阅读说明：本技术 基于块稀疏贝叶斯算法的上下链路通信感知一体化方法 (Up-down link communication perception integrated method based on block sparse Bayesian algorithm ) 是由 程知群 郭昊阳 李航 于 2021-06-02 设计创作，主要内容包括：本发明公开了基于块稀疏贝叶斯算法的上下链路通信感知一体化方法,将OFDM技术与压缩感知算法结合,首先是将数字基带信号通过PSK或QAM调制成数据符号,对数据符号进行空间预编码,经过对采样数进行IFFT后将时域信号发送给各个射频单元。其次通过射频单元的阵列响应向量确定频域信道矩阵；通过信道矩阵和时域信号确定射频单元中上/下行链路的接收和发射信号；建立稀疏表示类on-grid延迟量化模型；然后确定块稀疏信号,将参数估计问题转化为MMV块稀疏问题；最后利用快速边缘化快稀疏贝叶斯学习算法进行参数估计。(The invention discloses an uplink and downlink communication perception integrated method based on a block sparse Bayesian algorithm, which combines an OFDM technology with a compressed perception algorithm, firstly modulates a digital baseband signal into a data symbol through PSK or QAM, performs spatial precoding on the data symbol, and transmits a time domain signal to each radio frequency unit after IFFT is performed on a sampling number. Secondly, determining a frequency domain channel matrix through an array response vector of the radio frequency unit; determining the receiving and transmitting signals of an uplink/downlink in the radio frequency unit through the channel matrix and the time domain signal; establishing a sparse representation type on-grid delay quantization model; then determining a block sparse signal, and converting the parameter estimation problem into an MMV block sparse problem; and finally, performing parameter estimation by using a fast marginalization fast sparse Bayesian learning algorithm.)

1. An integrated method for sensing uplink and downlink communication based on a block sparse Bayesian algorithm is characterized by comprising the following steps:

s1: modulating the baseband signal to different subcarriers by utilizing a multicarrier modulation technology, and distributing the modulated time domain signal to different base station remote radio frequency units;

s2: determination of the steering vector a of the transmitting and receiving antennas by means of a uniform linear array radar per remote radio unit_M(σ)；

S3: determining a frequency domain channel matrix H on the corresponding nth sub-carrier on the t OFDM symbol in a channel with L multipath signals through array response vectors of a receiving end and a transmitting end_n；

S4: determining the received and transmitted signals Y of remote RF units in the uplink and downlink_n，tEstablishing a model after the quantitative delay of a sparse representation type on-grid direction-finding method

S5: converting the parameter estimation problem into an MMV block sparsity problem to obtain Y_tAnd block sparse signals, and obtaining the estimated values of the parameters through a fast marginalization block sparse Bayes learning algorithm.

2. The method of claim 1, wherein the baseband signal is modulated to different subcarriers by using a multi-carrier modulation technique, and the modulated time domain signal is distributed to different base station remote radio frequency units, in order to modulate the digital baseband signal to a data symbol by quadrature amplitude modulation first, then perform spatial pre-coding on the data symbol, provide power gain for a receiving end by pre-coding, and reduce signal processing difficulty of the receiving end; then, performing inverse fast Fourier transform on the frequency domain sampling data, and converting the frequency domain sampling data into time domain data; the resulting composite time domain signal x (t) is then distributed to the respective remote radio units.

3. Method according to claim 2, characterized in that the determination of the steering vector a of the transmitting and receiving antennas is carried out by a uniform linear array radar of each remote radio unit_M(σ), assuming that there is a planar wavefront in the signal propagation, the steering vector for a Uniform Linear Array (ULA) of M antennas is given by:

a_M(σ)＝[1，e^jπsinσ，...，e^{jπ(M-1)sinσ}]^T (1)

where M is the number of array antennas and σ is the angle of arrival or angle of departure.

