise data.

[0189] Clause 12: The method according to any one of Clauses 1 to 11, wherein the set of signal standards includes:

[0190] The first criterion is met when the power level associated with a set of noise data is known, and

[0191] The second criterion is satisfied when a set of angles associated with the spatial matrix of the reflection path is known.

[0192] Clause 13: The method according to any one of Clauses 1 to 12, wherein the set of angles includes the launch direction and the arrival direction.

[0193] Clause 14: The method according to any one of Clauses 1 to 13 further comprises:

[0194] Estimate the set of angles before determining whether the ratio is greater than the threshold; and

[0195] Based on this ratio, a set of objects associated with bistatic reflection is identified.

[0196] Clause 15: A system comprising:

[0197] Memory hardware configured to store instructions;

[0198] Processor hardware configured to execute the instruction, wherein the instruction performs the method described in any one of clauses 1 to 14.

[0199] in conclusion

[0200] The term non-transitory computer-readable medium does not include transient electrical or electromagnetic signals propagated through the medium (e.g., on a carrier wave). Non-limiting examples of non-transitory computer-readable media are non-volatile memory circuits (e.g., flash memory circuits, erasable programmable read-only memory circuits, or mask read-only memory circuits), volatile memory circuits (e.g., static random access memory circuits or dynamic random access memory circuits), magnetic storage media (e.g., analog or digital magnetic tape or hard disk drives), and optical storage media (e.g., CDs, DVDs, or Blu-ray discs).

[0201] The term "group" generally refers to a grouping of one or more elements. Elements in a group do not necessarily need to share any common characteristics or otherwise belong together. The phrase "at least one of A, B, and C" should be interpreted as representing the logic (A or B or C) using non-exclusive logical OR, and should not be interpreted as representing "at least one A, at least one B, and at least one C". The phrase "at least one of A, B, or C" should be interpreted as representing the logic (A or B or C) using non-exclusive logical OR. The foregoing description is merely illustrative in nature and is in no way intended to limit this disclosure, its application, or its uses.

[0202] The broad teachings of this disclosure can be implemented in various forms. Therefore, while this disclosure includes specific examples, its true scope should not be so limited, as other modifications will become apparent upon examination of the drawings, specification, and appended claims. In the written description and claims, one or more steps within the method may be performed in a different order (or simultaneously) without altering the principles of this disclosure. Similarly, one or more instructions stored in a non-transitory computer-readable medium may be performed in a different order (or simultaneously) without altering the principles of this disclosure. Unless otherwise stated, the numbering or other designation of instructions or method steps is for convenience of reference and does not indicate a fixed order.

[0203] Furthermore, although each of the above embodiments is described as having certain features, one or more of those features described with reference to any embodiment of this disclosure may be implemented in combination with and / or with features of any other embodiment of this disclosure, even if such combination is not explicitly described. In other words, the described embodiments are not mutually exclusive, and the arrangement of one or more embodiments with each other remains within the scope of this disclosure.

[0204] Spatial and functional relationships between components (e.g., between modules, circuit elements, semiconductor layers, etc.) are described using various terms, including “connected,” “joined,” “coupled,” “adjacent,” “closely adjacent,” “on top of,” “above,” “below,” and “set.” Unless explicitly described as “direct,” when the relationship between the first and second components is described in the foregoing disclosure, the relationship includes a direct relationship where no other intervening components are present between the first and second components, as well as an indirect relationship where one or more intervening components exist between the first and second components.

[0205] As described below, the term "group" generally refers to a grouping of one or more elements. However, in various implementations, a "group" can be an empty set in certain situations (in other words, in these cases the group has zero elements). For example, a set of search results generated by a query might depend on the query being an empty set. In situations where the context is ambiguous, the term "non-empty set" can be used to explicitly exclude the empty set—that is, a non-empty set will always have one or more elements.

[0206] A “subset” of the first group typically includes some elements of the first group. In various implementations, a subset of the first group is not necessarily a proper subset: in some cases, the subset can be extended together with (or equal to) the first group (in other words, the subset can include the same elements as the first group). In ambiguous contexts, the term “proper subset” can be used to explicitly indicate that a subset of the first group must exclude at least one element of the first group. Furthermore, in various implementations, the term “subset” does not necessarily exclude an empty set. For example, consider a set of candidates selected based on a first criterion and a subset of that set of candidates selected based on a second criterion; if no element in that set of candidates satisfies the second criterion, then the subset can be an empty set. In ambiguous contexts, the term “non-empty subset” can be used to explicitly indicate the exclusion of an empty set.

