CN113965881A - Millimeter wave integrated communication and sensing method under shielding effect - Google Patents

Abstract

The invention provides a millimeter wave integrated communication and perception method under the shielding effect, which utilizes pilot signals or other known data sequences of the existing communication system to carry out perception. And establishing a probabilistic reasoning model based on a factor graph according to the relation among the received data, the distribution of the environmental scatterers and the shielding effect. And finally, solving a compressed sensing reconstruction problem based on a bilinear approximate message transmission method utilizing the shielding relation, thereby realizing the sensing of the environment. Compared with the existing environment perception reconstruction algorithm, the millimeter wave environment perception algorithm based on the occlusion effect obviously improves the accuracy of environment perception, and provides an efficient environment perception method for the design of a future perception communication integrated system.

Description

Millimeter-wave integrated communication and perception method under occlusion effect

technical field

The invention relates to the field of wireless communication, in particular to the design of an integrated system of perception and communication in the field of new generation wireless communication.

Background technique

Millimeter wave (mmWave) has become a research focus in the current wireless communication field due to its high bandwidth, high reliability, and high integration, combined with massive multiple-input multiple-output (MIMO) technology. With the rapid development of the wireless communication industry, considering the rapidly increasing number of connected devices and services, and the increasingly complex application scenarios of wireless mobile communication, the propagation environment of wireless communication signals is increasingly complex. At the same time, with the increasingly dense deployment of wireless communication base stations, and the receiver and sender have more powerful computing power and physical performance than before, the information processing capacity has been greatly improved, which puts forward higher requirements for the communication system. We not only hope that they can continue to provide original communication services, but also hope that communication devices can use their powerful processing capabilities to complete the perception of the environment. Specifically, in future wireless communication scenarios, new technologies such as smart cities, autonomous driving, and drone positioning require not only wireless broadband connections, but also perception of accurate environmental information, including but not limited to stationary or moving objects in the environment The location, shape, state, and electromagnetic properties of the background scatterers.

How to use wireless communication equipment for environmental perception based on the wireless communication architecture and realize the integration of perception and communication is an important research direction of the next generation wireless communication system. Wherein, when using the uplink for environment perception, the base station implements passive environment perception while performing data communication by receiving the uplink communication signal sent by the user. One of the major challenges in the design of sensing-communication integrated systems lies in the potentially large number of unknown variables in the environment, so the sparsity of the target environment itself should be exploited. For example, in a cellular communication network, buildings are sparsely distributed within the wireless network coverage. In addition, there is generally a blocking effect between scatterers, and the scatterers located closer to the user will block the electromagnetic signal, so that they cannot reach the scatterers farther away in the same propagation direction. Therefore, not all scatterers within the sensing range will affect the multipath channel of the same user, and mutual occlusion causes different users to face different scatterer environments, and multiple users need to perform joint sensing from multiple perspectives. At present, the existing environment perception algorithms do not take advantage of the relationship between the distribution of environmental scatterers and the occlusion effect, and do not take this problem into account, and the performance is poor when solving the imaging model with the occlusion effect.

To sum up, considering the millimeter wave environment perception problem and the inconsistency of user observation targets under the occlusion effect, how to jointly realize the separation of environmental information in the uplink data and the resolution of the occlusion effect has high research difficulty and practical significance.

SUMMARY OF THE INVENTION

The purpose of the present invention is to solve the problem of how a base station uses uplink data sent by multiple users to perform environment perception in an uplink wireless communication scenario. The present invention utilizes the pilot signal or other known data sequence of the existing communication system for sensing, separates the sending and receiving processing, is compatible with the existing communication system, and realizes the integration of sensing and communication. Considering that due to the occlusion effect between ambient scatterers, different users face different scatterer environments, an integrated communication and perception method of mmWave under occlusion effect is proposed.

The concrete technical scheme adopted in the present invention is as follows:

A millimeter wave integrated communication and perception method under occlusion effect, comprising the following steps:

1) In the T-th time slot, the base station receives a pilot sequence s signal of length L sent by all active users in the space;

2) After the base station receives the signal, based on a millimeter wave channel model considering the occlusion effect, the environment perception problem is converted into a compressed sensing reconstruction problem;

3) Based on the relative positional relationship between environmental scatterers, a model of the relationship between the distribution and occlusion of environmental scatterers is established;

4) Calculate the relationship between the received data, the distribution of environmental scatterers and the occlusion effect, and establish a probabilistic inference model based on factor graphs;

5) Combining the models obtained in steps 3) and 4), based on a bilinear approximation message passing method using occlusion relations, the compressive sensing reconstruction problem in step 3) is solved to realize the perception of the environment.

