CN-122001429-A - Robust wave beam forming method of networked general sense integrated system

CN122001429ACN 122001429 ACN122001429 ACN 122001429ACN-122001429-A

Abstract

The invention discloses a robust beam forming method of a networked sense-of-general integrated system, which comprises the steps of establishing a networked sense-of-general integrated system with multiple base stations cooperated, constructing a closed solution of a received beam forming vector and a semi-planned sub-problem of transmitting end beam and power by taking joint design base station received beam forming, sensing transmitting beam forming and user transmitting power as variables, under the premise of meeting uplink communication interruption probability constraint and power budget, minimizing an objective function of a target position estimated Kramer, deriving to obtain a solvable convex approximation problem, designing a joint iterative algorithm based on alternate optimization and successive convex approximation to solve the convex approximation problem, and updating a closed solution of a received beam forming vector and a semi-planned sub-problem of transmitting end beam and power by alternate iteration, so as to obtain a robust beam forming design and power distribution strategy, maximizing equivalent beam forming gain of a target relative to the base station direction, further minimizing a target position estimation error and enhancing the robustness of the system.

Inventors

LV LING
XIA TIANLANG
DAI YANPENG
LI DEZHENG
Cai Cunchao

Assignees

大连海事大学

Dates

Publication Date: 20260508
Application Date: 20260320

Claims (9)