[0207] In the accompanying drawings, the direction indicated by the arrows typically illustrates the flow of information of interest (e.g., data or instructions). For example, when components A and B exchange various types of information, but the information transmitted from component A to component B is relevant to the description, the arrow may point from component A to component B. This unidirectional arrow does not imply that no other information is transmitted from component B to component A. Furthermore, for information sent from component A to component B, component B may send a request for or confirmation of receipt of that information to component A.

[0208] In this application, the term "module" is defined as follows, and may be replaced by the term "controller" or the term "circuit". The term "module" may refer to, partially include, or include: application-specific integrated circuits (ASICs); digital, analog, or mixed-signal analog / digital discrete circuits; digital, analog, or mixed-signal analog / digital integrated circuits; combinational logic circuits; field-programmable gate arrays (FPGAs); processor hardware (shared, dedicated, or grouped) that executes code; memory hardware (shared, dedicated, or grouped) coupled to the processor hardware and storing the code executed by the processor hardware; other suitable hardware components that provide the aforementioned functionality; or some or all of these combinations, such as in a system-on-a-chip.

[0209] Modules may include one or more interface circuits. In some examples, the interface circuits may implement wired or wireless interfaces for connecting to a local area network (LAN) or a wireless personal area network (WPAN). Examples of LANs are IEEE Standard 802.11-2020 (also known as the Wi-Fi wireless network standard) and IEEE Standard 802.3-2018 (also known as the Ethernet wired network standard). Examples of WPANs are IEEE Standard 802.15.4 (including the ZigBee Alliance's ZigBee standard) and Bluetooth wireless network standards from the Bluetooth Special Interest Group (SIG) (including Bluetooth SIG core specification versions 3.0, 4.0, 4.1, 4.2, 5.0, and 5.1).

[0210] Modules can communicate with other modules using interface circuitry. Although modules may be described in this disclosure as communicating logically directly with other modules, in various embodiments, modules may actually communicate via a communication system. Communication systems include physical and / or virtual network devices such as hubs, switches, routers, and gateways. In some embodiments, the communication system is connected to or traverses a wide area network (WAN), such as the Internet. For example, a communication system may include multiple LANs interconnected via the Internet or by point-to-point leased lines using technologies such as Multiprotocol Label Switching (MPLS) and Virtual Private Networks (VPNs).

[0211] In various implementations, the functionality of a module can be distributed among multiple modules connected via a communication system. For example, multiple modules may implement the same functionality distributed by a load balancing system. In a further example, the functionality of a module may be split between a server (also known as a remote or cloud) module and a client (or user) module. For example, a client module may include a native or web application that executes on a client device and communicates with the server module over a network.

[0212] Some or all of the hardware characteristics of a module can be defined using a hardware description language, such as IEEE Standard 1364-2005 (commonly referred to as "Verilog") and IEEE Standard 1076-2008 (commonly referred to as "VHDL"). Hardware description languages ​​can be used to manufacture and / or program hardware circuits. In some implementations, some or all of the module's characteristics can be defined by a language such as IEEE 1666-2005 (commonly referred to as "SystemC"), which contains the code and hardware description described below.

[0213] As described above, the term "code" can include software, firmware, and / or microcode, and can refer to programs, routines, functions, classes, and / or objects. Shared processor hardware includes a single microprocessor executing some or all of the code from multiple modules. Grouped processor hardware includes microprocessors that, in combination with additional microprocessors, execute some or all of the code from one or more modules. References to multiple microprocessors include multiple microprocessors on a discrete die, multiple microprocessors on a single die, multiple cores of a single microprocessor, multiple threads of a single microprocessor, or combinations thereof.

[0214] Memory hardware can store data together with or separately from code. Shared memory hardware includes a single storage device that stores some or all of the code from multiple modules. An example of shared memory hardware could be a Level 1 cache on or near a microprocessor die, which can store code from multiple modules. Another example of shared memory hardware could be persistent storage, such as a solid-state drive (SSD) or a magnetic hard disk drive (HDD), which can store code from multiple modules. Grouped memory hardware includes storage devices that combine with other storage devices to store some or all of the code from one or more modules. An example of grouped memory hardware is a Storage Area Network (SAN), which can store the code of a specific module on multiple physical devices. Another example of grouped memory hardware is the random access memory of each server in a group of servers, which collectively store the code of a specific module. The term memory hardware is a subset of the term computer-readable media.

