A beam training optimization method, system and device for RIS-assisted MIMO system

Abstract

A beam training optimization method, system and device for a RIS-assisted MIMO system, comprising the following steps: constructing a reconfigurable intelligent surface RIS-assisted MIMO system; taking maximizing the system receiving signal-to-noise ratio as the optimization goal, taking joint beam angle information and transmitting beam angle information as optimization variables, and designing an optimization model containing three optimization variables; designing a DFT codebook based on the beam angle information, and verifying it by comparing it with perfect beam alignment through simulation. The present invention aims at the problem that RIS-associated channel information is difficult to obtain, and adopts a beam-trained RIS-assisted communication system, but there is still a lack of comprehensive, complete, reliable and practical research solutions. This method can solve the energy consumption problem of emerging and future wireless networks, and the system performance corresponding to the optimized variables is close to the perfect beam alignment situation.

Description

Technical Field

[0001] The present invention belongs to the technical field of reconfigurable intelligent surface-assisted communication, and in particular relates to a beam training optimization method, system and device for a RIS-assisted MIMO system. Background Art

[0002] Reconfigurable intelligent surface (RIS) uses a large number of passive reflective elements to customize the wireless communication environment and can be deployed on building facades, rooms, factory ceilings, computer housings or dense networks. Due to its advantages such as low cost, high energy efficiency and full-duplex, it is considered a promising technology for the next generation of wireless communications, especially for millimeter wave (mmWave) communications between base stations and users that are susceptible to blocking and propagation losses.

[0003] At present, there have been a lot of studies on RIS in wireless communications, which mainly focus on performance analysis, beamforming optimization, physical layer security, channel estimation, etc., all of which require channel estimation or known channel state information (CSI). The acquisition of CSI is a prerequisite for fully realizing the potential of RIS-assisted mmWave systems. RIS cannot receive and transmit signals and lacks signal processing capabilities, so it is difficult to obtain complete CSI for BS-RIS links and RIS-User links. In addition, the size of the transceiver antenna array and the large number of passive components in RIS result in a large amount of training overhead for CSI acquisition. Therefore, in order to obtain sufficient beamforming gain without the need for channel estimation, multiple nodes in the system can perform beam training, and the designed codebook can be used to obtain the best beam alignment between nodes in the system.

[0004] On the other hand, existing RIS-based beam training research focuses on improving signal gain without considering the limitation of system energy efficiency. However, system energy consumption is a hot topic in emerging and future wireless networks, and system energy efficiency has become a key performance indicator to ensure green and sustainable wireless networks.

[0005] Therefore, the actual evaluation system of RIS-assisted MIMO system based on beam training still lacks a complete, reliable and practical research plan, and it is necessary to consider green and sustainable constraints such as power and energy efficiency.

[0006] In summary, the prior art has the following defects and deficiencies:

[0007] 1. RIS-assisted MIMO systems often require channel estimation or known channel state information. However, in practical applications, channel estimation is often too complex, resulting in inaccurate estimation or difficulty in obtaining channel information.

[0008] 2. Emerging and future wireless networks focus more on system energy consumption, but the beam training schemes for RIS-assisted MIMO systems often do not focus on energy efficiency;

[0009] 3. Existing solutions to the beam training problem of RIS-assisted communication systems often result in suboptimal solutions, which are far from perfect beam alignment. Summary of the invention

[0010] The object of the present invention is to provide a beam training optimization method, system and device for a RIS-assisted MIMO system to solve the problem that channel estimation is often too complex, resulting in inaccurate estimation and poor perfect beam alignment.

[0011] To achieve the above object, the present invention adopts the following technical solutions:

[0012] A beam training optimization method for a RIS-assisted MIMO system comprises the following steps:

[0013] Construct a reconfigurable intelligent surface RIS-assisted MIMO system, which has base stations and users in the system, and satisfies the communication barrier of the direct channel from the base station to the user in the system;

[0014] Taking maximizing the system receiving signal-to-noise ratio as the optimization goal, taking the joint beam angle information and the transmit beam angle information as the optimization variables, and taking the optimization variable codebook, RIS phase shift, transmit power at the transmitter and system energy efficiency as the constraints, an optimization model with three optimization variables is designed.

[0015] Design the DFT codebook based on the beam angle information, solve the optimization model, obtain the three angle information parameters after optimization, and analyze the impact of power and energy efficiency constraints on system performance;

[0016] The three optimized angle information parameters are combined with the achievable rate of the system and verified by comparing with the perfect beam alignment through simulation.