4. The method of claim 3, wherein the frequency domain channel matrix H on the nth sub-carrier corresponding to the tth OFDM symbol in the channel with L multipath signals is determined by the array response vectors of the receiving end and the transmitting end_nLet the angle of arrival or angle of departure of the first multipath be denoted as χ_lAnd σ_lThe time-domain baseband signal impulse response with M1 transmit antennas and M2 receive antennas is:

wherein, b_lIs the amplitude value, τ_lIs the propagation delay, f_D，lIs the associated propagation delay; converting the received signal to the frequency domain for processing, the frequency domain channel matrix at the nth subcarrier for the tth OFDM block is given by:

wherein the content of the first and second substances,is the effective OFDM symbol period T, T_sIs the effective OFDM length T plus the cyclic prefix length T_p。

5. The method of claim 4, wherein the determination of the received and transmitted signals y for remote RF units in the uplink and downlink is performed by comparing the received and transmitted signals y with a predetermined threshold value_n，tEstablishing a model after the quantitative delay of a sparse representation type on-grid direction-finding methodTo assume that there are a total of Q remote radio units, for downlink sensing, each remote radio unit receives reflected downlink signals from itself and the other Q-1 remote radio units, and the received signal at the qth remote radio unit at the nth subcarrier of the tth OFDM block can be expressed as:

wherein x_q，n，tIs the transmitted signal of the qth remote radio unit at the nth subcarrier of the tth OFDM block, b_q，l，τ_q，l，X_q，l，σ_q，lRespectively obtaining complex amplitude, propagation delay, arrival angle and departure angle of the qth long-distance radio frequency unit in the ith multipath; and wherein

x_n，t＝(x_1，n，t，...，x_Q，n，t)^T (7)

Wherein U is a diagonal matrix of array element departure angles,orColumn of (1) isOr z_n，tIs noise;

similarly, the received signal of the remote rf unit when sensing uplink is:

wherein b is_k，l，τ_k，l，χ_k，l，σ_k，lAnd f_D，k，lRespectively obtaining complex amplitude, propagation delay, arrival angle, departure angle and Doppler frequency shift of the kth remote radio frequency unit in the ith multipath;

then considering the on-grid sparse model after delay quantization, and using K to represent a transmitting end in an uplink or a downlink, namely a mobile station or a long-distance radio frequency unit; by M_TThe number of array elements used for receiving signals; the number of array elements used for reception is denoted by M,

wherein, C_nIs a quantized delay diagonal matrix andN_Pnot less than L and N_PLess than or equal to gN; p is an Np x L rectangular array matrix that maps signals from the subscriber/remote radio units to their multipath signals and each row has only a non-zero element with a value of 1.

6. The method of claim 5, wherein the converting the parameter estimation problem into the MMV block sparsity problem results in Y_tAnd block sparse signals, obtaining the estimated value of the parameters through a fast marginalization block sparse Bayesian learning algorithm, and rewriting (10) as:

whereinIs a quantized delay vector and

is M_Tk×M_Tk is an identity matrix;is the KR product;wherein P is_lIs a matrix P^TColumn l;

at this time, the process of the present invention,front half ofFor a known signal and depending on the subcarrier n, willObtaining an observation matrix Y as a row vector_t：

Wherein W is a sensing matrix with the nth behavior known signal The method is a block sparse matrix, and the block sparse matrix is called L-sparse because only L elements in the block sparse matrix are not zero, and L is called sparsity; at the moment, the parameter estimation problem is converted into a block sparsity problem of multiple measurement vectors, and the observation matrix is Y_tThe perceptual matrix is W and the block sparse matrix is

Sparse matrix of blocksDecomposed into according to a grid

Since the block sparse matrix is an L-sparse block sparse matrix, it is not zero only on the L (L belongs to L) th partition, so that the block sparse matrix can be obtained

Non-zero matrix in block sparse matrixExpressed as the number of transmitting array elements KThen it can be obtained:

by calculating B_l，kCross-correlation between rows and columns can yield an estimate of the angle of arrival or angle of departure,

Technical Field

The invention belongs to the technical field of communication, and relates to an uplink and downlink communication perception integrated method based on a block sparse Bayesian algorithm.

