A passive intelligent reflector-assisted cloud access network communication perception method

Abstract

The present invention discloses a passive intelligent reflector-assisted cloud access network communication perception method. This method enables interference management within the geographical coverage of distributed remote radio heads (RRHs) through the centralized information processing achieved through the cloud radio access network (C‑RAN). By jointly processing baseband signals at the BBU pool, an effective means of implementing network multi-input multi-output (MIMO) in heterogeneous wireless networks is provided. At the same time, the communication and radar perception signals of the dual-function radar system are enhanced by the passive intelligent reflector. The passive intelligent reflector and the beamforming at the RRH are jointly designed so that the reflected signal of the passive intelligent reflector is aligned with the target. This method can obtain the expected reflected beam of the reflector while ensuring the user's signal-to-interference-noise ratio.

Description

Technical Field

[0001] The present invention relates to the technical field of communication perception integration, and in particular to a passive intelligent reflector-assisted cloud access network communication perception method. Background Art

[0002] Wireless communications have gradually evolved from 1G to 5G, and research and development of 6G technology is now in full swing. The demand for spectrum resources is increasing. The limited availability of radio spectrum and the pursuit of ever-increasing communication speeds have led to overlap between radar and communications bands. Currently, military radars utilize many frequency bands below 10 GHz, and spectrum congestion is becoming a serious issue limiting wireless communication throughput. To address this congestion, proposals have been made to allow wireless communications and radar systems to share spectrum, allowing both functions to operate simultaneously over the same wide frequency band. Integrated communication and sensing technologies offer sensing and communication capabilities through a single platform, using the same waveform. Consequently, compared to systems that perform sensing and communication on separate platforms, they utilize spectrum efficiently and offer smaller device size and lower hardware costs. Summary of the Invention

[0003] The present invention uses a comprehensive beamforming method and communication signals for sensing. At the same time, the use of intelligent reflective surface technology can sense some obscured targets, greatly increasing the range of perceptible objects.

[0004] The technical solutions of the present invention are as follows:

[0005] A passive intelligent reflector-assisted cloud access network communication perception method is characterized in that: the BBU pool communicates with K single-antenna users through L multi-antenna RRHs and detects the required targets. A point-to-point wired fronthaul link is used between the BBU pool and the RRHs. Due to the obstruction between the RRHs and the detection targets, the signal is reflected to the target to be detected by the passive intelligent reflector, and the signal is sent to the user through the reflector to enhance communication. The transmit beamforming matrix of the BBU pool and RRH, the phase shift matrix of the reflector, and the covariance matrix of the forward quantization noise are jointly optimized, specifically including the following steps:

[0006] 1.1) In a dual-function radar communication system in a cloud wireless access network assisted by a passive intelligent reflector, it transmits communication symbols to K single-antenna users and simultaneously transmits radar detection waveforms to targets of interest. The system includes 1 baseband unit (BBU), L radar receiver hubs (RRHs), and M radar signal reception (IRS) units to assist in enhancing communication and radar perception. The BBU and RRHs are wired together. The RRHs use ULA array antennas with N antennas. R .

[0007] 1.2) The BBU pool first beamforms all user signals c transmitted to the lth RRH, and then superimposes them with the radar sensing signal x0. The resulting signal is: is the signal corresponding to the lth RRH, W c is the communication beamforming matrix, c~CN(0,I K ), x0~CN(0,R0). Since the capacity of the wireless fronthaul link is limited, the signal It is quantized and compressed before being transmitted to the RRH. By adopting the Gaussian test channel model, the compressed signal can be expressed as: x = W c c+x0+q

[0008] Where R0 is the covariance matrix of the perception signal; q~CN(0,Ω), Ω is the covariance matrix of the quantization noise, where is the covariance matrix of the compressed signal.

[0009] 1.3) The signal sent by the first RRH can be expressed as:

[0010]

[0011] q l ~CN(0,Ω l,l ) is independent of Quantization noise, Ω=diag({Ω l,l}),matrix The representative matrix intercepts the beamforming part corresponding to the RRH from the forward compressed signal of the BBU. The specific structure of the matrix is ​​[(l-1)N R +1,lN R ] The rows form a dimension of N R ×N R The unit matrix of RRH is , and the rest are all zero. Here we consider the single antenna power equation constraint of RRH

[0012] 1.4) The signal received by user k is: y k =(h k +g k ΨG)x+n k =z k x+n k

[0013] in represents the additive white Gaussian noise (AWGN) of the kth user, h k is the channel state from RRH to user k, g kis the channel state from the reflection surface to user k, Ψ is the phase reflection matrix of the reflection surface, G is the channel state from RRH to the reflection surface, z k =h k +g k ΨG.