In one or more embodiments, in order to convert the environmental perception problem in step 2) into a compressive sensing reconstruction problem, it is necessary to analyze the problems caused by environmental scatterers on the basis of a millimeter-wave channel model considering the occlusion effect. The multipath channel H ^S is estimated and the environmental information in the channel is separated.

In one or more embodiments, the millimeter-wave channel model and compressive sensing reconstruction problem in step 2) are:

2.a) The environmental information is discretized, and the environmental information in the entire space is regarded as a point cloud. Each point in the point cloud represents the environmental information of small cubes with sizes _{l r} , _wr and hr around them. The cube is called a pixel; the length, width and height of the perceived scatterer environment are L _r , W _r and H _r respectively, then the number of point clouds in the space is N=L _r /l _r ×W _r /w _r ×H _r /h _r ; each pixel may be empty or there may be scatterers; we use a scattering coefficient x _n to represent the scattering coefficient of the small cube where the nth point cloud point is located, if the interior of the small cube is empty , then x _n =0; so the environmental information of the entire space can be represented by the following variables:

x=[x ₁ , x ₂ , . . . , x _N ] ^T ;

2.b) The system model is expressed as follows, multiple users in space share time-frequency resources, and on any frequency resource block, the signal received by the n- _th BS receiving antenna is expressed as

表示N_R个BS天线的接收信号，

表示N_u个用户发送长度为L码元的导频，w为噪声，

表示用户到接收天线直视路径的信道系数，

表示用户到空间点云位置的信道系数，

表示用户到空间点云位置的信道系数，且

其中

表示用户至空间点云位置的遮挡矩阵，当V^s1出现全零列时，表示所有用户都无法感知到该像素点，即该像素点在感知范围外，V^s2(n_R)∈{0，1}^N×1表示空间点云位置直射至接收天线的遮挡矩阵，当V^s2(n_R)出现零元素时，表示接收天线无法感知到该像素点，该像素点也在感知范围外，且V(n_R)＝V^s1(n_R)diag(V^s2(n_R))；in

represents the received signals of _NR BS antennas,

Indicates that Nu users send pilots of length L symbols, _w is noise,

represents the channel coefficient of the direct line of sight from the user to the receiving antenna,

represents the channel coefficient from the user to the position of the spatial point cloud,

represents the channel coefficient from the user to the spatial point cloud location, and

in

Represents the occlusion matrix from the user to the position of the spatial point cloud. When all zero columns appear in V ^s1 , it means that all users cannot perceive the pixel, that is, the pixel is outside the perception range, V ^s2 (n _R )∈{0, 1} ^N×1 represents the occlusion matrix that the spatial point cloud position directly hits the receiving antenna. When a zero element appears in V ^s2 (n _R ), it means that the receiving antenna cannot perceive the pixel, and the pixel is also outside the perception range, and V(n _R )=V ^s1 (n _R )diag(V ^s2 (n _R ));

2.c) Based on the above system model, the optimization problem of solving environmental information is expressed as,

In the communication process, the channel uncertainty mainly comes from the unknown distribution of environmental scatterers. For the direct-view channel H ^LOS , a known statistical model is used to describe it, and the user sends the known pilot frequency s to estimate the channel H ^S. This transforms the solving problem into a compressive sensing reconstruction problem:

H已知，需要在遮挡矩阵V未知的条件下求解环境信息x。It splices the channel estimates obtained by the N _R BS receiving antennas into a matrix, where H ^S is known,

H is known, and the environment information x needs to be solved under the condition that the occlusion matrix V is unknown.

In one or more embodiments, the model of the relationship between ambient scatterer distribution and occlusion in step 3) is:

There is a pixel point C with a scattering coefficient between pixel points A and B. We propose a model of the relationship between the distribution of environmental scatterers and occlusion for occlusion judgment. Let the position of pixel point A be the coordinate origin, then pixel points B and C The coordinates of , are expressed as b and c, respectively, and it is judged that pixel C occludes the direct line of sight between pixels A and B by satisfying the following three conditions;

The distance d of the direct path between pixel C and pixels A and B is less than the threshold β:

The angle between the vector b and the vector c is an acute angle:

b·c>0;

Pixel C is located between pixels A and B:

||c·b||<||b|| ² .

