CN-122027403-A - Channel estimation method and system for transmissive reconfigurable intelligent super surface

CN122027403ACN 122027403 ACN122027403 ACN 122027403ACN-122027403-A

Abstract

The invention provides a channel estimation method and a system for a transmission type reconfigurable intelligent super-surface, wherein the channel estimation method and the system comprise the steps of constructing and estimating a composite channel through uplink and downlink pilot frequency transmission; and constructing a sparse recovery problem from TRIS to a user channel, and solving through a LAOMP algorithm. The method and the device effectively solve the problem of phase ambiguity in cascade channel estimation by decoupling and independently estimating the antenna-TRIS channel and the TRIS-user channel in stages, remarkably reduce pilot frequency overhead and calculation complexity, improve estimation precision and system robustness, realize high-efficiency calculation while guaranteeing precision, are also suitable for a reflective RIS system, and provide general guidance for RIS channel estimation.

Inventors

CHEN WEN
LIU ZIWEI
WU QINGQING

Assignees

上海交通大学

Dates

Publication Date: 20260512
Application Date: 20260331

Claims (10)

1. A channel estimation method for a transmissive reconfigurable intelligent subsurface, comprising: S1, constructing and estimating a composite channel through uplink and downlink pilot frequency transmission; S2, acquiring an antenna to TRIS channel based on a composite channel and solving through an L-BFGS algorithm; and step S3, constructing a sparse recovery problem from TRIS to a user channel, and solving the sparse recovery problem through LAOMP algorithm.
2. The channel estimation method for transmissive reconfigurable intelligent super-surface according to claim 1, wherein in step S1, the base station transmits a downlink pilot signal, and after the user equipment receives and feeds back, the base station receives an uplink signal to construct a composite channel The optimization problem is as follows: estimating composite channel using least squares method : Wherein, the Representing the total signal received by the base station; Representing a downlink pilot sequence; h represents a channel; Representing an estimate of the composite channel.
3. The method according to claim 1, wherein in step S2, the static channel from the antenna to the TRIS is decoupled from the composite channel by using the channel hardening effect, and the channel from the antenna to the TRIS is constructed Is to be solved: Wherein, the Representing a known diagonal matrix; A TRIS coefficient matrix representing the downlink of the first stage; a TRIS coefficient matrix representing the uplink of the first stage; Representing an estimate of the composite channel; K represents the number of users; Solving by adopting a limited memory quasi-Newton algorithm; In the step S3, the TRIS-to-user channel is represented in the angular domain as a sparse form: Wherein, the Representing a TRIS array response dictionary; Representing a sparse matrix; At the position of In a subframe, all user equipment simultaneously send a length of Orthogonal uplink pilot for (a) In the first place The base station receives the signal of each subframe as follows: The signals of all subframes and user equipment are collected by utilizing orthogonality processing, and the signals are obtained by vectorization: Wherein, the , ; D represents an intermediate variable; I represents an identity matrix; Representing gaussian white noise at the base station; a coefficient matrix representing TRIS in the third stage; representing a sparse dictionary matrix; representing the total signal received by the base station; Composite channel based on estimation And constructing a sparse recovery problem, and solving a TRIS to user channels.
4. The method of channel estimation for a transmissive reconfigurable intelligent subsurface of claim 3, wherein the solving using a finite memory quasi-newton algorithm comprises: Vector of complex channels Conversion to real parameter vectors Converting the solved objective function into: Wherein, the And From a matrix Real and imaginary part construction of (a): the gradient of the objective function is: Storing recently Vector pairs for multiple iterations And Implicitly updating search directions in a double-loop recursion manner So that ; Through reverse circulation and forward circulation, follow Updating parameters; Wherein, the Representing the step length, and meeting Wolfe conditions through linear search; representing the iteration number; The subscripts i, r represent the imaginary and real parts, respectively.
