PHASE CORRECTION FOR OBJECT DETECTION INCLUDING TDMA

Description

BACKGROUND

[0001] A variety of scenarios require signal handling. For example, radar systems and wireless communication systems include over-the-air signaling. One challenge associated with such systems is using multiple signals to increase information regarding detected objects while avoiding processing difficulties introduced by using multiple signals. For example, it is necessary for different signals to be orthogonal so that they can be individually detected and processed. One way of achieving orthogonality involves time division multiplexing (TDMA) in which signal transmissions are separated in time.

[0002] While TDMA can be an efficient way of increasing the number of signals available for object detection and analysis, TDMA techniques may introduce signaling errors that have to be addressed to provide accurate detection information. For example, TDMA may introduce an additional phase shift based on the difference in transmission timing and such a phase shift could skew object detection information if it is not properly recognized or its influence is not removed from the received detection signal processing techniques. When detecting moving objects TDMA-based phase shifts are related to the unknown velocity of the moving object, which increases the difficulties associated with processing detection signals including such a phase shift. EP 3 147 685 A1 discloses an object detection system comprising a first transmitter that respectively transmits a first signal at a first time, a second transmitter that respectively transmits a second signal at a second time, a plurality of receivers that receive the signals, and a processor that is configured to determine whether a transmission timing phase shift exists between the received first signal and the received second signal based on a relationship between matrices including the received signals.

[0003] WO 2018 / 071077 A2 discloses a radar system having a processor which is configured to estimate a velocity induced phase shift in a virtual array vector of signals based on signals received by each receiver corresponding to a sequence of chirps.

[0004] The paper from Roos Fabian et al.: "Enhancement of Doppler Unambiguity for Chirp-Sequence Modulated TDM-MIMO Radars", 2018 IEEE MTT-S INTERNATIONAL CONFERENCE ON MICROWAVES FOR INTELLIGENT MOBILITY (ICMIM), IEEE, 2018-04-15, pages 1-4, XP033391296, DOI: 10.1109 / ICMIM.2018.8443352, discloses a signal processing method for radar signals including a chirp-sequence modulation.

[0005] DE 10 2008 061932 A1 discloses a radar system having a processing unit which is configured to carry out a velocity correction process based on a fast Fourier transformation.SUMMARY

[0006] The invention provides an object detection system according to claim 1.

[0007] The invention further provides an object detection method according to claim 5.

[0008] Various features and advantages of at least one disclosed example embodiment will become apparent to those skilled in the art from the following detailed description. The drawings that accompany the detailed description can be briefly described as follows.BRIEF DESCRIPTION OF THE DRAWINGS

[0009] Figure 1 diagrammatically illustrates an object detection system designed according to an embodiment of this invention. Figure 2 is a flow chart diagram summarizing an example technique designed according to an embodiment of this invention. Figure 3 schematically illustrates matrices useful in an embodiment of this invention. DETAILED DESCRIPTION

[0010] Embodiments of this invention allow for effective and accurate object detection when radar signals are transmitted using time division multiplexing (TDMA). Any phase shift introduced by TDMA can be detected and effectively removed from the signal processing based on determining a relationship between matrices that include the received signals. In an example embodiment, a relationship between traces of covariance matrices including the received signals indicates whether a TDMA phase shift exists and received signals can be processed based on that information.

[0011] Figure 1 schematically illustrates an object detection system 20 supported on a vehicle 22. The system 20 includes a plurality of first transmitters 24 that transmit respective first signals at a first time. A plurality of second transmitters 26 respectively transmit second signals at a second time. The first and second signals are transmitted using TDMA. In the illustrated example embodiment the signals are linear frequency-modulated continuous wave signals, such as those typically used for automotive RADAR. Other signal types are used in some embodiments.