Background

In the present increasingly complex electromagnetic environment, the traditional single electronic countermeasure equipment is not the mainstream development direction, the research of the radar communication integrated system becomes more and more important, the electronic countermeasure equipment needs to have the communication function, the radar function, the signal processing function and the like at the same time, and at present, the application occasions needing the functions of detecting and positioning the radar and transmitting the communication and the message are more and more. Research on radar communication integrated systems relates to the field of radar electronic countermeasure and the field of communication electronic countermeasure, and the relation between a radar signal processing system and a communication signal processing system needs to be researched.

Since the concept of communication perception integration is proposed, various integrated design methods appear in sequence, the applications are most extensive on the basis of Linear Frequency Modulation (LFM) and Orthogonal Frequency Division Multiplexing (OFDM), and scholars propose an improved design method of fractional order Fourier transform on the basis of linear frequency modulation, so that the communication rate of integrated signals is improved; based on OFDM, researchers have combined Multiple Input Multiple Output (MIMO) with OFDM to improve the bandwidth utilization. However, currently, the research on parameter estimation of the integrated signal is few, and the test of the sensing performance of the integrated signal is not specific.

Disclosure of Invention

In order to solve the problems, the technical scheme of the invention comprises the following steps:

s1: modulating the baseband signal to different subcarriers by utilizing a multicarrier modulation technology, and distributing the modulated time domain signal to different base station remote radio frequency units;

s2: determination of the steering vector a of the transmitting and receiving antennas by means of a uniform linear array radar per remote radio unit_M(σ)；

S3: determining a frequency domain channel matrix H on the corresponding nth sub-carrier on the t OFDM symbol in a channel with L multipath signals through array response vectors of a receiving end and a transmitting end_n；

S4: determining the received and transmitted signals y of remote RF units in the uplink and downlink_n,tEstablishing a model after the quantitative delay of a sparse representation type on-grid direction-finding method

S5: converting the parameter estimation problem into an MMV block sparsity problem to obtain Y_tAnd block sparse signals, with fast edgesAnd obtaining the estimated value of the parameter by the edge block sparse Bayesian learning algorithm.

Preferably, the baseband signal is modulated to different subcarriers by using a multicarrier modulation technology, and the modulated time domain signal is distributed to different base station remote radio frequency units, in order to modulate the digital baseband signal to a data symbol by quadrature amplitude modulation, perform spatial precoding on the data symbol, provide power gain for a receiving end by precoding, and reduce the signal processing difficulty of the receiving end; then, performing inverse fast Fourier transform on the frequency domain sampling data, and converting the frequency domain sampling data into time domain data; the resulting composite time domain signal x (t) is then distributed to the respective remote radio units.

Preferably, said determination of the steering vector a of the transmitting and receiving antennas by means of a uniform linear array radar of each remote radio unit_M(σ), assuming that there is a planar wavefront in the signal propagation, the steering vector for a Uniform Linear Array (ULA) of M antennas is given by:

a_M(σ)＝[1,e^jπsinσ,…,e^{jπ(M-1)sinσ}]^T (1)

where M is the number of array antennas and σ is the angle of arrival or angle of departure.

Preferably, the frequency domain channel matrix H on the nth subcarrier corresponding to the tth OFDM symbol in the channel with L multipath signals is determined by the array response vectors of the receiving end and the transmitting end_nLet the angle of arrival or angle of departure of the first multipath be denoted as χ_lAnd σ_lThe time-domain baseband signal impulse response with M1 transmit antennas and M2 receive antennas is:

wherein, b_lIs the amplitude value, τ_lIs the propagation delay, f_D,lIs the associated propagation delay; converting the received signal to frequency domain for processing, the frequency domain channel matrix at the nth subcarrier of the t OFDM block is given byAnd (3) discharging:

wherein the content of the first and second substances,is the effective OFDM symbol period T, T_sIs the effective OFDM length T plus the cyclic prefix length T_p。