In one or more embodiments, a factor graph-based probabilistic inference model in step 4) is:

与x后验概率分解为：Will

The posterior probability with x is decomposed as:

和x，函数节点

p_x(x)和

并令H^S长度M＝N_uN_R，其中：A factor graph model is constructed according to the compressed sensing reconstruction problem in step 2) and the occlusion relationship in step 3); the factor graph contains variable nodes

and x, the function node

p _x (x) and

And let H ^S length M=N _u N _R , where:

where H ^S is the noised observation of z, and the noise is Gaussian white noise with variance σ ^w ; p _x (x) indicates that the prior distribution of x is a Gauss-Bernoulli distribution:

p _x (x)=(1-λ)δ(x)+λN(x; θ ^x , σ ^x );

表示环境信息x对信道的遮挡关系，具体如下：where λ represents the sparsity of the environmental information x,

Represents the occlusion relationship of the environmental information x to the channel, as follows:

In one or more embodiments, the bilinear approximation message passing method utilizing the occlusion relationship in step 5) is:

对于m＝1，2，...，M，n＝1，2，...，N，根据步骤4)中设定的p_x(x)和

生成x的估计的均值

和方差

的估计的均值

和方差

5.a) Initialize the algorithm parameters, let t be the number of iterations, and for m = 1, 2, ..., M, the residual mean is

For m = 1, 2, ..., M, n = 1, 2, ..., N, according to the p _x (x) and

generate the estimated mean of x

and variance

the estimated mean of

and variance

的观测均值

和方差

具体如下：5.b) For m=1,2,...,M, calculate

The observed mean of

and variance

details as follows:

in,

和方差

具体如下：5.c) For m = 1, 2, ..., M, compute the mean of the estimates of z _m

and variance

details as follows:

in,

where C is the normalization variable;

和方差

具体如下：5.d) For m = 1, 2, ..., M, calculate the mean of the residuals

and variance

details as follows:

的观测均值

和方差

对于n＝1，2，...，N，计算

的观测均值

和方差

具体如下：5.e) For m=1,2,...,M,n=1,2,...,N, compute

The observed mean of

and variance

For n=1,2,...,N, compute

The observed mean of

and variance

details as follows:

的估计均值

和方差

对于n＝1，2，...，N，计算x_n的观测均值

和方差

具体如下：5.f) For m=1,2,...,M,n=1,2,...,N, calculate

the estimated mean of

and variance

For n = 1, 2, ..., N, compute the observed mean of x _n

and variance

details as follows:

in,

得到环境信息x的估计值

5.g) Repeat step b) to step f) until the convergence condition is reached

Get an estimate of the environmental information x

The beneficial effects of the present invention are: in the uplink of wireless communication, in a scenario where data sent by users is used for environmental perception (for example, a multi-antenna base station uses uplink data sent by multiple single-antenna users for environmental perception), the present invention has the following advantages: The proposed millimeter-wave environment perception method under the occlusion effect, that is, a millimeter-wave integrated communication and perception method under the occlusion effect, uses the pilot signal of the existing communication system or other known data sequences for perception, which can be compared with the existing communication system. Compatible with communication systems to realize the integration of perception and communication. First, the present invention constructs a millimeter wave channel model considering the occlusion effect, and converts the environmental perception problem into a compressed sensing reconstruction problem. Then, based on the relative positional relationship between ambient scatterers, a model of the relationship between ambient scatterer distribution and occlusion is established. And according to the relationship between the received data, the distribution of environmental scatterers and the occlusion effect, a probabilistic inference model based on factor graph is established. Finally, based on a bilinear approximate message passing method using the occlusion relationship, the reconstruction problem of compressed sensing is solved to realize the perception of the environment. It uses multiple viewing angles of users to reconstruct the environment from multiple viewing angles. Compared with the existing environment perception reconstruction algorithm, the millimeter wave environment perception algorithm based on the occlusion effect of the present invention significantly improves the accuracy of environment perception, It provides an efficient environment perception method for the design of the future integrated system of perception and communication.