1. A robust wave beam forming method of a networked sense-of-general integrated system is characterized by comprising the following steps: the method comprises the following steps of S1, establishing a multi-base station cooperative networked sense-of-general integrated system, constructing an objective function of a Keramelteon boundary for minimizing target position estimation on the premise of meeting uplink communication interruption probability constraint and power budget by taking joint design base station receiving beam forming, sensing transmitting beam forming and user transmitting power as variables under the uncertainty condition that statistical errors exist on the basis of communication channel state information between uplink users and base stations; S2, carrying out salifying treatment on non-convex constraint in the objective function, converting interrupt probability constraint of a communication link into deterministic second-order cone programming and linear matrix inequality constraint by using a Bernstein inequality, and processing highly-coupled non-convex terms related to inverse of a Fisher information matrix in the objective function by using a Shultier theorem and a semi-positive relaxation method to derive a resolvable convex approximation problem; And S3, designing a joint iteration algorithm based on alternating optimization and successive approximation to solve the convex approximation problem, and updating a closed solution of a received beam forming vector and a semi-planned sub-problem of a transmitting end beam and power of a base station perceived signal through alternating iteration, so as to obtain a robust beam forming design and power distribution strategy, so that the beam forming gain is maximized in the direction of a target relative to the base station, and further, the target position error is minimized.
2. The method for forming the robust beam of the networked ventilation and induction integrated system according to claim 1, wherein the networked ventilation and induction integrated system model with the cooperation of the multiple base stations processes uplink communication signals from a plurality of communication users in a mode of coordinating reception beam forming by M base stations so as to effectively inhibit intra-cluster and inter-cluster interference and ensure uplink communication quality; a plurality of base stations connected to the CPU via a forward link, the base stations Configuration of Root transmitting antenna A receiving antenna, which adopts a uniform linear array with half wavelength interval and receives the coverage area The up communication signal of each single antenna user, the considered base station position is known, the three-dimensional position of base station m is expressed as , wherein, And the position coordinates of the base station along the x-axis direction in a three-dimensional rectangular coordinate system are represented. In the same way, the processing method comprises the steps of, And (3) with Respectively correspond to the coordinate values of the coordinate values in the directions of the y axis and the z axis. Location of target Is a key parameter that needs to be estimated by a sensing process, wherein, Representing coordinate values of the target in the x-axis direction in a preset coordinate system, and And The spatial position parameters of the targets are jointly formed; The establishment of the networked communication and sensing integrated system comprises an uplink communication model and a collaborative sensing model which consider the statistics of channel state information errors; the construction process of the uplink communication model is as follows: Transmitting signal of single antenna user k in base station m cluster The method comprises the following steps: Wherein: for the transmit power allocated to user k; is a communication data symbol and satisfies Communication symbols of different users And Are assumed to be independent of each other and k+.i; user k of coverage area of base station m and uplink communication channel thereof Including the estimation error, modeled as: Wherein: Representing the channel estimation vector(s), Representing a corresponding statistical channel estimation error, the statistical channel state information error being expressed as I.e. obeying a circularly symmetric complex gaussian distribution of zero mean, wherein: Is an error covariance matrix, and has , Wherein Is a dimensionless coefficient with a value range of E [0, 1) is the channel uncertainty level; Focusing on the same frequency bandwidth B, all M base stations and K users in the coverage area of the base stations perform uplink communication and collaborative sensing, For the average transmit power of the uplink user k, In order to receive the beamforming vector(s), Is an uplink communication channel; is the response matrix of the object and, Is a half-positive definite matrix, and the matrix is a half-positive definite matrix, Is the transmit covariance matrix of the perceived signal, Is the perceived transmit beamforming vector for base station j, and thus the signal-to-interference-and-noise ratio of the communication signal for communication user k at the current receiving base station n is given by: Wherein: For the power of the additive white gaussian noise on each antenna, Is the sum of squares of the modes of the elements in the vector, In order to receive the beamforming vector(s), Uplink communication channels for other users i; The construction process of the collaborative awareness model is as follows: perceived signal transmitted by base station m The method comprises the following steps: Wherein: is the perceived transmit beamforming vector for base station m; Is a normalized power perceived signal wave transmitted by base station m, i.e ; Current base station When the receiver is used as a perception receiver, echoes reflected by all base stations through targets are received The method comprises the following steps: Wherein: a target response matrix from the transmitting base station j to the receiving base station n via the target reflection, Is the perceived signal transmitted by the transmitting base station j, Perceiving additive noise of a receiving link for a receiving base station n; is the complex reflection coefficient, the radar cross section including the path loss and the target, the transmit steering vector for the transmitting base station j towards the target The method comprises the following steps: ; For a receiving base station n, it receives a steering vector of signals from a target direction The method comprises the following steps: ; From the perspective of perception, the received signal of the current base station n is sorted as follows: Wherein: The signal mean vector is the deterministic part of the signal depending on the parameter p to be estimated, here the sum of all perceived echoes; using interference plus noise covariance matrix The communication signals d k (t) of different users are independent of each other and independent of the thermal noise n (t), so the expectation of the cross terms in the above equation is zero, and thus the covariance matrix is finalized as: Wherein: Is that A rank identity matrix.