[0215] The apparatus and methods described in this application can be implemented, partially or entirely, by a special-purpose computer by configuring a general-purpose computer to perform one or more specific functions embodied in a computer program. Such apparatus and methods can be described as computerized or computer-implemented apparatus and methods. The aforementioned function blocks and flowchart elements serve as software specifications that can be converted into computer programs through the routine work of skilled technicians or programmers.

[0216] A computer program includes processor-executable instructions stored on one or more non-transitory computer-readable media. A computer program may also include or depend on stored data. A computer program may include a basic input / output system (BIOS) for interacting with the hardware of a special-purpose computer, device drivers for interacting with specific devices of the special-purpose computer, one or more operating systems, user applications, background services, background applications, etc.

[0217] Computer programs may include: (i) descriptive text to be parsed, such as HTML (Hypertext Markup Language), XML (Extensible Markup Language), or JSON (JavaScript Object Notation); (ii) assembly code; (iii) object code generated from source code by a compiler; (iv) source code executed by an interpreter; (v) source code compiled and executed by a just-in-time (JIT) compiler; and so on. As an example only, source code may be written using syntax from the following languages: C, C++, C#, Objective-C, Swift, Haskell, Go, SQL, R, Lisp, Java®, Fortran, Perl, Pascal, Curl, OCaml, JavaScript®, HTML5 (Hypertext Markup Language Version 5), Ada, ASP (Active Server Pages), PHP (PHP: Hypertext Preprocessor), Scala, Eiffel, Smalltalk, Erlang, Ruby, Flash®, Visual Basic®, Lua, MATLAB, SIMULINK, and Python®.

Claims

1. A method, the method comprising: Receive a set of radar signals incident on a set of objects; Determine a first hypothetical model representing the monostatic signal model of the set of radar signals; Determine a second hypothetical model representing the bistatic signal model of the aforementioned set of radar signals; Based on a set of signal criteria, a selected GLRT detector is chosen from a set of generalized likelihood ratio test GLRT detectors; Based on the selected GLRT detector, the ratio between the maximum likelihood function of the second hypothetical model and the maximum likelihood function of the first hypothetical model is determined; Determine whether the ratio is greater than a threshold; Determine whether a set of angles associated with the set of radar signals is available; In response to determining that the set of angles is available, update a set of data associated with a set of objects; as well as In response to determining that the set of angles is unavailable: Based on the aforementioned ratio selection angle estimation method; Estimate the set of angles using the selected angle estimation method; and Track the set of objects based on the set of angles.

2. The method according to claim 1, further comprising: In response to determining that the ratio is greater than a threshold, it is determined that the second hypothetical model is accurate; as well as In response to determining that the ratio is less than or equal to the threshold, it is determined that the first hypothesis model is accurate.

3. The method of claim 1, further comprising autonomously controlling the vehicle to avoid the group of objects.

4. The method according to claim 1, wherein, The angle estimation method selected includes: In response to determining that the ratio is greater than the threshold, a first angle estimation method is selected; and In response to determining that the ratio is less than the threshold, a second angle estimation method is selected.

5. The method according to claim 1, wherein, The set of GLRT detectors is derived from the following formula: 。 6. The method according to claim 5, wherein, The set of GLRT detectors includes: The first GLRT detector is defined as follows: , The second GLRT detector is defined as follows: , The third GLRT detector is defined as follows: ,as well as The fourth GLRT detector is defined as follows: .

7. The method according to claim 6, wherein: The selected angle estimation method is applicable to both direct path reflection and multipath reflection, and The selected GLRT detector is either the third GLRT detector or the fourth GLRT detector.