Description of drawings

Figure 1 is a schematic diagram of a millimeter-wave environment perception scene under the occlusion effect;

Figure 2 is a schematic diagram of the relationship between environmental scatterer distribution and occlusion;

Fig. 3 factor graph based on occlusion relationship;

Fig. 4 is the environment perception intuitive result diagram comparing the present invention with other compressed sensing reconstruction algorithms;

Fig. 5 is when comparing the present invention with other compressed sensing reconstruction algorithms, the relationship diagram of the number of users and the accuracy of environmental perception MSE;

FIG. 6 is a graph showing the relationship between the signal-to-noise ratio (SNR) and the environment perception accuracy (MSE) when the present invention is compared with other compressed sensing reconstruction algorithms.

Detailed ways

As shown in Figure 1, first we consider the following scenario, where a base station (BS) is deployed in an outdoor area and there are multiple active users (UE). In an uplink communication scenario, multiple single-antenna users send uplink communication signals to the BS simultaneously. The transmitted signal is scattered by the scatterer and multipath propagated to the BS for reception. There is an occlusion effect between scatterers in the environment. As shown in the figure above, the signal sent by user 1 is only scattered by target scatterers 1 and 3 and then received by the AP, while target scatterer 2 does not affect this process.

An embodiment of the present invention provides a millimeter-wave integrated communication and sensing method under occlusion effect, which includes the following steps:

1) In the T-th time slot, the base station receives a pilot sequence s signal of length L sent by all active users in the space;

2) After the base station receives the signal, based on a millimeter wave channel model considering the occlusion effect, the environment perception problem is converted into a compressed sensing reconstruction problem;

Among them, in an embodiment of the present invention, in order to convert the environmental perception problem in step 2) into a compressive sensing reconstruction problem, it is necessary to use a millimeter wave channel model considering the occlusion effect. The multipath channel H ^S is estimated, and the environmental information in the channel is separated.

Specifically, in an embodiment of the present invention, the millimeter wave channel model and the compressive sensing reconstruction problem described in step 2) are:

2.a) Discretize the environmental information, regard the environmental information in the whole space as a point cloud, each point in the point cloud represents the environmental information of the small cubes around it with sizes _{l r} , _wr and hr , these Small cubes are called pixels. The length, width and height of the perceived scatterer environment are L _r , W _r and H _r respectively, then the number of point clouds in the space is N= _{L r} /l _r ×W _r / _wr ×H _r /hr . The interior of each pixel may be empty, or there may be scatterers. We use a scattering coefficient x _n to represent the scattering coefficient of the small cube where the nth point cloud point is located. If the interior of the small cube is empty, then x _n =0. Therefore, the environmental information of the entire space can be represented by the following variables:

x=[x ₁ , x ₂ , . . . , x _N ] ^T ;

2.b) Express the system model as follows. Multiple users in the space share time-frequency resources, and on any frequency resource block, the signal received by the n _Rth BS receiving antenna is expressed as:

表示N_R个BS天线的接收信号，

表示N_u个用户发送长度为L码元的导频，w为噪声。

表示用户到接收天线直视路径的信道系数，

表示用户到空间点云位置的信道系数，

表示用户到空间点云位置的信道系数，且

其中

表示用户至空间点云位置的遮挡矩阵，当V^s1出现全零列时，表示所有用户都无法感知到该像素点，即该像素点在感知范围外。V^s2(n_R)∈{0，1}^N×1表示空间点云位置直射至接收天线的遮挡矩阵，当V^s2(n_R)出现零元素时，表示接收天线无法感知到该像素点，该像素点也在感知范围外，且V(n_R)＝V^s1(n_R)diag(V^s2(n_R))。in

represents the received signals of _NR BS antennas,

Indicates that Nu users send pilots with a length of L symbols, and _w is noise.

represents the channel coefficient of the direct line of sight from the user to the receiving antenna,

represents the channel coefficient from the user to the position of the spatial point cloud,

represents the channel coefficient from the user to the spatial point cloud location, and

in

It represents the occlusion matrix from the user to the position of the spatial point cloud. When all zero columns appear in V ^s1 , it means that all users cannot perceive the pixel, that is, the pixel is outside the perception range. V ^s2 (n _R )∈{0,1} ^N×1 represents the occlusion matrix that the spatial point cloud position directly hits the receiving antenna. When a zero element appears in V ^s2 (n _R ), it means that the receiving antenna cannot perceive the pixel, This pixel is also outside the perception range, and V(n _R )=V ^s1 (n _R )diag(V ^s2 (n _R )).