3. The robust beamforming method of a networked ventilation integrated system according to claim 1, wherein the process of constructing the target function with minimized target location estimation caramerol boundary is as follows: Target position parameter Snow-cost information matrix of (2) Derived from the generalized Slepian-bands formula, in particular the first of the Fisher information matrix Elements of which The definition is: Wherein the inverse of the covariance matrix The pre-whitening filter is used on the physical level; Taking a real part; furthermore, to obtain the signal mean value For position parameters 、、 The derivative of (a) is obtained by sampling the signal E times to obtain the discrete Fourier transform coefficients of E frequency points at the frequency points The signal mean value of (2) is: Wherein: is the perceived beamforming vector for base station j; Is at the frequency point A normalized power perceived signal wave transmitted by base station j; position component Comprises 、 And And (3) deriving to obtain: And vectorizing the derivatives of all the frequency points, namely obtaining the derivative from a substituted FIM matrix formula: After obtaining FIM of each receiving base station, constructing final perceived performance index (CRB) of Cramerro boundary, wherein CRB matrix corresponding to single base station sensing receiver is as follows And expanding CRB of a single base station to a whole networked multi-base station general sense integrated system, so as to obtain a fusion total CRB of the whole multi-base station collaborative networked general sense integrated system, wherein the fusion total CRB is as follows: The method comprises the steps of jointly optimizing the receiving beam shaper, the perception transmitting beam shaper and the transmitting power of users of a base station, minimizing the trace of the total CRB of a networking general sense integrated system with multiple base stations cooperated as an objective function, and simultaneously ensuring that the system meets the most basic communication service quality and physical power limit, and the original non-convex optimization problem of the objective function The expression of (2) is as follows: Wherein: =[ ,..., 0 represents the interruption probability of the signal-to-interference-and-noise ratio of the communication user, namely For the outage probability of user k, Representing the minimum signal-to-interference-and-noise threshold for user k, Representing a set of uplink user index sets covered by base station n, Indicating the maximum transmit power budget for base station j, The maximum transmit power upper limit of user k covered by base station n; the C1 constraint aims at ensuring the outage probability of each user, and ensures that the kth user can realize successful data reception with high enough signal-to-interference-and-noise ratio to exceed ; C2 constraint represents a transmit power limit of the base station; the C3 constraint represents a user transmit power limit.
4. A robust beamforming method of a networked ventilation and inductance integrated system as set forth in claim 3, wherein the original non-convex is optimized The process of performing the optimized transformation is as follows: Introducing an auxiliary scalar variable t to solve the original problem To minimize the problem of a linear variable t, while at the same time The objective function in (a) is moved into the constraint condition, and the optimization problem is converted into: introducing a new auxiliary matrix variable Dimension and The same applies to Representation of Inequality of Establishing is equivalent to the existence of a matrix , So that the following two conditions are simultaneously satisfied: The non-convex constraint containing the matrix inversion is converted into a standard linear matrix inequality that does not contain any matrix inversion in the relation of the corresponding substitution schulb theorem, as shown in the following formula: Wherein: Is a unit matrix of order 2, Is the transpose of the 2 nd order identity matrix.
5. The robust beamforming method of a networked general sense integrated system according to claim 1, wherein the method is characterized in that the non-convex constraint in the objective function is subjected to convex processing, the interrupt probability constraint of the communication link is converted into deterministic second order cone programming and linear matrix inequality constraint by using a bernstein type inequality, and the highly coupled non-convex term related to the inverse of the feesky information matrix in the objective function is processed by using a sultam theorem and a semi-positive relaxation method, and the specific process for deriving the resolvable convex approximation problem is as follows: Algebraic expansion and sorting of the inequality of the signal-to-interference-and-noise ratio inside the probability in the C1 constraint and channel vector Substitution with an expression containing an error term yields: Wherein: Indicating that the echo is perceived as interfering, Representing noise power; the inequality contains random errors The inequality of the term (d) being the desired signal power term is written as the standard quadratic form: In the case of the formula (I) of this patent, Is a quadratic coefficient matrix, which is related to the power of the expected user k Receiving beams Related to; for linear term coefficient vectors, random errors are described Cross-action with deterministic portions of the desired signal, constant terms Is a scalar, and the specific expression is as follows: applying the argument bernstein type inequality for any given variable , , A kind of electronic device So that the probability inequality holds: The application theory is based on The random vector is Covariance matrix And introducing relaxation variables And Converting the C1 constraint into: Construction of new optimization variables, i.e. perceptual covariance The C2 constraint is rewritten as a linear trace constraint If and only if And is also provided with When true, therefore, the original non-convex optimization problem Restated as: 。