[0017] Furthermore, the reconfigurable smart surface RIS-assisted MIMO system includes a multi-antenna base station BS, a RIS of a reflection unit and a multi-antenna user UE. The antenna arrays of the multi-antenna base station BS and the multi-antenna user UE are uniform linear arrays. The multi-antenna base station BS controls the phase shift of the RIS unit through a reconfigurable smart surface RIS controller.

[0018] Furthermore, when there are obstacles or deep fading in the direct channel from the base station to the user, or when communication is impossible due to long distance, communication is achieved with the help of RIS cascade channel; long distance is defined as: the signal received by the receiving end cannot reach the receiving threshold, resulting in unsuccessful communication.

[0019] Furthermore, the optimization model is specifically expressed as:

[0020]

[0021] stΩ ′ u ∈[-1,1],Ω ′ R ∈[-1,1],Φ ′ R ∈[-1,1].

[0022] [ψ] n,n |=1.

[0023]

[0024]

[0025] Among them, P max Indicates the maximum transmission power of the transmitter; η m represents the energy efficiency constraint of the system; SNR is the system signal-to-noise ratio.

[0026] Furthermore, the system signal-to-noise ratio:

[0027]

[0028] Where p is the user transmit power; represents the phase shift matrix of RIS, β = 1 means RIS lossless reflection; h B,R and h R,u Indicates the channel information of the RIS-BS and UE-RIS link; ω B and ω u Represents the antenna beam vector of the base station and the user end, which is related to the optimization variable; N UE ,N BS and N RIS Respectively represent the number of antennas at the user and base station ends and the number of reflection units of RIS; g r,B ,g u,r represents the Rayleigh fading channel gain of the RIS-BS and UE-RIS links; a RIS (ψ R ,Φ R ),a u (ψ u ) represent the relative steering vector of RIS and the steering vector information of the user respectively; a RIS (ψ ′ R ,Φ ′ R ),a u (ψ ′ u ) represents the joint beam vector of RIS and the transmit beam vector of the user.

[0029] Furthermore, we solve the optimization model with three optimization variables:

[0030] For three optimization variables and their range constraints Ω′ u ∈[-1,1],Ω′ R ∈[-1,1],Φ′ R ∈[-1,1], design a uniformly distributed codebook vector, solve the system receiving signal-to-noise ratio corresponding to all cases through an exhaustive search scheme, and further calculate the system achievable rate;

[0031] When only the variable range constraint and RIS phase shift constraint are considered, the RIS-UE distance and the system achievable rate are solved for the changes under different numbers of reflection units;

[0032] Add transmission power constraints and energy consumption constraints to solve the changes in RIS-UE distance and system achievable rate under different constraint parameters.

[0033] Further verification is as follows:

[0034] Through simulation, the relationship between the RIS-UE distance and the achievable rate of the system is compared under different numbers of RIS reflector units. The closer the RIS-UE distance is, the higher the achievable rate of the system is; and the more RIS reflector units there are, the closer the curve is to perfect beam alignment.

[0035] Furthermore, a beam training optimization system for a RIS-assisted MIMO system includes:

[0036] System building module, used to build a reconfigurable intelligent surface RIS-assisted MIMO system, the system has base stations and users, and the direct channel from the base station to the user in the system has communication barriers;

[0037] The optimization model building module is used to maximize the system receiving signal-to-noise ratio as the optimization goal, take the joint beam angle information and the transmit beam angle information as the optimization variables, and take the optimization variable codebook, RIS phase shift, transmit power of the transmitter and system energy efficiency as constraints to design an optimization model with three optimization variables;

[0038] The optimization module is used to design the DFT codebook based on the beam angle information, solve the optimization model, obtain the three angle information parameters after optimization, and analyze the impact of power and energy efficiency constraints on system performance;

[0039] The verification module is used to verify the three optimized angle information parameters in combination with the achievable rate of the system by comparing them with the perfect beam alignment through simulation.

[0040] Furthermore, a computer device includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, the steps of a beam training optimization method for a RIS-assisted MIMO system are implemented.

[0041] Compared with the prior art, the present invention has the following technical effects:

[0042] The present invention is aimed at the scenarios that may exist in the practical application of the RIS-assisted communication system (such as obstacles in the direct link or the reflection link, ultra-long-distance transmission coverage with a small number of base stations, or indoor-outdoor communication, etc.), and the RIS-assisted communication system can solve this problem;

[0043] The present invention aims to solve the problem that RIS associated channel information is difficult to obtain, and adopts a beam-trained RIS-assisted communication system. However, there is still a lack of comprehensive, complete, reliable, and practical research solutions. This method can solve the energy consumption problem of emerging and future wireless networks, and the system performance corresponding to the optimized variables is close to the perfect beam alignment situation.