[0012] The system 20 includes a plurality of receivers 28 that receive the first and second signals after those signals reflect off an object in a vicinity or pathway of the vehicle 22. Figure 1 shows two first transmitters 24, two second transmitters 26 and four receivers 28 but the illustrated embodiment includes more of each. One example embodiment of the system 20 includes four first transmitters 24, four second transmitters 26 and thirty-two receivers 28.

[0013] A processor 30 is configured to process the received signals to make determinations regarding a detected or tracked object, such as the range, angle and Doppler velocity of the object. The processor 30 is configured to compensate for a transmission timing phase shift that may exist between the received first signals and the received second signals. The processor 30 determines whether a transmission phase shift exists based on a relationship between matrices that include the received signals.

[0014] The processor includes at least one computing device, such as a microprocessor, and memory. The processor 30 may be a dedicated device or component or be realized through part of an existing controller or processor on the vehicle 22 that is also used for other purposes. Those skilled in the art who have the benefit of this description will realize what arrangement of a processor, transmitters and receivers will meet their particular needs.

[0015] Figure 2 is a flowchart diagram 40 that summarizes an embodiment using TDMA to achieve signal orthogonality in MIMO radars. At 42, the first transmitters 24 transmit the first signals at a first time. The second transmitters transmit the second signals at 44. The second transmission time of the second signals in this example is subsequent to the first transmission time. The TDMA technique of this example includes repeatedly transmitting the first and second signals at scheduled intervals.

[0016] At 46, the receivers 28 receive the signals reflected from an object. The received signal at any of the receivers 28, which can be considered a (n,m) receiver array element, can be represented as a delayed transmitted signal with the following equation. s nm t = Ae jφ R e jφ θ e jφ D e jφ S where t is the discretized time; φ R ,φ θ ,φ D are phase terms containing information about the detected object's range, angle, and velocity respectively; and φ S is the phase term due to the sequential transmitter activation.

[0017] The total delay in the signal represented in equation (1) includes four terms. Each of these terms provides information related to the target or detected object. The first delay θ R is due to the distance between the reference point of the array of receivers 28 and the object and can be used to compute the object range. The second delay φ θ is the result of the position of the receiver 28 with respect to the reference point of the receiver array. The second delay φ θ is used to compute the target angle. The third delay φ D is a consequence of a moving object or a moving platform and is useful to determine the object's velocity. The φ S delay is due to the TDMA sequential transmitter activation. The processor 30 is configured to determine whether φ S exists and to estimate φ S for proper data analysis.

[0018] In the embodiment the TDMA strategy includes activating the second transmitters 26 at one-half the pulse repetition interval of the first transmitters 24. This allows for φ S to be represented by the following equation. φ S = 2 π 2 v λ T where the pulse repetition interval is 2T, λ denotes the wavelength and ν is the object's true velocity. The following equation describes that velocity. v = v m + lv un , l = 0 , ± 1 , ± 2 , … where ν m is the measured target's velocity and ν un is the maximum unambigious detectable velocity, v un = λ / (4T). Replacing equation (3) into equation (2), φ S becomes: φ S = 4 π v m / λ T + 4 πl v un / λ T = 4 π v m / λ T + lπ

[0019] Based on equation (4), it can be observed that the φ S phase term introduces a phase shift of 0 or 180 degrees depending on the target's true velocity ν. Since the target's velocity is not known a priori, the processor 30 is configured to estimate this phase shift for proper data analysis.

[0020] In Figure 2, the processor determines a reference covariance matrix including the received first signals at 48. Figure 3 schematically shows received signals at 50. The illustration includes 256 received signals representing each of thirty-two receivers 28 receiving a reflected first signal from each of four first transmitters 24 and a reflected second signal from each of four second transmitters 26. The received first signals are included in a reference set enclosed in the box 52. All of the received first signals from a single first transmission time are included in the reference set. The processor 30 determines the reference covariance matrix using the first signals in the reference set 52.