2.c) Based on the above system model, the optimization problem of solving environmental information is expressed as,

In the communication process, the channel uncertainty mainly comes from the unknown distribution of environmental scatterers, and a known statistical model is used to describe the direct-view channel H ^LOS . Therefore, the user sends the known pilot frequency s to estimate the channel H ^S , thus transforming the solving problem into a compressed sensing reconstruction problem, namely:

H已知。需要在遮挡矩阵V未知的条件下求解环境信息x。It splices the channel estimates obtained by the N _R BS receiving antennas into a matrix, where H ^S is known,

H is known. The environment information x needs to be solved under the condition that the occlusion matrix V is unknown.

3) Based on the relative positional relationship between environmental scatterers, a model of the relationship between the distribution and occlusion of environmental scatterers is established;

Specifically, in an embodiment of the present invention, the model of the relationship between ambient scatterer distribution and occlusion in step 3) is:

As shown in Figure 2, there is a pixel C with a scattering coefficient between pixels A and B. We propose a model of the relationship between the distribution of ambient scatterers and occlusion for occlusion judgment. Let the position of pixel A be the origin of coordinates, then the coordinates of pixels B and C are represented as b and c, respectively. By satisfying the following three conditions, it is determined that the pixel point C occludes the direct view path between the pixel points A and B.

The distance d of the direct path between the pixel point C and the pixel points A and B is less than the threshold β,

The angle between the vector b and the vector c is an acute angle,

b·c>0;

Pixel C is located between pixels A and B,

||c·b||<||b|| ² ;

4) Calculate the relationship between the received data, the distribution of environmental scatterers and the occlusion effect, and establish a probabilistic inference model based on factor graphs.

Specifically, in an embodiment of the present invention, a factor graph-based probabilistic inference model in step 4) is as follows.

与x后验概率分解为，Will

The posterior probability with x is decomposed into,

和x，函数节点

p_x(x)和

为简洁表达，令H^S长度M＝N_uN_R。As shown in Figure 3, a factor graph model is constructed according to the compressed sensing reconstruction problem in step 2) and the occlusion relationship in step 3). Factor graph contains variable nodes

and x, the function node

p _x (x) and

For brevity, let ^HS length M=N _u N _R .

in

Denote that H ^S is the noised observation of z, and the noise is white Gaussian noise with variance σ ^w . p _x (x) indicates that the prior distribution of x is a Gauss-Bernoulli distribution,

p _x (x)=(1-λ)δ(x)+λN(x; θ ^x , σ ^x );

表示环境信息x对信道的遮挡关系，where λ represents the sparsity of the environmental information x.

represents the occlusion relationship of the environmental information x to the channel,

5) Combining the models obtained in steps 3) and 4), based on a bilinear approximation message passing method using occlusion relations, the compressive sensing reconstruction problem in step 3) is solved to realize the perception of the environment.

Specifically, in an embodiment of the present invention, the bilinear approximation message passing method using the occlusion relationship in this step is:

对于m＝1，2，...，M，n＝1，2，...，N，根据步骤4)中设定的p_x(x)和

生成x的估计的均值

和方差

的估计的均值

和方差

5.a) Initialize the algorithm parameters. Let t be the number of iterations, and for m = 1, 2, ..., M, the residual mean is

For m = 1, 2, ..., M, n = 1, 2, ..., N, according to the p _x (x) and

generate the estimated mean of x

and variance

the estimated mean of

and variance

的观测均值

和方差

5.b) For m=1,2,...,M, calculate

The observed mean of

and variance

in,

和方差

5.c) For m = 1, 2, ..., M, compute the mean of the estimates of z _m

and variance

in,

C is the normalization variable.