[0044] For the optimization problem with three optimization variables, the present invention adopts an exhaustive search scheme to obtain the optimal solution of system performance, which is close to perfect beam alignment;

[0045] In summary, the present invention can optimize the angle information associated with the departure angle and arrival angle of the user and the RIS node by designing a codebook, which is used to solve the performance improvement problem of the RIS-assisted wireless communication system based on beam training under constraints such as energy efficiency, and can be close to a perfect beam alignment situation. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] Figure 1 It is the system model designed by the present invention.

[0047] Figure 2 It is the RIS-UE distance vs. the achievable rate of the system under different numbers of RIS reflection units and perfect beam alignment in the present invention.

[0048] Figure 3 It is the RIS-UE distance vs. the achievable rate of the system under different energy efficiency constraints, power constraints and no energy efficiency power constraints in the present invention. DETAILED DESCRIPTION

[0049] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0050] A beam training optimization method, system and device for a RIS-assisted MIMO system, the described system includes a multi-antenna base station BS, a RIS with N reflection units and a multi-antenna user UE, the BS and UE antenna arrays are uniform linear arrays, and the BS controls the phase shift of the RIS unit through a RIS controller. When there are obstacles or deep fading in the direct channel from the base station to the user, or the communication distance is too far to communicate, it is necessary to use the RIS cascade channel to achieve communication.

[0051] In the actual system, the positions of the BS and RIS are determined, that is, the transmission angle (AOD) and arrival angle (AOA) related to the dominant path of the RIS-BS link are known. In order to achieve perfect beam alignment in the system, beam training involves the estimation of AOD and AOA related to the dominant path of the UE-RIS link. By designing the codebook of the UE's beamforming vector and the RIS's phase shift vector related angle information (involving the azimuth and elevation angles of AOD and AOA), the receiving gain of the system is improved through optimization. Note that for RIS, the goal is not to find AOA and AOD, but to find an optimal "relative reflection angle" to achieve beam alignment.

[0052] The optimization problem is proposed: maximizing the system receiving signal-to-noise ratio is the optimization goal, and the RIS-related joint beam angle information ψ ′ R (θ R ,ν R ), φ ′ R (ν R ), transmit beam angle information related to the user To optimize the variables, the optimization variable codebook, RIS phase shift, transmit power at the transmitter, and system energy efficiency are taken as constraints, and an optimization problem with three optimization variables is designed; (θ R ,ν R They represent the azimuth and elevation angles corresponding to the relative reflection angle of RIS, θ u ,ν u Respectively represent the azimuth and elevation angles corresponding to the UE's corresponding emission angle, where ν u =v R )

[0053] Solve the optimization problem: Design a DFT codebook based on angle information, propose an exhaustive search method to solve the above optimization problem, and analyze the impact of power and energy efficiency constraints on system performance;

[0054] Simulation verification problem: The three angle information parameters after optimization are obtained, combined with the system's achievable rate, and compared with the perfect beam alignment through simulation. The results show that the system's achievable rate corresponding to the optimized parameters is close to that of the perfect alignment, which reflects the effectiveness of this method.

[0055] like Figure 1 As shown, the system model designed by the present invention is a RIS-assisted MIMO system, including a base station, a RIS, and a multi-antenna user, and the system noise obeys a complex Gaussian distribution. The additive Gaussian white noise is used, and the direct channel is not considered, only the RIS cascade channel is considered. For uplink information transmission, based on the design of the angle information-related codebook, the system performance is improved through an exhaustive search scheme to achieve beam alignment. An optimization problem is proposed: Taking the maximization of the system received signal-to-noise ratio as the optimization goal, the RIS-related joint beam angle information ψ ′ R (θ R ,ν R ), φ ′ R (ν R ), transmit beam angle information related to the user In order to optimize the variables, the optimization variable codebook, RIS phase shift, transmit power of the transmitter and system energy efficiency are taken as constraints, and an optimization problem with three optimization variables is designed.