5. A channel estimation method for a transmissive reconfigurable intelligent super-surface according to claim 3, wherein the sparse recovery problem is: Wherein, the Representing a perception matrix; a coefficient matrix representing TRIS in the third stage; representing a sparse dictionary matrix; representing sparse vectors to be recovered; representing sparse representation of user channels under an overcomplete dictionary; defining the design of a perception matrix and the modeling of a sparse channel T-U, solving by adopting a prospective orthogonal matching pursuit algorithm, and recovering the complete TRIS to a user channel 。
6. A channel estimation system for a transmissive reconfigurable intelligent subsurface comprising a base station and a plurality of user equipments, characterized in that the base station is equipped with a TRIS transceiver; The TRIS transceiver comprises a transmission unit, a controller and a horn antenna; the projection unit controls the controller through an FPGA or a control chip; The controller stores and operates the channel estimation method for the transmissive reconfigurable intelligent super-surface according to any one of claims 1 to 5 to realize channel estimation; The horn antenna is connected to the baseband processing unit via a unique radio frequency link, and the baseband unit triggers the channel estimation process.
7. The channel estimation system for transmissive reconfigurable intelligent super-surface of claim 6, wherein said base station transmits a downlink pilot signal, and after user equipment receives and feeds back, the base station receives an uplink signal to construct a composite channel The optimization problem is as follows: estimating composite channel using least squares method : Wherein, the Representing the total signal received by the base station; Representing a downlink pilot sequence; h represents a channel; Representing an estimate of the composite channel.
8. The channel estimation system for a transmissive reconfigurable intelligent super surface of claim 6, wherein said controller uses a channel hardening effect to decouple antenna-to-TRIS static channels from composite channels to construct information about antenna-to-TRIS channels Is to be solved: Wherein, the Representing a known diagonal matrix; A TRIS coefficient matrix representing the downlink of the first stage; a TRIS coefficient matrix representing the uplink of the first stage; Representing an estimate of the composite channel; K represents the number of users; Solving by adopting a limited memory quasi-Newton algorithm; Expressing TRIS to user channels in the angular domain as a sparse form: Wherein, the Representing a TRIS array response dictionary; Representing a sparse matrix; At the position of In a subframe, all user equipment simultaneously send a length of Orthogonal uplink pilot for (a) In the first place The base station receives the signal of each subframe as follows: The signals of all subframes and user equipment are collected by utilizing orthogonality processing, and the signals are obtained by vectorization: Wherein, the , ; D represents an intermediate variable; I represents an identity matrix; Representing gaussian white noise at the base station; a coefficient matrix representing TRIS in the third stage; representing a sparse dictionary matrix; representing the total signal received by the base station; Composite channel based on estimation And constructing a sparse recovery problem, and solving a TRIS to user channels.
9. The channel estimation system for a transmissive reconfigurable intelligent super surface according to claim 8, wherein said solving using a finite memory quasi-newton algorithm comprises: Vector of complex channels Conversion to real parameter vectors Converting the solved objective function into: Wherein, the And From a matrix Real and imaginary part construction of (a): the gradient of the objective function is: Storing recently Vector pairs for multiple iterations And Implicitly updating search directions in a double-loop recursion manner So that ; Through reverse circulation and forward circulation, follow Updating parameters; Wherein, the Representing the step length, and meeting Wolfe conditions through linear search; representing the iteration number; The subscripts i, r represent the imaginary and real parts, respectively.
10. The channel estimation system for a transmissive reconfigurable intelligent super-surface of claim 8, wherein the sparse recovery problem is: Wherein, the Representing a perception matrix; a coefficient matrix representing TRIS in the third stage; representing a sparse dictionary matrix; representing sparse vectors to be recovered; representing sparse representation of user channels under an overcomplete dictionary; defining the design of a perception matrix and the modeling of a sparse channel T-U, solving by adopting a prospective orthogonal matching pursuit algorithm, and recovering the complete TRIS to a user channel 。