和方差

5.d) For m = 1, 2, ..., M, calculate the mean of the residuals

and variance

的观测均值

和方差

对于n＝1，2，...，N，计算

的观测均值

和方差

5.e) For m=1,2,...,M,n=1,2,...,N, compute

The observed mean of

and variance

For n=1,2,...,N, compute

The observed mean of

and variance

的估计均值

和方差

对于n＝1，2，...，N，计算x_n的观测均值

和方差

5.f) For m=1,2,...,M,n=1,2,...,N, calculate

the estimated mean of

and variance

For n = 1, 2, ..., N, compute the observed mean of x _n

and variance

in

得到环境信息x的估计值

5.g) Repeat step b) to step f) until the convergence condition is reached

Get an estimate of the environmental information x

It can be seen from the computer simulation that as shown in Figure 4, the GAMP algorithm completely ignores the occlusion effect. The Bilinear GAMP algorithm regards the occlusion matrix V and the environmental information x as independent variables, and does not use the occlusion relationship to solve them separately. Compared with the first two algorithms, the millimeter wave environment perception algorithm based on the occlusion effect of the present invention significantly improves the accuracy of environment perception. FIG. 5 shows that with the increase of the number of users, the environment perception effect of the method of the present invention is gradually improved and is superior to the existing algorithm. FIG. 5 shows that with the increase of the signal-to-noise ratio, the environmental perception effect of the method of the present invention is gradually improved and is superior to the existing algorithm.

To sum up, in the uplink of wireless communication, in a scenario where data sent by users is used for environmental perception (for example, a multi-antenna base station uses uplink data sent by multiple single-antenna users for environmental perception), in this embodiment of the present invention, The proposed millimeter-wave environment perception method under the occlusion effect, that is, a millimeter-wave integrated communication and perception method under the occlusion effect, uses the pilot signal of the existing communication system or other known data sequences for perception, which can be compared with the existing communication system. Compatible with communication systems to realize the integration of perception and communication. First, in the embodiment of the present invention, a millimeter-wave channel model considering the occlusion effect is constructed, and the environment perception problem is converted into a compressed sensing reconstruction problem. Then, based on the relative positional relationship between ambient scatterers, a model of the relationship between ambient scatterer distribution and occlusion is established. And according to the relationship between the received data, the distribution of environmental scatterers and the occlusion effect, a probabilistic inference model based on factor graph is established. Finally, based on a bilinear approximate message passing method using the occlusion relationship, the reconstruction problem of compressed sensing is solved to realize the perception of the environment. It uses multiple user viewing angles to reconstruct the environment from multiple perspectives. Compared with the existing environment perception reconstruction algorithm, the millimeter wave environment perception algorithm based on the occlusion effect in the embodiment of the present invention significantly improves the accuracy of environment perception. It provides an efficient environment perception method for the design of the integrated system of perception and communication in the future.

The above-mentioned embodiments are used to illustrate the present invention, but not to limit the present invention. Any modification or change made to the present invention within the spirit of the present invention and the protection scope of the claims shall fall within the protection scope of the present invention.

Claims (6)

1. a millimeter-wave integrated communication and perception method under occlusion effect, is characterized in that, comprises the steps: 1) In the T-th time slot, the base station receives a pilot sequence s signal of length L sent by all active users in the space; 2) After the base station receives the signal, based on a millimeter wave channel model considering the occlusion effect, the environment perception problem is converted into a compressed sensing reconstruction problem; 3) Based on the relative positional relationship between environmental scatterers, a model of the relationship between the distribution and occlusion of environmental scatterers is established; 4) Calculate the relationship between the received data, the distribution of environmental scatterers and the occlusion effect, and establish a probabilistic inference model based on factor graphs; 5) Combining the models obtained in steps 3) and 4), based on a bilinear approximation message passing method using occlusion relations, the compressive sensing reconstruction problem in step 3) is solved to realize the perception of the environment. 2. the millimeter-wave integrated communication and perception method under the occlusion effect according to claim 1, is characterized in that, in order to convert the environmental perception problem in step 2) into the compressive sensing reconstruction problem, need to consider occlusion based on a kind of On the basis of the millimeter-wave channel model of the effect, the multipath channel H ^S caused by the environmental scatterers is estimated, and the environmental information in the channel is separated. 3. the millimeter-wave integrated communication and sensing method under the occlusion effect according to claim 1 and 2, is characterized in that, the millimeter-wave channel model in step 2) and compressive sensing reconstruction problem are: 2.a) The environmental information is discretized, and the environmental information in the entire space is regarded as a point cloud. Each point in the point cloud represents the environmental information of small cubes with sizes _{l r} , _wr and hr around them. The cube is called a pixel; the length, width and height of the perceived scatterer environment are L _r , W _r and H _r respectively, then the number of point clouds in the space is N=L _r /l _r ×W _r /w _r ×H _r /h _r ; each pixel may be empty or there may be scatterers; we use a scattering coefficient x _n to represent the scattering coefficient of the small cube where the nth point cloud point is located, if the interior of the small cube is empty , then x _n =0; so the environmental information of the entire space can be represented by the following variables: x=[x ₁ , x ₂ , . . . , x _N ] ^T ; 2.b) The system model is expressed as follows, multiple users in space share time-frequency resources, and on any frequency resource block, the signal received by the n _Rth BS receiving antenna is expressed as:

in

represents the received signals of _NR BS antennas,

Indicates that Nu users send pilots of length L symbols, _w is noise,

represents the channel coefficient of the direct line of sight from the user to the receiving antenna,

represents the channel coefficient from the user to the position of the spatial point cloud,

represents the channel coefficient from the user to the spatial point cloud location, and

in

represents the occlusion matrix from the user to the position of the spatial point cloud, when

