CN-122017789-A - Multi-target DOA estimation method for single-base MIMO radar

CN122017789ACN 122017789 ACN122017789 ACN 122017789ACN-122017789-A

Abstract

The invention discloses a single-base MIMO radar multi-target DOA estimation method, and belongs to the field of signal processing. According to the Bayesian formula, the posterior probability density function of the multi-target reflection coefficient of the single-base MIMO radar based on the uniform linear array is firstly deduced, and the posterior probability density function of the multi-target direction angle when the received signal is known is further obtained by combining the deduction result. By searching spectral peaks of the posterior probability density function of the target direction angle, azimuth angles of a plurality of targets can be obtained simultaneously. The invention provides a DOA detection method aiming at the condition of dense distribution of a plurality of targets based on a posterior probability density function. The invention derives the posterior probability density function for multi-target DOA estimation of the uniform linear array single-base MIMO radar based on the information theory, and organically combines all available information through the Bayesian theory, thereby providing more comprehensive and comprehensive angle estimation.

Inventors

Cheng Huiran
XU DAZHUAN
XIE YUSHAN
SHI BINGZHI

Assignees

南京航空航天大学

Dates

Publication Date: 20260512
Application Date: 20251215

Claims (6)

1. The multi-target DOA estimation method of the single-base MIMO radar is characterized by comprising the following steps of: Step 1), constructing a joint conditional probability density function of a receiving signal, a target direction angle and a reflection coefficient in a detection system model by utilizing noise properties of the receiving signal based on a single-base MIMO radar detection system model of a uniform linear array; step 2), deducing a posterior probability density function of a reflection coefficient of a target to be detected of the detection system model and a posterior probability density function of multiple target direction angles according to a Bayes formula; step 3), searching a one-dimensional maximum value of the posterior probability density function of the direction angle of the target to be detected, and obtaining a direction angle dense distribution area of the target to be detected; And 4) aiming at the densely distributed area, carrying out multi-dimensional maximum search on the posterior probability density function of the target direction angle, and obtaining direction angle estimated values of a plurality of densely distributed targets.
2. The method of claim 1, wherein in step 1), the MIMO radar transmitting array and the receiving array are respectively provided with Sum of all The array elements are uniformly distributed and are distributed in a certain pattern, And (3) with Array element intervals of a transmitting array and a receiving array respectively , Is wavelength; setting a certain space with Q targets to be detected, wherein radar transmitting signals are equal-power orthogonal narrowband signal sequences, and the average transmitting power of each signal is The sequence length is K; Representing the mth transmitted signal sequence of length K, using Representing a matrix of radar transmit signals, then the radar receive signals Expressed as: ; Wherein, the , A complex Gaussian reflection coefficient representing the q-th target; For the radar received signal to carry a mean value of 0 and a variance of 0 Additive complex gaussian white noise; For the reception direction matrix of the radar, For the transmission direction matrix of the radar, the target direction vector is developed as , A signal departure direction angle corresponding to the q-th target; stretching the received signal matrix to obtain a received signal vector form expressed as: ; Wherein, the Represents the Khatri-Rao product, Is a reflection coefficient vector; Due to noise For additive complex Gaussian white noise samples, the vector is at the target direction angle And a reflection coefficient vector Under known conditions, a signal vector is received Is expressed as a multi-dimensional conditional probability density function: ; Wherein, the Representing the conjugate transpose operation.
3. The method according to claim 2, wherein in step 2), deriving a posterior probability density function of the detection system model multi-target reflectance vector in combination with a bayesian formulation comprises: ; Wherein, the Representing complex Gaussian reflection coefficient vectors Is 0 mean, variance is Is a complex gaussian distribution of Q dimensions.
4. A method according to claim 3, wherein in step 2) deriving a posterior probability density function of the detection system model multi-target direction angle in combination with a bayesian formula comprises: Setting the first Direction angle corresponding to each target At the position of Is subject to uniform distribution in the observation region of (2), then the target azimuth vector Probability of (2) Obtaining the detection system model at a given received signal according to a Bayesian formula with a constant Target direction angle vector in case The posterior probability density function of (2) is: ; wherein the matrix Sum matrix The following respectively satisfy: ; 。
5. The method according to claim 4, wherein in step 3), given a target search range, the known received signal is processed Posterior probability density function of target direction angle under condition And carrying out one-dimensional maximum value search to obtain the direction angle estimated value of the resolvable target and the distribution area of the dense inseparable target.
6. The method of claim 5, wherein in step 4), a pair function is employed for dense non-separable target distribution areas based on one-dimensional maximum search results And carrying out multidimensional maximum value search to obtain a direction angle estimated value of each target, wherein the angle corresponding to the maximum value is the target direction angle estimated value.

