Method for assisting user arrival angle estimation by active intelligent reflecting surface in cloud access network

By using active intelligent reflective surfaces and multi-antenna remote RF heads to coordinate the signal processing in the cloud access network, the target arrival angle is estimated using the atomic norm minimization method, the challenge of target positioning and perception capabilities in future communication scenarios is solved, and the goal positioning and perception capabilities are achieved. High-precision target positioning and communication performance improvements.

CN119011341APending Publication Date: 2024-11-22ZHEJIANG UNIV OF TECH
View PDF 0 Cites 0 Cited by

Patent Information

Application Number
CN202410951517.2
Authority / Receiving Office
CN · China
Patent Type
Applications(China)
Current Assignee / Owner
Filing Date
2024-07-16
Publication Date
2024-11-22

AI Technical Summary

Technical Problem

In future communication scenarios, devices need to have both communication and perception capabilities, especially in the cloud access network environment. How to effectively estimate the target arrival angle to achieve target positioning and improve communication performance is a challenge.

Method used

By introducing active intelligent reflection surface (Active-IRS) into the cloud access network, the reflection amplitude and phase are adjusted in snap time using the atomic norm minimization method, combining multi-antenna remote radio frequency head (RRH) and baseband processing unit The pool (BBU) collaborates with the signal to estimate the target arrival angle.

Benefits of technology

It realizes high-precision perception function for unobstructed targets within the service scope, improves wireless communication performance, has smaller errors and higher accuracy than other algorithms, and can estimate the arrival angles of LOS and NLOS channels at the same time.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN119011341A_ABST
    Figure CN119011341A_ABST
Patent Text Reader

Abstract

The invention relates to the technical field of communication perception integration, and discloses a method for assisting user arrival angle estimation by an active intelligent reflecting surface in a cloud access network, which comprises the following steps that: firstly, a multi-antenna remote radio head (RRH) transmits a signal to K targets, each receiving RRH receives a reflection path echo signal from an IRS, the RRH uploads the received signal to a baseband processing unit (BBU), and the BBU transmits the received signal to the K targets; the BBU estimates an angle of arrival from the target to each IRS from the received signal; by adopting the scheme, the BBU can estimate the arrival angle from the target to the IRS with relatively high precision, and high-precision target positioning is facilitated.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of communication perception integration, and in particular to a method for estimating user arrival angles assisted by active intelligent reflective surfaces in a cloud access network. Background Art

[0002] After communications enter the post-5G and 6G era in the future, a series of emerging application scenarios will emerge, including autonomous driving, industrial automation, virtual reality, etc. These scenarios require the system to be not limited to the communication between devices, but also require the devices to have a certain perception of the surrounding environment. The integrated communication and perception technology can not only meet the needs of new application scenarios for devices to have both communication and perception capabilities, but also integrate the communication and perception systems to achieve a certain perception of the target while ensuring the communication quality. The cloud access network (C-RAN) software-based network functions of wireless access are virtualized functions and deployed in a standard cloud environment, which can improve design flexibility and computing scalability, improve energy efficiency and reduce integration costs. The intelligent reflecting surface (IRS) is a planar array composed of a large number of reconfigurable passive elements, each of which can cause a certain phase shift independently of the incident signal, thereby synergistically changing the propagation of the reflected signal, reconstructing the entire wireless channel environment, and improving wireless communication performance. Summary of the invention

[0003] The purpose of the present invention is to improve the communication performance by solving the basic problem of target positioning in an active intelligent reflecting surface (Active-IRS) assisted system under the framework of a cloud access network (C-RAN) and utilizing the intelligent reflecting surface technology.

[0004] A method for estimating the arrival angle of a target with active IRS assistance in a cloud access network, wherein the RRH responsible for sending signals transmits signals to K targets, and the BBU pool processes the received signals from each receiving RRH. The reflection amplitude and phase of the IRS are adjusted in each snapshot time, and the arrival angle of the target to each IRS is estimated by using the atomic norm minimization method.

