CN111505601A - Linear motion demodulation implementation method based on improved differential cross multiplication - Google Patents

Abstract

A linear motion demodulation method based on improved difference cross multiplication is based on trigonometric function geometric theorem, directly performs differential processing on an I/Q signal after direct current calibration without inverse orthotropic transformation, and then performs cross multiplication and logical addition operation on the I/Q signal to extract a motion signal to be detected.

Description

Linear motion demodulation implementation method based on improved differential cross multiplication

Technical Field

The invention relates to a technology in the field of radar detection, in particular to a linear motion demodulation implementation method based on improved differential cross multiplication.

Background

In the radar detection of the target motion track, a signal processing technology based on phase information is needed. The target movement causes the nonlinear phase modulation of the radar electromagnetic wave signal, so a high-linearity demodulation method is required. The most commonly used linear demodulation algorithms at present are both the arctan algorithm and the differential cross multiplication (DACM) algorithm. First, the arctangent algorithm is simple and highly linear, once the arctangent algorithm has attracted a lot of attention for linear demodulation. However, the arctangent algorithm has strict value-taking domain constraint on the motion signal to be measured, and the universal application of the algorithm is influenced. When the phase locus of the modulation signal is not in the unique quadrant or the range of the first quadrant and the fourth quadrant due to the motion information to be detected, the demodulation signal generates phase ambiguity, and the waveform is subjected to jumping truncation.

In order to break the limitation of the value range and make the demodulation algorithm applicable to the whole range, scientists further propose the DACM algorithm. The DACM algorithm is realized by performing differentiation and integration operations again on the basis of the arctan algorithm, which causes that the DACM algorithm needs to perform nested operation and a large amount of approximation, increases the accumulated noise and approximation error of demodulation results, and reduces the accuracy and stability of demodulation signals. In addition, both the arctangent algorithm and the DACM algorithm depend on the accuracy of the pre-dc offset calibration step excessively, and the robustness and the anti-noise capability of the algorithm are affected.

Disclosure of Invention

The invention provides a linear motion demodulation implementation method based on improved differential cross multiplication aiming at the phase ambiguity problem of the demodulation signals in the prior art, which utilizes the geometric theorem of difference and trigonometric function, does not need inverse tangential transformation, only relates to the first-order operation of I/Q signals, greatly simplifies the operation process, reduces operation approximation and accumulated noise, and improves the linearity, stability and noise resistance of the algorithm.

The invention is realized by the following technical scheme:

the invention relates to a linear motion demodulation implementation method based on improved differential cross multiplication, which is based on trigonometric function geometric theorem, directly performs differential processing on an I/Q signal after direct current calibration without inverse tangent transformation, and performs cross multiplication and logical addition operation on the I/Q signal and an original I/Q signal to extract a motion signal to be detected.

The original I/Q signals refer to: and amplifying, filtering and carrying out down-conversion treatment on the radar echo signal received by the radio frequency receiving end to obtain the radar echo signal.

Technical effects

The invention integrally solves the technical problems that in the prior art, the value taking domain of a measurable motion signal is strictly limited, the phase of a demodulated signal is fuzzy, the accuracy of a preposed direct current offset calibration algorithm is excessively depended on, the linearity and the stability of the algorithm are influenced, the conventional DACM algorithm is realized by a large number of nested operations, the operation error and the accumulated noise are introduced, the anti-noise capability of the algorithm is reduced, and the complex algorithm operation occupies too many computer resources and influences the execution efficiency of the algorithm.

Compared with the prior art, the method removes the limit of the value taking domain of the measurable motion signal, does not need tangent and anti-tangent transformation, and directly carries out differential cross multiplication demodulation on the original I/Q signal to obtain the motion signal, thereby greatly reducing the iteration times, the accumulated noise and the approximate error of the algorithm, ensuring the high linearity of the algorithm, improving the operation efficiency of the algorithm and simultaneously showing robustness to the noise.