Preferably, the received and transmitted signals y of the remote radio frequency units in the uplink and downlink are determined_n,tEstablishing a model after the quantitative delay of a sparse representation type on-grid direction-finding methodTo assume that there are a total of Q remote radio units, for downlink sensing, each remote radio unit receives reflected downlink signals from itself and the other Q-1 remote radio units, and the received signal at the qth remote radio unit at the nth subcarrier of the tth OFDM block can be expressed as:

wherein x_q,n,tIs the transmitted signal of the qth remote radio unit at the nth subcarrier of the tth OFDM block, b_q,l，τ_q,l，χ_q,l，σ_q,lRespectively obtaining complex amplitude, propagation delay, arrival angle and departure angle of the qth long-distance radio frequency unit in the ith multipath; and wherein

x_n，t＝(x_1，n，t，…，x_O，n，t)^T (7)

Wherein U is a diagonal matrix of array element departure angles,orColumn of (1) isOr z_n,tIs noise;

similarly, the received signal of the remote rf unit when sensing uplink is:

wherein b is_k,l，τ_k,l，χ_k,l，σ_k,lAnd f_D,k,lRespectively obtaining complex amplitude, propagation delay, arrival angle, departure angle and Doppler frequency shift of the kth remote radio frequency unit in the ith multipath;

then considering the on-grid sparse model after delay quantization, and using K to represent a transmitting end in an uplink or a downlink, namely a mobile station or a long-distance radio frequency unit; by M_TThe number of array elements used for receiving signals; the number of array elements used for reception is denoted by M,

wherein, C_nIs a quantized delay diagonal matrix andand N is_PLess than or equal to gN; p is an Np x L rectangular array matrix that maps signals from the subscriber/remote radio units to their multipath signals and each row has only a non-zero element with a value of 1.

Preferably, the parameter estimation problem is converted into an MMV block sparsity problem to obtain Y_tAnd block sparse signals, obtaining the estimated value of the parameters through a fast marginalization block sparse Bayesian learning algorithm, and rewriting (10) as:

whereinIs a quantized delay vector and is thatThe identity matrix of (1);is the KR product;wherein P is_lIs a matrix P^TColumn l;

at this time, the process of the present invention,front half ofFor a known signal and depending on the subcarrier n, willObtaining an observation matrix Y as a row vector_t：

Wherein W is a sensing matrix with the nth behavior known signal The method is a block sparse matrix, and the block sparse matrix is called L-sparse because only L elements in the block sparse matrix are not zero, and L is called sparsity; at the moment, the parameter estimation problem is converted into a block sparsity problem of multiple measurement vectors, and the observation matrix is Y_tThe perceptual matrix is W and the block sparse matrix is

Sparse matrix of blocksDecomposed into according to a grid

Since the block sparse matrix is an L-sparse block sparse matrix, it is not zero only on the L (L belongs to L) th partition, so that the block sparse matrix can be obtained

Non-zero matrix in block sparse matrixAccording to the number of transmitting array elementsK is represented byThen it can be obtained:

by calculating B_l,kCross-correlation between rows and columns can yield an estimate of the angle of arrival or angle of departure,

the invention has at least the following beneficial effects: the invention provides a scheme for estimating sensing parameters by using a sparse Bayesian algorithm aiming at the radar sensing function of a communication sensing integrated signal, the scheme determines the transmitting and echo signals of the integrated signal through a frequency domain channel matrix between a base station and a mobile station and an array response vector of an antenna, and adopts an on-grid direction finding method to assume that the signal falls on a grid without error, thereby realizing the estimation of the incoming direction of the signal by reconstructing the sparse signal by directly using a sparse representation method. Then, parameters such as a departure angle, an arrival angle, a Doppler frequency shift and the like of the signal are estimated by using a block sparse Bayesian learning reconstruction algorithm. The final simulation results demonstrate that the estimation is quite robust and accurate with a sufficiently high received signal-to-noise ratio.