[0056] The system's received signal-to-noise ratio SNR is

[0057]

[0058] Where p is the user transmit power; represents the phase shift matrix of RIS, β = 1 means RIS lossless reflection; h B,R and h R,u Indicates the channel information of the RIS-BS and UE-RIS link; ω B and ω u Represents the antenna beam vector of the base station and the user end, which is related to the optimization variable; N UE ,N BS and N RIS Respectively represent the number of antennas at the user and base station ends and the number of reflection units of RIS; g r,B ,g u,r represents the Rayleigh fading channel gain of the RIS-BS and UE-RIS links; a RIS (ψ R ,Φ R ),a u (ψ u ) represent the relative steering vector of RIS and the steering vector information of the user respectively; a RIS (ψ ′ R ,Φ ′ R ),a u (ψ ′ u) represents the joint beam vector of RIS and the transmit beam vector of the user.

[0059] The optimization problem can be expressed as

[0060]

[0061] stΩ ′ u ∈[-1,1],Ω ′ R ∈[-1,1],Φ ′ R ∈[-1,1].

[0062] [ψ] n,n |=1.

[0063]

[0064]

[0065] Among them, P max Indicates the maximum transmission power of the transmitter; η m Represents the energy efficiency constraint of the system.

[0066] Solve the above optimization problem with three optimization variables:

[0067] For three optimization variables and their range constraints Ω′ u ∈[-1,1],Ω′ R ∈[-1,1],Φ′ R ∈[-1,1], design a uniformly distributed codebook vector, solve the system receiving signal-to-noise ratio corresponding to all cases through an exhaustive search scheme, and further calculate the system achievable rate;

[0068] When only the variable range constraints and RIS phase shift constraints are considered, the changes in RIS-UE distance and system achievable rate under different numbers of reflection units are solved and analyzed to facilitate performance analysis when power and energy consumption constraints are added in the next step.

[0069] Based on 2, transmit power constraints and energy consumption constraints are added to solve the changes in RIS-UE distance and system achievable rate under different constraint parameters and analyze them.

[0070] Figure 2 Through simulation, the relationship between the RIS-UE distance and the achievable rate of the system is compared under different numbers of RIS reflector units. The closer the RIS-UE distance is, the higher the achievable rate of the system is; and the more RIS reflector units there are, the closer the curve is to perfect beam alignment. Figure 3Compare the relationship between RIS-UE distance and system achievable rate under different system energy efficiency and power. Figure 2 If power constraints are added on this basis, the curve moves downward; after adding energy efficiency constraints, there is a theoretical upper limit to the rate that the system can achieve.

Claims

1. A beam training optimization method for a RIS-assisted MIMO system, characterized in that: The following steps are involved: Construct a reconfigurable intelligent surface RIS-assisted MIMO system, which has base stations and users in the system, and satisfies the communication barrier of the direct channel from the base station to the user in the system; Taking maximizing the system receiving signal-to-noise ratio as the optimization goal, taking the joint beam angle information and the transmit beam angle information as the optimization variables, and taking the optimization variable codebook, RIS phase shift, transmit power at the transmitter and system energy efficiency as the constraints, an optimization model with three optimization variables is designed. Design the DFT codebook based on the beam angle information, solve the optimization model, obtain the three angle information parameters after optimization, and analyze the impact of power and energy efficiency constraints on system performance; The optimized three angle information parameters are combined with the system's achievable rate and verified by comparing with the perfect beam alignment through simulation. The optimization model is specifically expressed as: stΩ′ u ∈[-1,1],Ω′ R ∈[-1,1],Φ′ R ∈[-1,1]. |[ψ] n,n |=1. Among them, P max Indicates the maximum transmission power of the transmitter; η m represents the energy efficiency constraint of the system; SNR is the system signal-to-noise ratio; System signal-to-noise ratio: Where p is the user transmit power; represents the phase shift matrix of RIS, β = 1 means RIS lossless reflection; h B,R and h R,u Indicates the channel information of the RIS-BS and UE-RIS links; ω B and ω u Represents the antenna beam vector of the base station and the user end, which is related to the optimization variable; N UE ,N BS and N RIS Respectively represent the number of antennas at the user and base station ends and the number of reflection units of RIS; g r,B ,g u,r represents the Rayleigh fading channel gain of the RIS-BS and UE-RIS links; a RIS (ψ R ,Φ R ),a u (ψ u ) represent the relative steering vector of RIS and the steering vector information of the user respectively; a RIS (ψ′ R ,Φ′ R ),a u (ψ′ u ) represents the joint beam vector of RIS and the transmit beam vector of the user; Solve the optimization model with three optimization variables: For three optimization variables and their range constraints Ω u ∈[-1,1],Ω R ∈[-1,1],Φ R ∈[-1,1], design a uniformly distributed codebook vector, solve the system receiving signal-to-noise ratio corresponding to all cases through an exhaustive search scheme, and further calculate the system achievable rate; When only the variable range constraint and RIS phase shift constraint are considered, the RIS-UE distance and the system achievable rate are solved for the changes under different numbers of reflection units; Add transmission power constraints and energy consumption constraints to solve the changes in RIS-UE distance and system achievable rate under different constraint parameters.