[0014] Furthermore, in order to minimize the mean square error between the ideal beam and the beam of the smart reflector, the dual-function radar communication system of the cloud wireless access network assisted by the passive smart reflector needs to be performed under the conditions of user SINR limitation and forward compression rate limitation. Finally, the covariance matrix R of the RRH transmission signal, the covariance matrix Ω of the forward quantization noise, the phase reflection matrix Ψ of the smart reflector and the beamforming matrix W of the communication signal are calculated. c Optimize.

[0015] The optimization problem can be expressed as:

[0016]

[0017] in is the spatial frequency steering vector, the IRS has a linearly spaced isotropic array with element spacing d and wavelength λ. (d) is the linear user SINR constraint, Γ is the threshold, and w i W c The i-th column of (e) is the forward compression rate constraint (the original constraint is amplified and the amplified condition is required to meet the constraint), and the constraint (f) Indicates that the matrix is ​​positive semidefinite.

[0018] Furthermore, the optimization problem needs to be decomposed into two sub-problems and solved alternately.

[0019] 3.1) For the first sub-problem, we first need to fix Ψ in the optimization problem to optimize the RRH beamforming matrix W c and the forward quantization noise covariance matrix Ω.

[0020] 3.1.1) For constraint (b), define Since the solution of R0 to the subproblem can be omitted, we can get:

[0021]

[0022] 3.1.2) Rewrite the constraint (d) in the optimization problem as follows:

[0023]

[0024] 3.1.3) Constraint (e) can be further transformed into the following approximate constraint:

[0025]

[0026] S l is an auxiliary variable, which is required to be a semi-positive definite matrix. , the equality of the above equation holds. 3.1.4) First, define The optimization problem can be transformed into:

[0027]

[0028] This problem can be directly optimized to obtain the optimal w k and R.

[0029] 3.2) For the second sub-problem, we first need to fix the optimization variables R and w k and Ω to optimize the reflection surface matrix Ψ.

[0030] 3.2.1) User k’s achievable rate R k It can be written as:

[0031]

[0032] where w k Auxiliary constant variables,

[0033] u k is a linear estimator for decoding the signal received by user k.

[0034] Constraint (d) can be expressed as:

[0035]

[0036] For a fixed transmit beamformer, the optimal solution for a known linear receiver is given by:

[0037]

[0038] 3.2.2) Each iteration updates w k and R, so that the constraints meet the constant constraint requirements. Definition The constraint condition in 3.2.1) can be finally rewritten as The convex form of :

[0039]

[0040] log(w k )-Tr(ψ H a k B k ψ-2ψ H d k +c)+1≥log(1+Γ)

[0041] in:

[0042] 3.2.3) Rewrite about Objective function:

[0043]

[0044] Where Q(θ)=diag(a(θ) H )GRG H diag(a(θ)), Using auxiliary variable t j There are constraints on the optimization objective function:

[0045]

[0046] 3.2.4) Use the semi-positive definite relaxation method to directly Optimize and obtain the result ψ. The objective function is calculated by Taylor's formula in ψ t Expand to get in ψ t The initial value of is obtained by satisfying the reflection condition of the passive intelligent reflective surface.

[0047] 3.2.5) Finally, we get the optimized form of sub-problem 2 in SDR form:

[0048]

[0049] ψ H Q(θ)ψ-αd(θ)≤t j ,t j ≥0

[0050]

[0051] 3.2.6) According to the optimization in step 3.2.5) Perform Gaussian randomization a sufficient number of times to obtain the best performance ψ, and determine whether the optimization has converged. If it has not converged, update the variable S according to 3.1.3). l , optimize the sub-problem 1 in 3.1.4), update the variable w k and R k , and then update the relevant variables in sub-problem 2, and optimize the sub-problem in 3.2.5) until convergence.

[0052] The present invention has the following beneficial effects: A dual-function radar communication system in a cloud radio access network assisted by a passive intelligent reflector uses communication signals for both communicating with users and detecting targets of interest, effectively alleviating spectrum congestion. The cloud radio access network improves computational and energy efficiency. Signals reflected by the IRS can be constructively added to signals from other paths to enhance the power of the desired signal at the receiver, or destructively cancel undesired signals such as co-channel interference. This system can be densely deployed with scalable costs and low energy consumption, eliminating the need for complex interference management between passive IRSs. BRIEF DESCRIPTION OF THE DRAWINGS

[0053] Figure 1 Comparison of integrated beamforming schemes with different numbers of passive intelligent reflector units;

[0054] Figure 2 This is a model diagram of the dual-function radar communication system of the cloud wireless access network assisted by the passive intelligent reflector in the present invention. DETAILED DESCRIPTION

[0055] The present invention will be further described below with reference to the accompanying drawings.