[0005] Furthermore, a method for estimating target arrival angle using active IRS in a cloud access network specifically includes the following steps:

[0006] 1.1) The system for active IRS-assisted target arrival angle estimation in the cloud access network includes a BBU pool, P RRHs, and J IRSs. A wired fronthaul link is used to transmit information between the BBU pool and each RRH. Each receiving RRH is equipped with a uniform linear antenna array (ULA) composed of M antennas, which is responsible for sensing the target angle information in the cell. The active IRS consists of N reflection units. The BBU pool knows the status information of the line-of-sight (LOS) channel from the receiving RRH to the IRS. The target is blocked from the RRH, and the transmitting RRH is blocked from the IRS.

[0007] 1.2) Send RRH transmission signals to K targets. The target echo signals are reflected by IRS and then transmitted to the receiving RRH. The BBU pool processes the received signals from P RRHs at the same time. ρ represents the signal transmission power. Suppose the target reflection signal at the i-th snapshot is

[0008] S i =ρ[s 1,i ,s 2,i ,...,s K,i ] T (i=1,2,...,I).

[0009] s K,i represents the reflection signal of the Pth target at the i-th snapshot

[0010] 1.3) The channel from the jth IRS to the pth RRH is

[0011]

[0012] in It is the first IR The small-scale fading coefficient from the jth IRS to the pth RRH under the non-line-of-sight path, (l IR =1,2,…,L IR -1)

[0013]

[0014] μ0 represents the path loss coefficient when the reference distance is d0, d j,p represents the distance from the jth IRS to the pth RRH path, Q is the Ricean channel factor, L IR represents the total number of paths, γ j,p is the path loss coefficient from the jth IRS to the pth RRH, It is the first IR The complex channel gain of the non-line-of-sight path, τ~CN(0,1);

[0015] is the large-scale fading coefficient from the jth IRS to the pth RRH in the line-of-sight path,

[0016]

[0017] The array response vector (l IR =0,1,…,L IR -1)

[0018]

[0019] Among them, d RRH Indicates the distance between two adjacent RRH antennas, d IRS represents the distance between two adjacent reflection elements of IRS; λ represents the carrier wavelength; and Respectively represent the first IR The departure angle and arrival angle from the jth IRS to the pth RRH in the path;

[0020]

[0021]

[0022] and They represent the non-line-of-sight channel and line-of-sight channel matrices from the j-th IRS to the p-th RRH respectively.

[0023] 1.4) The channels between all targets and the jth IRS are

[0024]

[0025] in is the large-scale fading coefficient from the target to the jth IRS under the line-of-sight path,

[0026]

[0027] The array response vector of the jth IRS arrival angle (j=1,2,…,J) is

[0028]

[0029] represents the arrival angle of all targets to the jth IRS

[0030]

[0031] 1.5) The j-th IRS control matrix can be expressed as

[0032]

[0033] Among them, vj,N and ω j,N represent the reflection amplitude and phase of the N reflection elements in the j-th IRS respectively.

[0034] 1.6) Each snapshot IRS changes the reflection coefficient L times RRH forms L signals into vectors, so the i-th snapshot signal received by IRS is

[0035]

[0036] in

[0037]

[0038] N j,i,1 represents the noise generated by the jth active reflector at the i-th snapshot, which is related to the inherent device noise of the active IRS element of the input noise, where represents the noise power, I is the unit matrix; N j,i,2 represents the static noise of the jth active reflector at the i-th snapshot, which can usually be ignored.

[0039] 1.7) The i-th snapshot signal received by the p-th RRH is

[0040]

[0041] in

[0042]

[0043] n p,i represents the Gaussian white noise received by the pth RRH at the i-th snapshot, where

[0044] 1.8) The I-th snapshot signal received by the p-th RRH is

[0045]

[0046] The signal received by the mth antenna of the pth RRH is n p,m is the additive white Gaussian noise of the channel between the IRS and the mth antenna of the pth RRH,

[0047] 1.9) After the RRH receives the signal from the target, each RRH quantizes it and then sends the quantized signal to the BBU pool. First, the signal received by the mth antenna of the pth RRH is normalized to

[0048]

[0049] where η p,m is a scaling factor; the quantized signal can be calculated from the normalized signal as:

[0050]

[0051] in Represents the integer closest to x. p,m It represents the resolution of the quantizer when quantizing the received signal at the mth antenna of the pth RRH. The quantized signal is then sent to the BBU pool through an error-free fronthaul link.