6. the method for forming a robust beam of a networked ventilation integrated system according to claim 1, wherein the joint optimization algorithm based on alternating optimization and successive convex approximation comprises the following steps: First, the original non-convex optimization problem is solved based on an alternate optimization algorithm framework Decoupling is the problem of outer layer optimized receiving beam forming and the problem of inner layer optimized transmitting beam and power distribution; The outer layer optimized receiving beam forming problem converts the receiving beam forming sub problem into generalized Rayleigh quotient maximization problem under the condition of fixed sensing beam and user power, and directly deduces the optimal closed solution of the receiving beam forming weight Updating receive beamforming vectors for each base station Solution to the problem of inner layer optimization of transmit beam and power allocation by fixing Is to (1) Performing successive convex approximation and joint optimization; Then, under the condition of inner layer fixed receiving beam forming, constructing a perception transmitting beam and a power distribution sub-problem, carrying out local linearization convex approximation on a Kramer Luo Jieyao beam by utilizing first-order Taylor expansion, constructing a semi-definite planning model by combining the deterministic interrupt probability constraint after conversion, and carrying out iterative solution by adopting a successive convex approximation algorithm; and finally, updating a closed solution of the outer layer receiving beam forming sub-problem and a semi-definite programming solution of the inner layer transmitting beam and power distribution sub-problem through alternate iteration until a system objective function converges, and recovering a beam forming vector meeting rank-one constraint from a relaxation solution by utilizing a Gaussian randomization method.
7. The robust beamforming method of a networked ventilation integrated system according to claim 6, wherein the specific process of closed-form solution of the received beamforming vector is as follows: Assuming that the current outer layer iteration is at the first Wheel for forming matrix of sensing wave beam of transmitting end And user transmit power Fixing the solution for the previous iteration, updating the receive beamforming vector To maximize the communication performance of the user, serving the base station n Individual users, modeling the interference suffered by their received signals, at a given transmitting end Interference plus noise covariance matrix The method comprises the following steps: Wherein: Representing a user To a base station Is used for the uplink channel vector of (a), Representing slave base stations The transmission reaching the base station via reflection from the target Is a matrix of equivalent channels; Thus, the problem of maximizing the signal-to-interference-and-noise ratio of the user is expressed in terms of generalized Rayleigh maximization: The first step of The normalized closed-form solution of the iterative update is: The closed solution guarantees the optimality of the receiving end under a fixed transmit optimization variable.
8. The method for forming a robust beam of a networked ventilation integrated system according to claim 6, wherein the solving of the problems of optimizing the transmit beam and the power distribution by the inner layer is as follows: Determining C5 constrained convex approximation by using C5 constrained first-order Taylor expansion by adopting a successive convex approximation method, wherein the C5 constrained convex approximation is positioned at the r-th iteration of an inner layer successive convex approximation algorithm, and all approximations are based on the last iteration, namely the r-th iteration Solution 1 time to construct; the non-convex function At the r Iteration point of 1 time Where approximated according to the first-order Taylor expansion of the polynary function, i.e 。 Is that Is specifically expanded as follows: Wherein: Is the last iteration point, will Substituting the specific numerical matrix calculated after the FIM formula; is FIM function vs. perceived beam covariance matrix A specific numerical matrix obtained from the previous iteration point; Is FIM function versus user power And a specific numerical matrix obtained at the last iteration point, Is the optimized variable to be solved in the iteration; thus, in the r-th iteration, the translated constraint The method comprises the following steps: inner layer successive convex approximation (LJL) In the iteration, directly utilizing the solution of the previous iteration And is fixed To update the constraint coefficients of the bernstein type inequality; Defining a constant matrix Due to Fixing the quadratic coefficient matrix at the inner layer The form in the iteration that is related to the last round of power only is: The general expression of the linear term coefficient is expressed as Using updated quadratic coefficient matrices First, the The linear term coefficient vector used for each iteration is: similarly, constant term Also as a function of the optimization variables, in the first In the iteration, the value is updated based on the solution of the previous round: 。
9. The method for forming a robust beam of a networked ventilation integrated system according to claim 6, wherein the process of constructing the perceived transmit beam and the power allocation sub-problem under the condition of inner layer fixed receive beam forming is as follows: suppose the outer layer alternate optimization is in the first place Wheel, where the receive beamforming vector is fixed as Inner layer successive convex approximation In the iteration, the solution obtained by the previous iteration is utilized And As local linearization points, the original non-convex problem is converted into a sub-problem of the form: The sub-problem of standard semi-definite programming is effectively solved by the CVX existing tool box, Variable optimization at the transmitting end is processed by adopting a successive convex approximation method, and the variable optimization is performed at the first layer of the inner layer In the iteration, the solution of the previous round is utilized Updating first-order Taylor expansion coefficients in the outage probability constraint and the Schlemm's complement constraint, and solving a corresponding semi-planned sub-problem to obtain a current optimal update step, wherein the process is repeated until a convergence threshold is met Or (b) ; Iterative at the outer layer In the round, two steps are alternately executed, firstly, the outer layer updates the receiving wave beam And then, the inner layer is called to perform joint optimization on the perceived transmission beam and the power so as to minimize the Keramelteon on the premise of meeting the constraint of the signal-to-interference-and-noise ratio of communication.