Drawings

FIG. 1 is a schematic diagram of a uniform linear array of an integrated uplink and downlink communication sensing method based on a block sparse Bayesian algorithm according to an embodiment of the present invention;

FIG. 2 is a schematic flow chart of an uplink and downlink communication perception integrated method based on a block sparse Bayesian algorithm according to an embodiment of the present invention;

FIG. 3 is a schematic diagram illustrating simulation of the uplink arrival angle of an integrated uplink and downlink communication sensing method based on a block sparse Bayesian algorithm according to an embodiment of the present invention;

fig. 4 is a schematic diagram illustrating simulation of downlink arrival angles of an uplink and downlink communication perception integrated method based on a block sparse bayesian algorithm according to an embodiment of the present invention.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, the present invention is described in further detail below with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are merely illustrative of the invention and are not intended to limit the invention.

On the contrary, the invention is intended to cover alternatives, modifications, equivalents and alternatives which may be included within the spirit and scope of the invention as defined by the appended claims. Furthermore, in the following detailed description of the present invention, certain specific details are set forth in order to provide a better understanding of the present invention. It will be apparent to one skilled in the art that the present invention may be practiced without these specific details.

Referring to fig. 1, a schematic of a Uniform Linear Array (ULA) of the present invention is shown. Wherein the triangles represent radio frequency units in a uniform linear array for transmitting and receiving signals; the arrow represents one of the received multipath signals; θ represents the angle of arrival AoA.

Referring to fig. 2, which is a schematic diagram of a design flow of the present invention, in view of a lack of the prior art in the aspect of integrated parameter sensing, the applicant proposes a scheme for estimating sensing parameters by using a block sparse bayesian algorithm for a radar sensing function of a communication sensing integrated signal, wherein the scheme determines transmission and echo signals of the integrated signal by using a frequency domain channel matrix between a base station and a mobile station and an array response vector of an antenna, and uses an on-grid direction finding method, i.e., we perform grid sampling on a source signal arrival direction, and assume that the signal falls on the grid without error, so that the estimation of the signal arrival direction can be realized by directly using a sparse representation method through reconstructing a sparse signal. Then, parameters such as a departure angle, an arrival angle, a Doppler frequency shift and the like of the signal are estimated by using a block sparse Bayesian learning reconstruction algorithm. The final simulation results demonstrate that the estimation is quite robust and accurate with a sufficiently high received signal-to-noise ratio.

The invention discloses an uplink and downlink communication perception integrated scheme based on a block sparse Bayesian algorithm, which comprises the following steps:

s1: the baseband signal is modulated onto different subcarriers using OFDM multicarrier modulation techniques. And the modulated time domain signals are distributed to different base station remote radio frequency units.

Firstly modulating a digital baseband signal into a data symbol through QAM modulation, then carrying out space pre-coding on the data symbol, and providing power gain (power gain) for a receiving end through pre-coding, and on the other hand, reducing the signal processing difficulty of the receiving end; then IFFT is carried out on the frequency domain sampling data, and the frequency domain sampling data is converted into time domain data; the resulting composite time domain signal x (t) is then distributed to the respective remote radio units.

S2: determining an array response vector a of transmit and receive antennas by a uniform linear array radar (ULA) of each remote radio unit_M(σ)。

Assuming a planar wavefront exists in the signal propagation, the steering vector for a Uniform Linear Array (ULA) of M antennas is given by:

a_M(σ)＝[1,e^jπsinσ,…,e^{jπ(M-1)sinσ}]^T (1)

where M is the number of array antennas and σ is the angle of arrival (AoA) or angle of departure (AoD).

S3: determining a frequency domain channel matrix H on an nth subcarrier corresponding to a t OFDM symbol in a channel with L multipath signals through array response vectors of a receiving end and a transmitting end_n。

Let us denote the angle of arrival (AoA) or angle of departure (AoD), respectively, of the first multipath as χ_lAnd σ_lThe time-domain baseband signal impulse response with M1 transmit antennas and M2 receive antennas is:

wherein b is_lIs the amplitude value, τ_lIs the propagation delay, f_D,lIs the associated propagation delay. The received signal is converted to the frequency domain for processing. The frequency domain channel matrix at the nth subcarrier for the tth OFDM block is given by:

whereinIs the effective OFDM symbol period T, T_sIs the effective OFDM length T plus the cyclic prefix length T_p。