[0052] The signal finally received by the BBU pool is

[0053]

[0054] Furthermore, in a system with active IRS-assisted target arrival angle estimation in a cloud access network, the angle between the IRS and the RRH in the LOS channel is known. and The arrival angle between the target and all IRSs needs to be estimated The specific steps are as follows:

[0055] 2.1) Deform the i-th snapshot signal received by the p-th RRH of the NLOS channel when the l-th reflection surface changes

[0056]

[0057] ω j,l,i represents the reflection coefficient of the jth reflector at the lth change of the i-th snapshot. The i-th snapshot signal received by the p-th RRH of the NLOS channel is

[0058]

[0059] Among them I M is the identity matrix of M rows and M columns;

[0060]

[0061] The definition is

[0062]

[0063] Where ◇ represents the Khatri-Rao product, represents the Kronecker product, * represents conjugate, and T represents transpose.

[0064] 2.2) In order to estimate the received signal Y B Angle YB Rewrite

[0065]

[0066] in

[0067]

[0068] S=[S1,...,S I ], (j=1,2,…,J; p=1,2,…,P).

[0069] Z 1D,j (:,i) represents Z 1D,j The i-th column (i=1,2,…,I).

[0070] 2.3) Define composition Z 1D,j The atomic set A 1D,j (j=1,2,…,J)

[0071]

[0072] express Array response vector of ;

[0073] Define composition Z 2D,j,p The atomic set A 2D,j,p (j=1,2,…,J; p=1,2,…,P)

[0074]

[0075] and express and An array of response vectors.

[0076] 2.4) The following optimization problem is introduced

[0077]

[0078] where λ1, λ2, λ3 and λ4 are regularization coefficients,

[0079]

[0080] is the matrix to be optimized, is a Toeplitz matrix that can be decomposed by Vandermonde. 1D,j Perform Vandermonde decomposition or root-music to get the estimated angle

[0081] 2.6) T 1D,jPerform Vandermonde decomposition to get the estimated angle

[0082]

[0083] The design idea of ​​the present invention is as follows: first, a multi-antenna remote radio head (RRH) transmits a signal to K targets, and each receiver (RRH) receives an echo signal from the reflection path of the IRS. The RRH uploads the received signal to the baseband processing unit pool (BBU), and the BBU estimates the arrival angle from the target to each IRS from the received signal.

[0084] The beneficial effect of the present invention is to realize the perception function of unobstructed targets within its service range through DOA estimation. On the basis of estimating DOA, the present invention can simultaneously estimate the DOA of the LOS channel and the RRH-IRS cascade channel of the NLOS. Compared with other algorithms, the present invention has smaller error and higher accuracy, which can improve communication performance and provide a reference for future communications. BRIEF DESCRIPTION OF THE DRAWINGS

[0085] Figure 1 Schematic diagram of the system of the present invention;

[0086] Figure 2 RMSE waveform diagram under different algorithms. DETAILED DESCRIPTION

[0087] The present invention will be further described below in conjunction with the accompanying drawings.

[0088] A method for estimating the arrival angle of a target with active IRS assistance in a cloud access network can realize the positioning of the target with the assistance of an active intelligent reflective surface in the context of the cloud access network, realize the perception function of unobstructed targets within its service range, and improve the wireless communication performance.