Description

Robust wave beam forming method of networked general sense integrated system Technical Field The invention relates to the technical field of wireless communication, in particular to a robust beam forming method of a networked sense-of-general integrated system. Background The 6G aims at realizing 'everything intelligent combination and digital twin', namely realizing the whole-domain coverage, whole-scene perception and intelligent interaction by constructing the deep fusion of the physical world and the digital world. Under the great technical prospect, the traditional communication network architecture faces the unprecedented challenges that frequency spectrum resources are increasingly exhausted, hardware cost and energy consumption are continuously increased, and single communication function cannot meet the requirements of environment sensing high precision and low time delay of emerging vertical industries such as intelligent transportation, low-altitude economy, industrial Internet of things and the like. In this context, sense of general integration has been developed as a key enabling technology for 6G. Many studies have investigated the sense of general integration systems from different angles, with beamforming designs being particularly important. Beamforming allows the transmitted or received wireless signals to be concentrated in a desired direction and also provides a rich degree of spatial freedom, enabling high speed data transmission and high precision target sensing. However, most of the current focus is on single base station architecture. With the rise of intelligent transportation, smart cities and other everything interconnection applications, a single sense-through integrated node is more difficult to meet the sensing requirements of persistence, high precision and wide coverage. More specifically, the limited perceptibility and strict resource constraints of a single sense-of-general integrated node often limit its perceived performance, especially in 6G vertical application scenarios. Although the architecture is simple and convenient in theory analysis, in a practical complex electromagnetic propagation environment, the perception of a single visual angle is extremely easy to be shielded by obstacles such as buildings, trees and the like, so that a blind area for target detection is caused. Therefore, there is a need to explore a multi-node collaborative networked sense-of-general integrated system to achieve high spatial diversity for efficient communication and multi-view perception. In practical deployments, achieving perfect channel state information estimation in practical communication systems remains challenging and impractical due to various factors, such as noise, signal attenuation, and multipath propagation. At present, the beam forming design under imperfect channel state information in the studied all-in-one system only considers the all-in-one system architecture of a single base station. Therefore, it is necessary to study the effect of the imperfections of channel state information on communication and perceptual performance in the case of multi-base station coordination to limit the effect of channel uncertainty on the performance of a practically all-in-one system. In summary, the prior art has the problems that the existing general sense integrated research is limited to a single base station architecture, is limited to single-view shielding and path loss, is difficult to meet the wide coverage and high-precision sensing requirements in a complex environment, the existing networked general sense integrated system research is mostly built on perfect channel state information assumption, ignores the influence of channel estimation errors on system performance in practical application, realizes the evolution from single-node independent sensing to multi-base station networked cooperative sensing, breaks through physical sensing bottleneck, and designs robust beam forming and cooperative resource optimization strategies capable of effectively coping with channel uncertainty in a complex interference environment of multi-base station cooperation. Disclosure of Invention In order to solve the problems, the technical scheme adopted by the invention is that the robust beam forming method of the networked sense-of-general integrated system comprises the following steps: the method comprises the following steps of S1, establishing a multi-base station cooperative networked sense-of-general integrated system, constructing an objective function of a Keramelteon boundary for minimizing target position estimation on the premise of meeting uplink communication interruption probability constraint and power budget by taking joint design base station receiving beam forming, sensing transmitting beam forming and user transmitting power as variables under the uncertainty condition that statistical errors exist on the basis of communication channel state information between uplink u