S4: determining the received and transmitted signals y of remote RF units in the uplink and downlink_n,tEstablishing a model after the quantitative delay of a sparse representation type on-grid direction-finding method

Assuming that there are a total of Q remote radio units, each remote radio unit receives reflected downlink signals from itself and other Q-1 RRUs for downlink sensing. Its received signal at the qth remote radio unit at the nth subcarrier of the tth OFDM block can be expressed as:

wherein x_q,n,tIs the transmitted signal of the qth remote radio unit at the nth subcarrier of the tth OFDM block, b_q,l，τ_q,l，χ_q,l，σ_q,lRespectively obtaining complex amplitude, propagation delay, arrival angle and departure angle of the qth long-distance radio frequency unit in the ith multipath; and wherein

x_n，t＝(x_1，n，t，…，x_O，n，t)^T (7)

Where U is a diagonal matrix of array element departure angles,orColumn of (1) isOr z_n,tIs noise.

Similarly, the received signal of the remote rf unit when sensing uplink is:

wherein b is_k,l，τ_k,l，χ_k,l，σ_k,lAnd f_D,k,lRespectively obtaining complex amplitude, propagation delay, arrival angle, departure angle and Doppler frequency shift of the kth remote radio frequency unit in the ith multipath;

the on-grid sparse model after delay quantization is then considered, at this pointK denotes the transmitting end in uplink or downlink, i.e. the mobile station or the remote radio unit; by M_TThe number of array elements used for receiving signals; the number of array elements used for reception is denoted by M.

Wherein C is_nIs a quantized delay diagonal matrix andand N is_PLess than or equal to gN; p is an npxl rectangular array matrix that maps signals from the subscriber/remote radio units to their multipath signals and each row has only a non-zero element with a value of 1;

s5: converting the parameter estimation problem into an MMV block sparse problem to obtain an observation matrix Y_tAnd block sparse signals, and obtaining the estimated values of the parameters through a fast marginalization block sparse Bayes learning algorithm.

Rewrite (10) to:

whereinIs a quantized delay vector and is M_Tk×M_Tk is an identity matrix;is the KR product;wherein P is_lIs a matrix P^TColumn l;

at this time, the process of the present invention,front half ofFor a known signal and depending on the subcarrier n, willObtaining an observation matrix Y as a row vector_t：

Wherein W is a sensing matrix with the nth behavior known signal For a block sparse matrix, we call it L-sparse, L sparsity, since only L elements in the block sparse matrix are non-zero. At the moment, the parameter estimation problem is converted into a block sparsity problem of multiple measurement vectors, and the observation matrix is Y_tThe perceptual matrix is W and the block sparse matrix is

Sparse matrix of blocksDecomposed into according to a grid

Since the block sparse matrix is an L-sparse block sparse matrix, it is not zero only on the L (L belongs to L) th partition, so that the block sparse matrix can be obtained

Non-zero matrix in block sparse matrixExpressed as the number of transmitting array elements KThen it can be obtained:

by calculating B_l,kCross-correlation between rows and columns can result in an estimate of angle of arrival (AoA) or angle of departure (AoD).

Fig. 3 and 4 are diagrams illustrating uplink angle-of-arrival and downlink angle-of-arrival simulations, respectively, of the present invention. Assuming that the system has 4 Remote Radio Units (RRUs), the connections are provided to 4 users by multi-user MIMO. Each RRU has 4 antennas and each mobile station has 1 antenna. The carrier frequency is 2.35gHz, the signal bandwidth is 100mHz, and λ is c/2.35 gHz. The multipath signal for each RRU/MS is generated with a uniform distribution of offsets of-75, 75 degrees, distances of [50, 180] meters, and moving speeds of-40, 40 meters per second, simulating the reflected signal from the object. In fig. 3 and 4, a plus sign and a circle represent an estimated value and an actual value, respectively. It can be seen that the estimated and actual values of the arrival angles of multipath signals estimated by using the method are better in conformity with the conditions of 50 meters to 150 meters, and have errors above 150 meters but still be acceptable.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents and improvements made within the spirit and principle of the present invention are intended to be included within the scope of the present invention.