Description

Channel estimation method and system for transmissive reconfigurable intelligent super surface Technical Field The invention belongs to the technical field of wireless communication, in particular relates to a channel estimation method and a channel estimation system for a transmission type reconfigurable intelligent super-surface, and particularly relates to a staged channel estimation scheme for a transmission type reconfigurable intelligent super-surface wireless communication system, which utilizes a channel hardening effect to reduce estimation complexity and cost. Background The reconfigurable intelligent super surface is oriented to the sixth generation mobile communication technology, and has great potential because the reconfigurable intelligent super surface can dynamically regulate and control the electromagnetic wave propagation environment. The transmission type reconfigurable intelligent super surface (TRANSMISSIVE RECONFIGURABLE INTELLIGENT SURFACE, TRIS) overcomes the half-space coverage limit of the traditional reflection type reconfigurable intelligent super surface RIS (‌ Reconfigurable Intelligent Surface) through the unique transmission type framework, and achieves full-space coverage and flexible deployment. The transmissive reconfigurable intelligent super surface (TRIS) is a key technology for innovating wireless channel characteristics, and its performance is highly dependent on accurate channel state information. However, under the practical constraint that TRIS hardware is usually only equipped with a single radio frequency link, it is difficult to realize low-overhead and high-precision channel estimation, and practical application of TRIS system faces serious channel estimation challenges: The TRIS system relates to two sections of independent channels from antenna to TRIS (A-T) and TRIS to user (T-U), and the direct estimation of the cascade channel can introduce the problem of phase ambiguity, so that the optimal TRIS phase configuration cannot be obtained. High-dimensional estimation challenge-TRIS is typically composed of a large number of passive units, resulting in extremely high dimensions of the channel to be estimated. While TRIS transceivers are typically equipped with only a single radio frequency link, further limiting the amount of data available, making high-dimensional channel parameter estimation extremely difficult. Hardware non-idealities-quantization errors, phase noise and inter-unit cross coupling of the actual TRIS exist, and these factors deteriorate the estimation accuracy. Dynamic environmental adaptability-user movement causes rapid time-varying of the channel, requiring an estimation scheme that not only accurately acquires the channel state, but also tracks its variation with low overhead. Existing RIS channel estimation studies are many, but mostly directed to cascaded channels or reflective architectures. For example, patent literature (reconfigurable intelligent surface channel feedback system), electronic equipment and media thereof (CN 115021864A) and patent literature (RIS-based two-stage super-resolution parameter channel estimation method and apparatus (CN 117614781 a) are both channel studies based on reflective architecture reconfigurable intelligent surface communication systems. The patent literature, "large-scale antenna channel estimation method based on millimeter wave intelligent reflection surface communication" (CN 112769461A) is a study aiming at cascade channels, and converts the cascade channel estimation problem into a sparse signal recovery problem. The staged low-overhead channel estimation scheme for the hardware architecture (single radio frequency link) specific to the TRIS system and the channel model (near-field a-T channel and far-field sparse T-U channel) is still blank. Therefore, a need exists for a specific method that can efficiently and accurately separate the estimated A-T and T-U channels. Disclosure of Invention Aiming at the defects in the prior art, the invention aims to provide a channel estimation method and a system for a transmission type reconfigurable intelligent super-surface, so as to solve the problems of phase ambiguity, high complexity and high pilot frequency overhead caused by cascade channel estimation in a TRIS system. The channel estimation method for the transmissive reconfigurable intelligent super-surface provided by the invention comprises the following steps: S1, constructing and estimating a composite channel through uplink and downlink pilot frequency transmission; S2, acquiring an antenna to TRIS channel based on a composite channel and solving through an L-BFGS algorithm; and step S3, constructing a sparse recovery problem from TRIS to a user channel, and solving the sparse recovery problem through LAOMP algorithm. Preferably, in the step S1, the base station transmits a downlink pilot signal, and after the ue receives and feeds back the downlink pilot signal, the base station receives