Drawings

FIG. 1 is a block diagram of the system of the present invention;

FIG. 2 is a flow chart of the method of the present invention;

FIG. 3 is a diagram illustrating simulation results of small-amplitude motion;

in the figure: (a) the I/Q phase track, and a demodulation signal obtained by (b) an arc tangent algorithm and (c) the method; simulation result of large amplitude motion: (d) the I/Q phase track, and a demodulation signal obtained by (e) an arc tangent algorithm and (f) the method;

FIG. 4 is a diagram illustrating the simulation results of motion signals at 25 dB;

in the figure: (a) an I/Q phase locus, and a demodulation signal obtained by processing (b) an arc tangent algorithm, (c) an existing DACM algorithm and (d) the method;

FIG. 5 is a diagram illustrating the results of demodulating a motion signal having an amplitude of about 10mm measured using a 2-4 GHz radar module;

in the figure: (a) I/Q phase locus, (b), (c) is the inverse tangent algorithm and demodulation of this method restores the signal oscillogram separately;

FIG. 6 is a schematic diagram showing the experimental results of measuring about 4mm linear motion using a 120GHz radar module;

in the figure: (a) I/Q phase trajectory, and via (b) an arctangent algorithm, (c) an existing DACM algorithm, and (d) the present method, respectively.

Detailed Description

The invention relates to a linear motion demodulation implementation method based on improved differential cross multiplication of the system, which specifically implements scenes including small-amplitude motion detection, large-amplitude motion detection and motion information demodulation simulation under large noise.

Example 1

Small amplitude motion detection

This embodiment utilizes 2.4 GHz's radar sensor system to detect the mechanical motion of target object within 10mm, based on trigonometric function geometric theorem, does not pass through anti-tangential transformation, directly carries out the cross multiplication to calibration signal and handles in order to realize motion information demodulation, specifically does:

step ① differentiates the calibration signal directly in the time domain to yield:

step ② utilizes the geometric theorem cos of trigonometric functions²Φ+sin²And (3) directly extracting a motion signal to be measured from the differential result of the step ①, wherein phi is 1:

because the differential signal is easily interfered by noise and is not beneficial to the processing of a subsequent circuit, the motion signal to be detected is restored to the motion signal under the digital domain in the next step, and a final digitized algorithm expression is obtained:

the results of the experiment are recorded in figure 5. As shown in fig. 5 (a), when the motion signal to be measured is not within the range of the value-taking region of the arctangent algorithm, the waveform of the demodulated signal is subjected to a skip truncation, and a phase ambiguity occurs. In contrast, the motion signal reconstructed by the modified DACM algorithm proposed in fig. 3 (b) is free from the limit of the value range, and has higher algorithm universality and robustness.

Example 2

Large amplitude motion detection

In the embodiment, the measurement of the large-amplitude motion signal is realized by using the 120GHz radar detection module with the wavelength of only 2.5mm, and except for small-amplitude motion which is not in the range of the value taking domain, the large-displacement motion can also cause phase aliasing and cause phase ambiguity. In fig. 3 (a) - (c), the I/Q signal phase trace spans the first and second quadrants and the trace extent is less than pi/2. In this case, the motion signal reconstructed by the arctan algorithm is phase blurred, whereas the demodulated signal obtained by the present method is free from phase distortion. Further, when the I/Q phase trajectory is far beyond 2 π, the arctan demodulation result again exhibits phase discontinuity, as shown in (d) - (f) of FIG. 3. On the contrary, the method can still maintain high linearity, authenticity and integrity of the target motion information.

This embodiment is based on the trigonometric function geometric theorem, and does not pass through inverse tangential transformation, and directly performs cross multiplication processing on the calibration signal to realize motion information demodulation, specifically:

step ① differentiates the calibration signal directly in the time domain to yield:

step ② utilizes the geometric theorem cos of trigonometric functions²Φ+sin²And (3) directly extracting a motion signal to be measured from the differential result of the step ①, wherein phi is 1:

because the differential signal is easily interfered by noise and is not beneficial to the processing of a subsequent circuit, the motion signal to be detected is restored to the motion signal under the digital domain in the next step, and a final digitized algorithm expression is obtained:

due to the extremely short wavelength of the large-amplitude motion, the I/Q phase locus can easily overflow by 2 pi by weak displacement, so that a large amount of phase aliasing is caused, and the large-amplitude motion characteristic is achieved. In this experiment, the object to be measured is moved linearly by 4mm in a single direction, and fig. 6 records the I/Q phase locus and the demodulation results obtained by the three algorithms, respectively. As shown in fig. 6 (a), due to the defect of the pre-dc offset calibration algorithm, the I/Q signal after calibration has a mismatch phenomenon, so that part of dc offset remains in the I/Q signal, and the amplitude cannot be completely normalized. This has a significant impact on the current DACM algorithm, which relies heavily on calibration accuracy.

Comparing the demodulation results of the algorithms in fig. 6, it is known that large amplitude motion causes phase ambiguity and waveform distortion of the arctan demodulation result. However, although the existing DACM algorithm does not generate phase ambiguity, due to the imperfection of the pre-calibration algorithm, the demodulation signal still changes with time, and the linearity of the signal is reduced. Compared with the traditional DACM algorithm, the improved DACM algorithm provided by the method not only well solves the problem of phase ambiguity, but also can perfectly restore the motion signal to be detected under large displacement, the Root Mean Square Error (RMSE) of the improved DACM algorithm is improved by 32dB compared with the traditional DACM algorithm, and the linearity of the demodulation algorithm after calibration mismatch is obviously improved.

Example 3

Motion information demodulation simulation under large noise

In the traditional DACM algorithm, Q/I is subjected to inverse tangential transformation as a whole, differentiation and cross multiplication are performed on the basis of the inverse tangential transformation, a large amount of approximation errors and accumulated noise are introduced in the operation process, the anti-noise capability of the algorithm is seriously influenced, and higher requirements are provided for the precision of a pre-calibration step. Compared with the DACM algorithm in the method, the DACM algorithm has a more simplified expression, the operation times and the approximation error are greatly reduced, and the stability and the anti-noise capability of the demodulation process are optimized. Fig. 4 shows the demodulated signal obtained by each algorithm at a signal-to-noise ratio of 25 dB. It can be seen that compared with the existing DACM algorithm, the variance of the arctangent algorithm is minimum, and the algorithm proposed by the method also has higher stability and robustness, and the Root Mean Square Error (RMSE) of the algorithm is close to the arctangent result and is improved by about 9dB compared with the existing DACM algorithm.

In summary, compared with the existing differential cross multiplication algorithm, the present invention firstly has no phase ambiguity, secondly has high linearity under large displacement measurement (the root mean square error is improved by 32dB compared with the traditional DACM algorithm), and simultaneously has high stability and robustness to noise (the root mean square error is improved by 9dB compared with the traditional DACM algorithm under 25dB signal-to-noise ratio). In addition, compared with the existing DACM algorithm, the method does not need tangent and inverse tangent transformation, reduces the iteration times of the algorithm, saves the computing resources and improves the operation efficiency of the algorithm.

The foregoing embodiments may be modified in many different ways by those skilled in the art without departing from the spirit and scope of the invention, which is defined by the appended claims and all changes that come within the meaning and range of equivalency of the claims are therefore intended to be embraced therein.

Claims (2)

1. A linear motion demodulation implementation method based on improved differential cross multiplication is characterized in that a motion signal to be detected can be extracted by directly carrying out differential processing on an I/Q signal subjected to direct current calibration based on trigonometric function geometric theorem without inverse tangent transformation, and then carrying out cross multiplication and logical addition operation on the I/Q signal and an original I/Q signal.

2. The linear motion demodulation implementation method based on the improved differential cross multiplication as claimed in claim 1, characterized by specifically:

step ① differentiates the calibration signal directly in the time domain to yield:

step ② utilizes the geometric theorem cos of trigonometric functions²Φ+sin²And (3) directly extracting a motion signal to be measured from the differential result of the step ①, wherein phi is 1:

because the differential signal is easily interfered by noise and is not beneficial to the processing of a subsequent circuit, the motion signal to be detected is restored to the motion signal under the digital domain in the next step, and a final digitized algorithm expression is obtained:

Images