8. The method according to claim 6, wherein: First Hypothesis Model Defined as: , Second Hypothesis Model Defined as: , It is a set of array observations. It is the first spatial matrix of the reflection path. It is the second space matrix of the reflection path. It is a set of transmitted signals. It is a set of noisy data. θ is a set of angle data containing K elements. N is the number of radar elements. K is the number of reflection paths. M is the number of observations. η is the power level associated with the set of noise data. γ is the threshold. p1 is the likelihood function under the second hypothesis model. p0 is the likelihood function under the first hypothetical model. Denotes the Frobenius norm. P A It is the first projection matrix that maps the vector to the projection onto the subspace formed by A, where, , P B It is the second projection matrix that maps the vector to the projection onto the subspace formed by B, where, , Defined as ,and Defined as , where I is the identity matrix.

9. The method according to claim 8, further comprising: Determine the probability that the second hypothetical model is correct. The probability is defined as follows: 。 10. The method according to claim 8, further comprising: Determine the probability that the first hypothetical model is correct. The probability is defined as follows: 。 11. The method according to claim 8, wherein, The first hypothesis model is based on: The number of elements in the radar array that receive the set of radar signals. The first number of reflection paths, The first number of observations, The set of array observations The first spatial matrix of the reflection path The set of transmitted signals, and The set of noise data.

12. The method according to claim 11, wherein, The second hypothesis model is based on: The number of elements in the radar array that receive the set of radar signals. The second number of reflection paths, The second number of observations, The set of array observations The second spatial matrix of the reflection path The set of transmitted signals, and The set of noise data.

13. The method according to claim 1, wherein, The set of signal standards includes: The first criterion is met when the power level associated with a set of noise data is known, and The second criterion is satisfied when a set of angles associated with the spatial matrix of the reflection path is known.

14. The method according to claim 1, wherein, The set of angles includes the launch direction and the arrival direction.

15. The method according to claim 1, further comprising: Estimate the set of angles before determining whether the ratio is greater than the threshold; as well as Based on the ratio, a set of objects associated with bistatic reflection is identified.

16. A system comprising: Memory hardware configured to store instructions; Processor hardware configured to execute the instructions, wherein the instructions include: Receive a set of radar signals incident on a set of objects; Determine a first hypothetical model representing the monostatic signal model of the set of radar signals; Determine a second hypothetical model representing the bistatic signal model of the aforementioned set of radar signals; Based on a set of signal criteria, a selected GLRT detector is chosen from a set of generalized likelihood ratio test GLRT detectors; Based on the selected GLRT detector, the ratio between the maximum likelihood function of the second hypothetical model and the maximum likelihood function of the first hypothetical model is determined; Determine whether the ratio is greater than a threshold; Determine whether a set of angles associated with the set of radar signals is available; In response to determining that the set of angles is available, update a set of data associated with the set of objects; and In response to determining that the set of angles is unavailable: Based on the aforementioned ratio selection angle estimation method; Estimate the set of angles using the selected angle estimation method; and Track the set of objects based on the set of angles.

17. The system according to claim 16, wherein, The instructions include: In response to determining that the ratio is greater than a threshold, it is determined that the second hypothetical model is accurate; and In response to determining that the ratio is less than or equal to the threshold, it is determined that the first hypothesis model is accurate.

18. The system according to claim 16, wherein, The angle estimation method selected includes: In response to determining that the ratio is greater than the threshold, a first angle estimation method is selected; and In response to determining that the ratio is less than the threshold, a second angle estimation method is selected.

19. The system according to claim 16, wherein: The set of GLRT detectors is derived from the following formula. ,and The set of GLRT detectors includes: The first GLRT detector is defined as follows: , The second GLRT detector is defined as follows: , The third GLRT detector is defined as follows: ,as well as The fourth GLRT detector is defined as follows: .

20. The system according to claim 19, wherein: First Hypothesis Model Defined as: , Second Hypothesis Model Defined as: , It is a set of array observations. It is the first spatial matrix of the reflection path. It is the second space matrix of the reflection path. It is a set of transmitted signals. It is a set of noisy data. θ is a set of angle data containing K elements. N is the number of radar elements. K is the number of reflection paths. M is the number of observations. η is the power level associated with the set of noise data. γ is the threshold. p1 is the likelihood function under the second hypothesis model. p0 is the likelihood function under the first hypothetical model. Denotes the Frobenius norm. P A It is the first projection matrix that maps the vector to the projection onto the subspace formed by A, where, , P B It is the second projection matrix that maps the vector to the projection onto the subspace formed by B, where, , Defined as ,and Defined as , where I is the identity matrix.

Images