[0089] like Figure 1 As shown in FIG. 1 , a method for estimating the arrival angle of a target using active IRS in a cloud access network is provided. The RRH responsible for sending the signal transmits the signal to K targets, and the BBU pool processes the received signals from each receiving RRH. The reflection amplitude and phase of the IRS are adjusted within each snapshot time, and the arrival angle of the target to each IRS is estimated using the atomic norm minimization method. Specifically, the following steps are included:

[0090] 1.1) The system for active IRS-assisted target arrival angle estimation in the cloud access network includes a BBU pool, P RRHs, and J IRSs. A wired fronthaul link is used to transmit information between the BBU pool and each RRH. Each receiving RRH is equipped with a uniform linear antenna array (ULA) composed of M antennas, which is responsible for sensing the target angle information in the cell. The active IRS consists of N reflection units. The BBU pool knows the status information of the line-of-sight (LOS) channel from the receiving RRH to the IRS. The target is blocked from the RRH, and the transmitting RRH is blocked from the IRS.

[0091] 1.2) Send RRH transmission signals to K targets. The target echo signals are reflected by IRS and then transmitted to the receiving RRH. The BBU pool processes the received signals from P RRHs at the same time. ρ represents the signal transmission power. Suppose the target reflection signal at the i-th snapshot is

[0092] S i =ρ[s 1,i ,s 2,i ,...,s K,i ] T (i=1,2,...,I).

[0093] s K,i represents the reflection signal of the Pth target at the i-th snapshot

[0094] 1.3) The channel from the jth IRS to the pth RRH is

[0095]

[0096] in It is the first IR The small-scale fading coefficient from the jth IRS to the pth RRH under the non-line-of-sight path, (l IR =1,2,…,L IR -1)

[0097]

[0098] μ0 represents the path loss coefficient when the reference distance is d0, d j,p represents the distance from the jth IRS to the pth RRH path, Q is the Ricean channel factor, L IR represents the total number of paths, γ j,p is the path loss coefficient from the jth IRS to the pth RRH, It is the first IR The complex channel gain of the non-line-of-sight path, τ~CN(0,1);

[0099] is the large-scale fading coefficient from the jth IRS to the pth RRH in the line-of-sight path,

[0100]

[0101] The array response vector (l IR =0,1,…,L IR -1)

[0102]

[0103] Among them, d RRH Indicates the distance between two adjacent RRH antennas, d IRS represents the distance between two adjacent reflection elements of IRS; λ represents the carrier wavelength; and Respectively represent the first IR The departure angle and arrival angle from the jth IRS to the pth RRH in the path;

[0104]

[0105] and They represent the non-line-of-sight channel and line-of-sight channel matrices from the j-th IRS to the p-th RRH respectively.

[0106] 1.4) The channels between all targets and the jth IRS are

[0107]

[0108] in is the large-scale fading coefficient from the target to the jth IRS under the line-of-sight path,

[0109]

[0110] The array response vector of the jth IRS arrival angle (j=1,2,…,J) is

[0111]

[0112] represents the arrival angle of all targets to the jth IRS

[0113]

[0114] 1.5) The j-th IRS control matrix can be expressed as

[0115]

[0116] Among them, v j,N and ω j,Nrepresent the reflection amplitude and phase of the N reflection elements in the j-th IRS respectively.

[0117] 1.6) Each snapshot IRS changes the reflection coefficient L times RRH forms L signals into vectors, so the i-th snapshot signal received by IRS is

[0118]

[0119] in

[0120]

[0121] N j,i,1 represents the noise generated by the jth active reflector at the i-th snapshot, which is related to the inherent device noise of the active IRS element of the input noise, where represents the noise power, I is the unit matrix; N j,i,2 represents the static noise of the jth active reflector at the i-th snapshot, which can usually be ignored.

[0122] 1.7) The i-th snapshot signal received by the p-th RRH is

[0123]

[0124] in

[0125]

[0126] n p,i represents the Gaussian white noise received by the pth RRH at the i-th snapshot, where

[0127] 1.8) The I-th snapshot signal received by the p-th RRH is

[0128]

[0129] The signal received by the mth antenna of the pth RRH is n p,m is the additive white Gaussian noise of the channel between the IRS and the mth antenna of the pth RRH,

[0130] 1.9) After the RRH receives the signal from the target, each RRH quantizes it and then sends the quantized signal to the BBU pool. First, the signal received by the mth antenna of the pth RRH is normalized to

[0131]

[0132] where η p,mis a scaling factor; the quantized signal can be calculated from the normalized signal as:

[0133]

[0134] in Represents the integer closest to x. p,m It represents the resolution of the quantizer when quantizing the received signal at the mth antenna of the pth RRH. The quantized signal is then sent to the BBU pool through an error-free fronthaul link.