When an all-zero column appears, it means that all users cannot perceive the pixel, that is, the pixel is outside the perception range.

represents the occlusion matrix that the position of the spatial point cloud directly hits the receiving antenna, when

When a zero element appears, it means that the receiving antenna cannot perceive the pixel, the pixel is also outside the perception range, and

2.c) Based on the above system model, the optimization problem of solving environmental information is expressed as,

In the communication process, the channel uncertainty mainly comes from the unknown distribution of environmental scatterers. For the direct-view channel H ^LOS , a known statistical model is used to describe it, and the user sends the known pilot frequency s to estimate the channel H ^S. This transforms the solving problem into a compressive sensing reconstruction problem:

It splices the channel estimates obtained by the N _R BS receiving antennas into a matrix, where H ^S is known,

H is known, and the environment information x needs to be solved under the condition that the occlusion matrix V is unknown. 4. The millimeter-wave integrated communication and perception method under occlusion effect according to claim 1 and 2, is characterized in that, the model of environmental scatterer distribution and occlusion relation in step 3) is: There is a pixel point C with a scattering coefficient between pixel points A and B. We propose a model of the relationship between the distribution of environmental scatterers and occlusion for occlusion judgment. Let the position of pixel point A be the coordinate origin, then pixel points B and C The coordinates of , are expressed as b and c, respectively, and it is judged that pixel C occludes the direct line of sight between pixels A and B by satisfying the following three conditions; The distance d of the direct path between pixel C and pixels A and B is less than the threshold β:

The angle between the vector b and the vector c is an acute angle: b·c>0; Pixel C is located between pixels A and B: ||c·b||<||b|| ² . 5. a kind of millimeter wave environment perception method according to claim 3, is characterized in that, a kind of probability inference model based on factor graph in step 4) is: Will

The posterior probability with x is decomposed as:

A factor graph model is constructed according to the compressed sensing reconstruction problem in step 2) and the occlusion relationship in step 3); the factor graph contains variable nodes

and x, the function node

and

And let H ^S length M=N _u N _R , where:

where H ^S is the noised observation of z, and the noise is Gaussian white noise with variance σ ^w ; p _x (x) indicates that the prior distribution of x is a Gauss-Bernoulli distribution: p _x (x)=(1-λ)δ(x)+λN(x; θ ^x , σ ^x ); where λ represents the sparsity of the environmental information x,

Represents the occlusion relationship of the environmental information x to the channel, as follows:

6. millimeter wave integrated communication and perception method under occlusion effect according to claim 5, is characterized in that, the bilinear approximate message passing method utilizing occlusion relation in step 5) is: 5.a) Initialize the algorithm parameters, let t be the number of iterations, and for m = 1, 2, ..., M, the residual mean is

For m = 1, 2, ..., M, n = 1, 2, ..., N, according to the p _x (x) and

generate the estimated mean of x

and variance

the estimated mean of

and variance

5.b) For m=1,2,...,M, calculate

The observed mean of

and variance

details as follows:

in,

5.c) For m = 1, 2, ..., M, compute the mean of the estimates of z _m

and variance

details as follows:

in,

where C is the normalization variable; 5.d) For m = 1, 2, ..., M, calculate the mean of the residuals

and variance

details as follows:

5.e) For m=1,2,...,M,n=1,2,...,N, compute

The observed mean of

and variance

For n=1,2,...,N, compute

The observed mean of

and variance

details as follows:

5.f) For m=1,2,...,M,n=1,2,...,N, calculate

The estimated mean of

and variance

For n = 1, 2, ..., N, compute the observed mean of x _n

and variance

details as follows:

in,

5.g) Repeat step b) to step f) until the convergence condition is reached

Get an estimate of the environmental information x

Images