2. The beam training optimization method of a RIS-assisted MIMO system according to claim 1, characterized in that: The reconfigurable intelligent surface RIS-assisted MIMO system includes a multi-antenna base station BS, a RIS of a reflection unit and a multi-antenna user UE. The antenna arrays of the multi-antenna base station BS and the multi-antenna user UE are uniform linear arrays. The multi-antenna base station BS controls the phase shift of the RIS unit through a reconfigurable intelligent surface RIS controller.

3. The beam training optimization method of a RIS-assisted MIMO system according to claim 1, characterized in that: When there are obstacles or deep fading in the direct channel from the base station to the user, or when communication is impossible due to long distance, communication is achieved with the help of RIS cascade channel; long distance is defined as: the signal received by the receiving end cannot reach the receiving threshold, resulting in unsuccessful communication.

4. The beam training optimization method of a RIS-assisted MIMO system according to claim 1, characterized in that: The verification is as follows: Through simulation, the relationship between the RIS-UE distance and the achievable rate of the system is compared under different numbers of RIS reflector units. The closer the RIS-UE distance is, the higher the achievable rate of the system is; and the more RIS reflector units there are, the closer the curve is to perfect beam alignment.

5. A beam training optimization system for a RIS-assisted MIMO system, characterized in that: include: System building module, used to build a reconfigurable intelligent surface RIS-assisted MIMO system, the system has base stations and users, and the direct channel from the base station to the user in the system has communication barriers; The optimization model building module is used to maximize the system receiving signal-to-noise ratio as the optimization goal, take the joint beam angle information and the transmit beam angle information as the optimization variables, and take the optimization variable codebook, RIS phase shift, transmit power of the transmitter and system energy efficiency as constraints to design an optimization model with three optimization variables; The optimization module is used to design the DFT codebook based on the beam angle information, solve the optimization model, obtain the three angle information parameters after optimization, and analyze the impact of power and energy efficiency constraints on system performance; The verification module is used to verify the three optimized angle information parameters by comparing them with the perfect beam alignment through simulation in combination with the achievable rate of the system; The optimization model is specifically expressed as: stΩ′ u ∈[-1,1],Ω′ R ∈[-1,1],Φ′ R ∈[-1,1]. |[ψ] n,n |=1. Among them, P max Indicates the maximum transmission power of the transmitter; η m represents the energy efficiency constraint of the system; SNR is the system signal-to-noise ratio; System signal-to-noise ratio: Where p is the user transmit power; represents the phase shift matrix of RIS, β = 1 means RIS lossless reflection; h B,R and h R,u Indicates the channel information of the RIS-BS and UE-RIS link; ω B and ω u Represents the antenna beam vector of the base station and the user end, which is related to the optimization variable; N UE ,N BS and N RIS Respectively represent the number of antennas at the user and base station ends and the number of reflection units of RIS; g r,B ,g u,r represents the Rayleigh fading channel gain of the RIS-BS and UE-RIS links; a RIS (ψ R ,Φ R ),a u (ψ u ) represent the relative steering vector of RIS and the steering vector information of the user respectively; a RIS (ψ′ R ,Φ′ R ),a u (ψ′ u ) represents the joint beam vector of RIS and the transmit beam vector of the user; Solve the optimization model with three optimization variables: For three optimization variables and their range constraints Ω′ u ∈[-1,1],Ω′ R ∈[-1,1],Φ′ R ∈[-1,1], design a uniformly distributed codebook vector, solve the system receiving signal-to-noise ratio corresponding to all cases through an exhaustive search scheme, and further calculate the system achievable rate; When only the variable range constraint and RIS phase shift constraint are considered, the RIS-UE distance and the system achievable rate are solved for the changes under different numbers of reflection units; Add transmission power constraints and energy consumption constraints to solve the changes in RIS-UE distance and system achievable rate under different constraint parameters.

6. A computer device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that: When the processor executes the computer program, the steps of the beam training optimization method for a RIS-assisted MIMO system as claimed in any one of claims 1 to 4 are implemented.

Images