[0056] like Figure 1-2 As shown, a passive intelligent reflector-assisted cloud access network communication perception method is described. The BBU pool communicates with K single-antenna users through L multi-antenna RRHs and detects the required targets. A point-to-point wired fronthaul link is used between the BBU pool and the RRHs. Due to the obstruction between the RRHs and the detection targets, the signal is reflected to the target to be detected through the passive intelligent reflector, and the signal is sent to the user through the reflector to enhance communication. The transmit beamforming matrix of the BBU pool and RRH, the phase shift matrix of the reflector, and the covariance matrix of the forward quantization noise are jointly optimized, specifically including the following steps:

[0057] 1.1) In a dual-function radar communication system in a cloud wireless access network assisted by a passive intelligent reflector, it transmits communication symbols to K single-antenna users and simultaneously transmits radar detection waveforms to targets of interest. The system includes 1 baseband unit (BBU), L radar receiver hubs (RRHs), and M radar signal reception (IRS) units to assist in enhancing communication and radar perception. The BBU and RRHs are wired together. The RRHs use ULA array antennas with N antennas. R .

[0058] 1.2) The BBU pool first beamforms all user signals c transmitted to the lth RRH, and then superimposes them with the radar sensing signal x0. The resulting signal is: is the signal corresponding to the lth RRH, W cis the communication beamforming matrix, c~CN(0,I K ), x0~CN(0,R0). Since the capacity of the wireless fronthaul link is limited, the signal It is quantized and compressed before being transmitted to the RRH. By adopting the Gaussian test channel model, the compressed signal can be expressed as: x = W c c+x0+q

[0059] Where R0 is the covariance matrix of the perception signal; q~CN(0,Ω), Ω is the covariance matrix of the quantization noise, where is the covariance matrix of the compressed signal.

[0060] 1.3) The signal sent by the first RRH can be expressed as:

[0061]

[0062] q l ~CN(0,Ω l,l ) is independent of Quantization noise, Ω=diag({Ω l,l}),matrix The representative matrix intercepts the beamforming part corresponding to the RRH from the forward compressed signal of the BBU. The specific structure of the matrix is ​​[(l-1)N R +1,lN R ] The rows form a dimension of N R ×N R The unit matrix of RRH is , and the rest are all zero. Here we consider the single antenna power equation constraint of RRH

[0063] 1.4) The signal received by user k is: y k =(h k +g k ΨG)x+n k =z k x+n k

[0064] in represents the additive white Gaussian noise (AWGN) of the kth user, h k is the channel state from RRH to user k, g k is the channel state from the reflection surface to user k, Ψ is the phase reflection matrix of the reflection surface, G is the channel state from RRH to the reflection surface, z k =h k +g k ΨG.

[0065] In order to minimize the mean square error between the ideal beam and the beam of the smart reflector, the dual-function radar communication system of the cloud wireless access network assisted by the passive smart reflector needs to be performed under the conditions of user SINR limitation and forward compression rate limitation. Finally, the covariance matrix R of the RRH transmission signal, the covariance matrix Ω of the forward quantization noise, the phase reflection matrix Ψ of the smart reflector, and the beamforming matrix W of the communication signal are calculated. c Optimize.

[0066] In order to minimize the mean square error between the ideal beam and the beam of the smart reflector, the dual-function radar communication system of the cloud wireless access network assisted by the passive smart reflector needs to be performed under the conditions of user SINR limitation and forward compression rate limitation. Finally, the covariance matrix R of the RRH transmission signal, the covariance matrix Ω of the forward quantization noise, the phase reflection matrix Ψ of the smart reflector and the beamforming matrix W of the communication signal are calculated. c Optimize

[0067] The optimization problem is expressed as:

[0068]

[0069] Where J is the number of sampling angles, is the spatial frequency steering vector, an IRS with a linearly spaced isotropic array, element spacing d, wavelength λ, α is the scaling factor, d(θ) is the ideal beam; (d) is the linear user SINR constraint, Γ is the threshold, w i W c The i-th column of represents the covariance of Gaussian white noise; (e) is the forward compression rate constraint, which amplifies the original constraint and requires that the amplified condition meets the forward transmission limit capacity C l , in the constraint condition (f) Indicates that the matrix is ​​positive semidefinite.