[0021] At 54 in Figure 2, the processor determines an unmodified covariance matrix including at least some of the received first signals and at least some of the received second signals. In this example, a set of the received first signals enclosed in the box 56 in Figure 3 is included in the unmodified covariance matrix. This set 56 is a subset of the received first signals and can be represented by Y 1 . The unmodified covariance matrix also includes a set of the received second signals enclosed in the box 58. This set of second signals is a subset of all of the received second signals and can be represented by Y 2 . In this example, the unmodified covariance matrix can be represented by R xy of matrix [Y 1 ,Y 2 ].

[0022] At 60 in Figure 2, the processor 30 determines a modified covariance matrix using the first signals in Y 1 and a modified version of the set 58 of second signals. The processor 30 determines modified values of the second signals in the set 58 by multiplying each by -1. That set of modified values can be represented as -Y 2 . The modified covariance matrix determined at 60 can be represented by R xy' of matrix [Y 1 ,-Y 2 ].

[0023] The modified covariance matrix includes using -1 as a factor applied to the received second signals in the set 58 because φ S is part of an exponential function. Accordingly, Euler's identity indicates that the received signal is effectively multiplied by 1 if there is no transmission timing phase shift and multiplied by -1 if there is such a phase shift (i.e., e jφs< = e j(4π (v m / λ)T+lπ)< = (-1) l e j4π (v m / λ)T< ).

[0024] At 62, the processor 30 determines respective traces of the covariance matrices. The processor uses a relationship between the covariance matrices as basis to determine whether a TDMA or transmission timing phase shift φ S exists.

[0025] A covariance matrix of an NxN square matrix X is defined as R XX , and its trace value can be computed in a known manner as tr R = ∑ i = 1 N R ii

[0026] The processor 30 determines at 64 which of R xy or R xy' is more similar to the reference covariance matrix. At 66, the processor 30 determines whether a TDMA phase shift exists based on which of the traces is more similar to the trace of the reference covariance matrix. If the trace of the modified covariance matrix is more similar to the reference covariance matrix that indicates that the transmission timing phase shift does exist. If, on the other hand, the trace of the unmodified covariance matrix is more similar to the reference covariance matrix that indicates that the transmission timing phase shift does not exist.

[0027] The processor 30 is configured to process the received signals based on a technique that accounts for the existence or absence of a transmission timing phase shift φ S at 68. When there is no phase shift the fourth phase term of equation (1) is eliminated and signal processing to determine the object's range, angle and Doppler velocity proceeds without φ S . On the other hand when φ S exists, the processor includes a 180 degree shift value for φ S in equation (1).

[0028] The embodiment facilitates using cost-effective TDMA signaling techniques without difficulties or errors otherwise associated with introducing a TDMA phase shift between signals. The disclosed system and method improve object detection and tracking efficiency and accuracy.

Claims

1. An object detection system (20), comprising: a plurality of first transmitters (24) that respectively transmit a first radar signal at a first time; a plurality of second transmitters (26) that respectively transmit a second radar signal at a second time; a plurality of receivers (28) that receive the radar signals; wherein the receivers (28) receive the radar signals after the radar signals reflect off an object; and a processor (30) that is configured to determine whether a Time Division Multiplexing (TDMA) transmission timing phase shift exists between the received first radar signals and the received second radar signals based on a relationship between matrices including the received signals; characterized in that the matrices comprise covariance matrices; the processor (30) is configured to: determine a reference covariance matrix of the received first radar signals; determine an unmodified covariance matrix including at least part of the received first radar signals and at least part of the received second radar signals; determine a set of modified second signal values for the part of the received second radar signals multiplied by negative one; determine a modified covariance matrix including the part of the received first radar signals and the set of modified second signal values; and determine the relationship between the matrices based on (i) a relationship between the reference covariance matrix and the unmodified covariance matrix and (ii) a relationship between the reference covariance matrix and the modified covariance matrix; the processor (30) is configured to: determine a reference trace T of the reference covariance matrix; determine a trace Tu of the unmodified covariance matrix; determine a trace Tm of the modified covariance matrix; determine the relationship between the reference covariance matrix and the unmodified covariance matrix based on a relationship between the reference trace T and the trace Tu; and determine the relationship between the reference covariance matrix and the modified covariance matrix based on a relationship between the reference trace T and the trace Tm; and the processor (30) is configured to: determine whether the reference trace T is more similar to the trace Tu or the trace Tm; and determine whether the TDMA transmission timing phase shift exists based on which of the traces Tu or Tm is more similar to the reference trace T.