[0135] The signal finally received by the BBU pool is

[0136]

[0137] In a system with active IRS-assisted target arrival angle estimation in a cloud access network, the angle between the IRS and RRH in the LOS channel is known. and The arrival angle between the target and all IRSs needs to be estimated The specific steps are as follows:

[0138] 2.1) Deform the i-th snapshot signal received by the p-th RRH of the NLOS channel when the l-th reflection surface changes

[0139]

[0140] ω j,l,i represents the reflection coefficient of the jth reflector at the lth change of the i-th snapshot. The i-th snapshot signal received by the p-th RRH of the NLOS channel is

[0141]

[0142] Among them I M is the identity matrix of M rows and M columns;

[0143]

[0144] The definition is

[0145]

[0146] Where ◇ represents the Khatri-Rao product, represents the Kronecker product, * represents conjugate, and T represents transpose.

[0147] 2.2) In order to estimate the received signal Y B Angle Y B Rewrite

[0148]

[0149] in

[0150]

[0151] S=[S1,...,S I ], (j=1,2,…,J; p=1,2,…,P).

[0152] Z 1D,j (:,i) represents Z 1D,j The i-th column (i=1,2,…,I).

[0153] 2.3) Define composition Z 1D,j The atomic set A 1D,j (j=1,2,…,J)

[0154]

[0155] express Array response vector of ;

[0156] Define composition Z 2D,j,p The atomic set A 2D,j,p (j=1,2,…,J; p=1,2,…,P)

[0157]

[0158] and express and An array of response vectors.

[0159] 2.4) The following optimization problem is introduced

[0160]

[0161] where λ1, λ2, λ3 and λ4 are regularization coefficients,

[0162]

[0163] is the matrix to be optimized, is a Toeplitz matrix that can be decomposed by Vandermonde. 1D,j Perform Vandermonde decomposition or root-music to get the estimated angle

[0164] 2.6) T 1D,j Perform Vandermonde decomposition to get the estimated angle

[0165]

[0166] like Figure 2 As shown, the RMSE waveforms of the present invention and the comparative schemes: OMP algorithm and L1 norm algorithm under three different algorithms, where:

[0167]

[0168] The use of this solution enables the BBU to estimate the arrival angle from the target to the IRS with high accuracy, which is helpful for high-precision target positioning.