[0070] The above optimization problem needs to be decomposed into two sub-problems and solved alternately.

[0071] 3.1) For the first sub-problem, we first need to fix Ψ in the optimization problem to optimize the RRH beamforming matrix W c and the forward quantization noise covariance matrix Ω.

[0072] 3.1.1) For constraint (b), define Since the solution of R0 to the subproblem can be omitted, we can get:

[0073]

[0074] 3.1.2) Rewrite the constraint (d) in the optimization problem as follows:

[0075]

[0076] 3.1.3) Constraint (e) can be further transformed into the following approximate constraint:

[0077]

[0078] S l is an auxiliary variable, which is required to be a semi-positive definite matrix. , the equality of the above equation holds. 3.1.4) First, define The optimization problem can be transformed into:

[0079]

[0080] This problem can be directly optimized to obtain the optimal w k and R.

[0081] 3.2) For the second sub-problem, we first need to fix the optimization variables R and w k and Ω to optimize the reflection surface matrix Ψ.

[0082] 3.2.1) User k’s achievable rate R k It can be written as:

[0083]

[0084] where w k Auxiliary constant variables,

[0085] u k is a linear estimator for decoding the signal received by user k.

[0086] Constraint (d) can be expressed as:

[0087]

[0088] For a fixed transmit beamformer, the optimal solution for a known linear receiver is given by:

[0089]

[0090] 3.2.2) Each iteration updates w k and R, so that the constraints meet the constant constraint requirements. Definition The constraint condition in 3.2.1) can be finally rewritten as The convex form of :

[0091]

[0092] log(w k )-Tr(ψ H a k B k ψ-2ψ H d k +c)+1≥log(1+Γ)

[0093] in:

[0094] 3.2.3) Rewrite about Objective function

[0095]

[0096] Where Q(θ)=diag(a(θ) H )GRG H diag(a(θ)), Using auxiliary variable t j There are constraints on the optimization objective function:

[0097]

[0098] 3.2.4) Use the semi-positive definite relaxation method to directly Optimize and obtain the result ψ. The objective function is calculated by Taylor's formula in ψ t Expand to get in ψ t The initial value of is obtained by satisfying the reflection condition of the passive intelligent reflective surface.

[0099] 3.2.5) Finally, we get the optimized form of sub-problem 2 in SDR form:

[0100]

[0101] ψ H Q(θ)ψ-αd(θ)≤t j ,t j ≥0

[0102]

[0103] 3.2.6) According to the optimization in step 3.2.5) Perform Gaussian randomization a sufficient number of times to obtain the best performance ψ, and determine whether the optimization has converged. If it has not converged, update the variable S according to 3.1.3). l , optimize the sub-problem 1 in 3.1.4), update the variable w k and R k, and then update the relevant variables in sub-problem 2, and optimize the sub-problem in 3.2.5) until convergence.

Claims

1. A passive intelligent reflector-assisted cloud access network communication perception method, characterized in that: a baseband unit (BBU) pool communicates with K single-antenna users via L multi-antenna relay hubs (RRHs) and detects the desired target; a point-to-point wired fronthaul link is used between the BBU pool and the RRHs; due to obstruction between the RRHs and the detected target, the passive intelligent reflector reflects the signal to the target to be detected, and simultaneously sends the signal to the user via the reflector to enhance communication; and the method jointly optimizes the transmit beamforming matrices of the BBU pool and RRHs, the phase shift matrix of the reflector, and the covariance matrix of the forward quantization noise, including the following steps: 1.1) In a dual-function radar communication system in a cloud wireless access network assisted by a passive intelligent reflector, it sends communication symbols to K single-antenna users and simultaneously sends radar detection waveforms to the target of interest. The system includes 1 BBU, L RRHs, and M IRS units to assist in enhanced communication and radar perception. The BBU and RRHs are connected by wires, and the RRHs use ULA array antennas with N antennas. R ; 1.2) The BBU pool first beamforms all user signals c transmitted to the lth RRH, and then superimposes them with the radar sensing signal x0. The resulting signal is: is the signal corresponding to the lth RRH, W c is the communication beamforming matrix, c~CN(0,I K ), x0~CN(0,R0); Since the capacity of the wireless fronthaul link is limited, the signal It is quantized and compressed before being transmitted to the RRH; by adopting the Gaussian test channel model, the compressed signal can be expressed as: x = W c c+x0+q Where R0 is the covariance matrix of the perception signal; q~CN(0,Ω), Ω is the covariance matrix of the quantization noise, where is the covariance matrix of the compressed signal; 1.3) The signal sent by the first RRH is expressed as: q l ~CN(0,Ω l,l ) is independent of Quantization noise, Ω=diag({Ω l,l}),matrix The representative matrix intercepts the beamforming part corresponding to the RRH from the forward compressed signal of the BBU; the specific structure of the matrix is ​​[(l-1)N R +1,lN R ] The rows form a dimension of NR ×N R The unit matrix, the rest are zero; here we consider the single antenna power equation limitation of RRH 1.4) The signal received by user k is: y k =(h k +g k ΨG)x+n k =z k x+n k in represents the additive white Gaussian noise AWGN of the kth user, h k is the channel state from RRH to user k, g k is the channel state from the reflection surface to user k, Ψ is the phase reflection matrix of the reflection surface, G is the channel state from RRH to the reflection surface, z k =h k +g k ΨG.