2. The object detection system (20) of claim 1, wherein the processor (30) is configured to: determine that the TDMA transmission timing phase shift exists when the reference trace T is more similar to the trace Tm than the trace Tu; determine that the TDMA transmission timing phase shift does not exist when the reference trace T is more similar to the trace Tu than the trace Tm; and use a phase shift value of one hundred eighty degrees for processing at least some of the received radar signals when the TDMA transmission timing phase shift exists.

3. The object detection system (20) of any of claims 1 - 2, wherein: there are four first transmitters (24); there are four second transmitters (26); there are thirty-two receivers (28); there are one hundred twenty-eight received first radar signals in the reference covariance matrix; there are sixty-four received first signals in the part of the received first radar signals; and there are sixty-four received second signals in the part of the received second radar signals.

4. The object detection system (20) of any of claims 1 - 3, wherein the processor (30) is configured to use a phase shift value of one hundred eighty degrees for processing at least some of the received radar signals when the TDMA transmission timing phase shift exists.

5. An object detection method, comprising: transmitting first radar signals at a first time from a plurality of first transmitters (24); transmitting second radar signals at a second time from a plurality of second transmitters (26); receiving the radar signals at a plurality of receivers (28); wherein the receivers (28) receive the radar signals after the radar signals reflect off an object; and determining whether a Time Division Multiplexing (TDMA) transmission timing phase shift exists between the received first radar signals and the received second radar signals based on a relationship between matrices including the received signals; characterized in that the matrices comprise covariance matrices; the method comprises: determining a reference covariance matrix of the received first radar signals; determining an unmodified covariance matrix including at least part of the received first radar signals and at least part of the received second radar signals; determining a set of modified second signal values for the part of the received second radar signals multiplied by negative one; determining a modified covariance matrix including the part of the first radar signals and the set of modified second signal values; and determining the relationship between the matrices based on (i) a relationship between the reference covariance matrix and the unmodified covariance matrix and (ii) a relationship between the reference covariance matrix and the modified covariance matrix; the method comprises: determining a reference trace T of the reference covariance matrix; determining a trace Tu of the unmodified covariance matrix; determining a trace Tm of the modified covariance matrix; determining the relationship between the reference covariance matrix and the unmodified covariance matrix based on a relationship between the reference trace T and the trace Tu; and determining the relationship between the reference covariance matrix and the modified covariance matrix based on a relationship between the reference trace T and the trace Tm; and the method comprises: determining whether the reference trace T is more similar to the trace Tu or the trace Tm; and determining whether the TDMA transmission timing phase shift exists based on which of the traces Tu or Tm is more similar to the reference trace T.

6. The method of claim 5, further comprising determining that the TDMA transmission timing phase shift exists when the reference trace T is more similar to the trace Tm than the trace Tu; determining that the TDMA transmission timing phase shift does not exist when the reference trace T is more similar to the trace Tu than the trace Tm; and using a phase shift value of one hundred eighty degrees for processing at least some of the received radar signals when the TDMA transmission timing phase shift exists.

7. The method of any of claims 5 - 6, wherein there are four first transmitters (24); there are four second transmitters (26); there are thirty-two receivers (28); there are one hundred twenty-eight received first radar signals in the reference covariance matrix; there are sixty-four received first signals in the part of the received first radar signals; and there are sixty-four received second signals in the part of the received second radar signals.

8. The method of any of claims 5 - 7, comprising using a phase shift value of one hundred eighty degrees for processing at least some of the received radar signals when the TDMA transmission timing phase shift exists.