Claims

1. A method for estimating user arrival angle using active intelligent reflective surface assistance in a cloud access network, characterized in that: 1.1) The system for active IRS-assisted target arrival angle estimation in the cloud access network includes a BBU pool, P RRHs, and J IRSs. A wired fronthaul link is used between the BBU pool and each RRH for information transmission. Each receiving RRH is equipped with a uniform linear antenna array (ULA) consisting of M antennas, which is responsible for sensing the target angle information in the cell. The active IRS consists of N reflective units. The BBU pool knows the state information of the line-of-sight (LOS) channel from the receiving RRH to the IRS. The target is blocked from the RRH, and the transmitting RRH is blocked from the IRS. 1.2) Send RRH transmission signals to K targets. The target echo signals are reflected by IRS and then transmitted to the receiving RRH. The BBU pool processes the received signals from P RRHs at the same time. ρ represents the signal transmission power. Suppose the target reflection signal at the i-th snapshot is: S i =ρ[s 1,i ,s 2,i ,...,s K,i ] T (i=1,2,...,I) s K,i represents the reflection signal of the kth target at the i-th snapshot; 1.3) The channel from the jth IRS to the pth RRH is: in It is the first IR The small-scale fading coefficient from the jth IRS to the pth RRH under the non-line-of-sight path, l IR =1,2,…,L IR -1; μ 0 Indicates that the reference distance is d 0 The path loss coefficient at time d j,p represents the distance from the jth IRS to the pth RRH path, Q is the Ricean channel factor, L IR represents the total number of paths, γ j,p is the path loss coefficient from the jth IRS to the pth RRH, It is the first IR The complex channel gain of the non-line-of-sight path, τ~CN(0,1); is the large-scale fading coefficient from the jth IRS to the pth RRH in the line-of-sight path, The array response vector (l IR =0,1,…,L IR -1) Among them, d RRH Indicates the distance between two adjacent RRH antennas, d IRS represents the distance between two adjacent reflection elements of IRS; λ represents the carrier wavelength; and Respectively represent the first IR The departure angle and arrival angle from the jth IRS to the pth RRH in the path; and They represent the non-line-of-sight channel and line-of-sight channel matrices from the j-th IRS to the p-th RRH respectively; 1.4) The channels between all targets and the jth IRS are in is the large-scale fading coefficient from the target to the jth IRS under the line-of-sight path, The array response vector of the jth IRS arrival angle (j=1,2,…,J) is represents the arrival angle of all targets to the jth IRS 1.5) The j-th IRS control matrix is ​​expressed as Among them, v j,N and ωj,N denote the reflection amplitude and phase of the N reflection elements in the jth IRS respectively; 1.6) Each snapshot IRS changes the reflection coefficient L times RRH forms L signals into vectors, so the i-th snapshot signal received by IRS is in N j,i,1 represents the noise generated by the jth active reflector at the i-th snapshot, where represents the noise power, I is the unit matrix; N j,i,2 represents the static noise of the jth active reflector at the i-th snapshot; 1.7) The i-th snapshot signal received by the p-th RRH is in n p,i represents the Gaussian white noise received by the pth RRH at the i-th snapshot, where 1.8) The I-th snapshot signal received by the p-th RRH is The signal received by the mth antenna of the pth RRH is n p,m is the additive white Gaussian noise of the channel between the IRS and the mth antenna of the pth RRH, 1.9) After the RRH receives the signal from the target, each RRH quantizes it and then sends the quantized signal to the BBU pool. First, the signal received by the mth antenna of the pth RRH is normalized to where η p,m is a scaling factor; the quantized signal is calculated from the normalized signal as: in represents the integer closest to x, B p,m It represents the resolution of the quantizer when quantizing the received signal at the mth antenna of the pth RRH. The quantized signal will then be sent to the BBU pool through an error-free fronthaul link. The signal finally received by the BBU pool is 2. According to claim 1, a method for estimating user arrival angle using active intelligent reflector in a cloud access network, It is characterized in that In the system for active IRS-assisted target arrival angle estimation in the cloud access network, the angle between the IRS and the RRH in the LOS channel is known. and The arrival angle between the target and all IRSs needs to be estimated (j=1,2,…,J); the specific steps are as follows: 2.1) Deform the i-th snapshot signal received by the p-th RRH of the NLOS channel when the l-th reflection surface changes ω j,l,i represents the reflection coefficient of the jth reflection surface at the lth change in the i-th snapshot; The i-th snapshot signal received by the p-th RRH of the NLOS channel is Among them I M is the identity matrix of M rows and M columns; The definition is where ◇ represents the Khatri-Rao product, represents Kronecker product, * represents conjugation, and T represents transpose; 2.2) In order to estimate the received signal Y B Angle Y B Rewrite in S=[(S 1 ) T ,...,(S I ) T ] T j=1,2,…,J;p=1,2,…,P Z 1D,j (:,i) represents Z 1D,j The i-th column, i=1,2,…,I; 2.3) Define composition Z 1D,j The atomic set A 1D,j , j = 1, 2, ..., J; express Array response vector of ; Define composition Z 2D,j,p The atomic set A 2D,j,p (j=1,2,…,J; p=1,2,…,P) and express and Array response vector of ; 2.4) The following optimization problem is introduced where λ 1 , 2 , 3 and λ 4 is the regularization coefficient, is the matrix to be optimized, is a Toeplitz matrix that can be decomposed by Vandermonde. 1D,j Perform Vandermonde decomposition or root-music to get the estimated angle 2.6) T 1D,j Perform Vandermonde decomposition to get the estimated angle