2. The passive intelligent reflector-assisted cloud access network communication perception method according to claim 1, characterized in that: In order to achieve the minimum mean square error between the ideal beam and the beam of the smart reflector, the dual-function radar communication system of the cloud wireless access network assisted by the passive smart reflector needs to be performed under the conditions of user SINR limitation and forward compression rate limitation. Finally, the covariance matrix R of the RRH transmission signal, the forward quantization noise covariance matrix Ω, the phase reflection matrix Ψ of the smart reflector and the beamforming matrix W of the communication signal are calculated. c Optimize The optimization problem is expressed as: R≥0,Ω≥0(f) Where J is the number of sampling angles, is the spatial frequency steering vector, an IRS with a linearly spaced isotropic array, element spacing d, wavelength λ, α is the scaling factor, d(θ) is the ideal beam; (d) is the linear user SINR constraint, Γ is the threshold, and w i W c The i-th column of represents the covariance of Gaussian white noise; (e) is the forward compression rate constraint, which amplifies the original constraint and requires that the amplified condition meets the forward transmission limit capacity C l , in the constraint (f), ≥ indicates the positive semidefiniteness of the matrix.

3. The passive intelligent reflector-assisted cloud access network communication perception method according to claim 2, characterized in that: The optimization problem needs to be decomposed into two sub-problems and solved alternately; 3.1) For the first sub-problem, we first need to fix Ψ in the optimization problem to optimize the RRH beamforming matrix W c and the forward quantization noise covariance matrix Ω; 3.1.1) For constraint (b), define Since the solution of R0 to the sub-problem can be omitted, we can get: 3.1.2) Rewrite the constraint (d) in the optimization problem as follows: 3.1.3) Constraint (e) can be further transformed into the following approximate constraint: S l is an auxiliary variable, which is required to be a semi-positive definite matrix; when When , the equality sign in the above formula holds; 3.1.4) First define The optimization problem is transformed into: R≥0,R k ≥0,Ω≥0 This problem is directly optimized to obtain the optimal w k and R; 3.2) For the second sub-problem, we first need to fix the optimization variables R and w k and Ω to optimize the reflection surface matrix Ψ; 3.2.1) User k’s achievable rate R k Written as: where w k Auxiliary constant variables; u k is the linear estimator for decoding the signal received by user k; Constraint (d) is expressed as: For a fixed transmit beamformer, the optimal solution for a known linear receiver is given by: 3.2.2) Each iteration updates w k and R, so that the constraints meet the constant constraint requirements; define The constraint condition in 3.2.1) was finally rewritten as The convex form of : log(w k )-Tr(ψ H a k B k ψ-2ψ H d k +c)+1≥log(1+Γ) in: 3.2.3) Rewrite about Objective function: Where Q(θ)=diag(a(θ) H )GRG H diag(a(θ)), Using auxiliary variable t j There are constraints on the optimization objective function: 3.2.4) Use the semi-definite relaxation method to directly Optimize and obtain the result ψ; the objective function is calculated by Taylor formula in ψ t Expand to get in ψ t The initial value of is obtained by satisfying the reflection condition of the passive intelligent reflective surface; 3.2.5) Finally, we get the optimized form of sub-problem 2 in SDR form: ψ H Q(θ)ψ-αd(θ)≤t j ,t j ≥0 3.2.6) According to the optimization in step 3.2.5) Perform Gaussian randomization a sufficient number of times to obtain the best performance ψ, and determine whether the optimization has converged. If it has not converged, update the variable S according to 3.1.3). l , optimize the sub-problem 1 in 3.1.4), update the variable w k and R k , and then update the relevant variables in sub-problem 2, and optimize the sub-problem in 3.2.5) until convergence.

Images