Description

Multi-target DOA estimation method for single-base MIMO radar Technical Field The invention relates to the field of radar signal processing, in particular to a single-base MIMO radar multi-target direction angle estimation method. Background The application background of the target direction angle estimation widely covers a plurality of key fields, and the application of the technology is mainly focused on the fields of radar systems, sonar systems, communication systems, aerospace, unmanned systems, earth observation and the like. Most of the existing direction angle estimation uses MUSIC, RD-MUSIC, ESPRIT, PM and other algorithms, and the algorithms can accurately estimate the target direction under ideal conditions. However, the above algorithm is relatively sensitive to noise and is susceptible to large effects at low signal-to-noise ratios. Disclosure of Invention The invention provides a multi-target direction angle estimation method of a single-base MIMO radar, which provides a more accurate method for multi-target DOA estimation of a uniform linear array single-base MIMO radar at a low signal-to-noise ratio (SNR). The embodiment of the invention provides a single-base MIMO radar multi-target direction angle estimation method based on a uniform linear array, which comprises the following steps: Step 1), constructing a single-base MIMO radar detection system model based on a uniform linear array, and constructing a joint conditional probability density function of a radar detection system model receiving signal, a target direction angle and a reflection coefficient by utilizing the property of noise in the receiving signal; Step 2), deriving a posterior probability density function of a reflection coefficient of a target to be detected and a posterior probability density function of a multi-target direction angle in a single-base MIMO radar detection system model by combining a Bayesian formula; step 3), scanning a one-dimensional maximum value of the target direction angle posterior probability density function in the region to be detected to obtain a direction angle estimated value of the sparse resolvable target and a distribution region of the dense unresolved target; And 4) aiming at the dense target distribution area, carrying out multidimensional resolution on the target direction angle posterior probability density function to obtain the direction angle estimated value of the dense distribution target. Optionally, in one embodiment of the present invention, in step 1), the MIMO radar transmitting array and the receiving array are respectively provided withSum of allThe array elements are uniformly distributed and are distributed in a certain pattern,And (3) withArray element intervals of a transmitting array and a receiving array respectively,Is wavelength; setting a certain space with Q targets to be detected, wherein radar transmitting signals are equal-power orthogonal narrowband signal sequences, and the average transmitting power of each signal is The sequence length is K; Representing the mth transmitted signal sequence of length K, using Representing a radar transmit signal matrix, the radar receive signal Y is represented as: Wherein, the ,Complex Gaussian reflection coefficient representing the q-th objectFrom mean 0, variance ofIs composed of additive complex Gaussian white noise random variables; For the reception direction matrix of the radar, For the transmission direction matrix of the radar, the target direction vector is developed as,A signal departure direction angle (arrival direction angle) corresponding to the q-th target; stretching the received signal matrix to obtain a received signal vector form expressed as: Wherein, the Represents the Khatri-Rao product,As a vector of the reflection coefficient,Is an additive complex Gaussian white noise vector; Due to noise For additive complex Gaussian white noise samples, at target direction anglesAnd reflectance coefficientUnder known conditions, a signal is receivedIs expressed as a multi-dimensional conditional probability density function: Optionally, in an embodiment of the present invention, in step 2), deriving a joint posterior probability density function of multiple target reflection coefficients in a uniform linear array single-base MIMO radar detection system model in combination with a bayesian formula includes: Wherein, the Representing complex Gaussian reflection coefficient vectorsIs 0 mean, variance isIs a complex gaussian distribution of Q dimensions. Alternatively, in one embodiment of the present invention, in step 2), a uniform linear array single base MIMO radar detection system model is derived from a bayesian formula at a given received signalIn the case of (a), a posterior probability density function of multi-target direction angle estimation, comprising: Assuming each target direction angle Are all in the observation intervalThe target azimuth angle vector is distributed uniformly by the oral administr