A beamforming method based on partial channel state information in active ris

Abstract

The application discloses a beamforming design method under active reconfigurable intelligent surface (RIS) assisted communication based on partial channel state information, and belongs to the technical field of wireless communication physical layer. Based on the partial channel state information of a two-hop channel (from a transmitter to a RIS, and from the RIS to a receiver), the application first deduces an analytical expression of a lower bound of an average sum rate and an average transmit power of an active reconfigurable intelligent surface, then establishes an optimization model for maximizing the average sum rate of a system while guaranteeing the average transmit power of the RIS end and the transmit power constraint of a base station, and finally obtains the beamforming of the base station end and the RIS end based on gradient descent and alternating direction multiplier technology. Simulation results show that the beamforming method provided by the application can effectively reduce the performance loss caused by the partial channel state information, thereby providing a solution to the problem that the two-hop channel state information cannot be independently identified under active reconfigurable intelligent surface assisted communication.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication physical layer technology, and relates to beamforming technology, array signal processing technology, active reconfigurable intelligent planar coverage enhancement technology, and more specifically, to a beamforming design method for a reconfigurable intelligent surface-assisted communication system based on partial channel state information. Background Technology

[0002] Reconfigurable smart plane technology has rapidly attracted widespread attention and research from academia and industry due to its programmability, ease of deployment, low overhead, and low power consumption, and is expected to become one of the key technologies for future 6G mobile communication networks. To address the severe multiplicative fading effect in passive RIS-assisted systems, a novel active RIS architecture has been proposed. Unlike passive RIS technology, by configuring additional amplification circuits for each phase-shifting circuit-driven reflective element, active RIS technology can not only adjust the phase of the reflected signal to form a directional beam, but also amplify the power of the reflected signal to mitigate the multiplicative fading effect. Therefore, efficient beamforming design based on active RIS technology relies on independent, high-precision channel state information for the two-hop channels. However, perfectly obtaining the independent channel state information for the two-hop channels is quite difficult. In traditional passive RIS-assisted systems, passive beamforming technology only needs to utilize the cascaded channel state information from the transmitter to the RIS and then to the receiver to satisfy transverse mode constraints. Extending this technology to an active RIS-assisted system requires addressing two issues: (1) how to consider beamforming schemes when two-hop channels cannot be perfectly acquired in an active RIS; and (2) how to perform joint power allocation between the base station and the RIS. Summary of the Invention

[0003] The purpose of this invention is to design a beamforming scheme for active RIS (Radio Router System), which can reduce the performance loss caused by only using partial channel state information; and maximize the system average sum rate to improve the quality of service by jointly allocating the base station and RIS transmit power.

[0004] The technical solution of this invention is as follows: First, based on partial channel state information of a two-hop channel, analytical expressions for the user average rate and the RIS average transmit power are derived. Second, a system and average rate maximization problem based on joint base station and RIS transmit power allocation is constructed, and optimization sub-problems for base station beamforming and active RIS beamforming are constructed respectively based on an alternating optimization method. Finally, gradient descent and alternating direction multiplier techniques are used to update the base station beamforming and active RIS beamforming respectively, and the optimal beamforming scheme is obtained through iterative optimization. To achieve the above objectives, this invention adopts the following technical solution:

[0005] A beamforming method based on partial channel state information in an active RIS includes the following steps:

[0006] Step S1, System Configuration

[0007] Step S101, Signal Transmission Model:

[0008] The base station is equipped with N antennas to transmit downlink data. Given K single-antenna users, where the transmitted signal satisfies A RIS consisting of M active reflective elements is configured between the base station and the user to construct an additional communication link and increase spatial diversity gain. The base station precoding matrix is ​​as follows: And the base station's transmission power meets the requirements. Where P N That is the maximum transmission power of the base station. and Let represent the channels from the base station to the active RIS, and from the base station and the active RIS to user k, respectively. Then the received signal for the k-th user is:

[0009]

[0010] in, Let n be the RIS reflection matrix. k z and z are the additive white Gaussian noise at the active RIS and the user receiver, respectively, and they follow the distribution. and and This refers to the corresponding noise power. The emitter power of an active RISC. Where w = [w1, ..., w M ] H This is the beamforming vector for the active RIS terminal. The achievable rate for the k-th user is...

[0011]

[0012] Step S102, Channel Model:

[0013] Channel H from base station to RIS dr and RIS to user channel H r,k Follows Rice distribution:

[0014]

[0015]

[0016] in, It is the Rice factor. It is a distance-dependent large-scale fading coefficient, a direct path component. and The angle of the direct path is determined. The component of the indirect path follows a Rayleigh distribution, i.e. and Where Σ B ≥0 is the spatial correlation matrix of the antenna at the base station, which has all 1 diagonal elements. R ≥0 and Σ r,k ≥0 represent RIS reflector arrays and and The relevant spatial correlation matrix has all-1 diagonal elements. Therefore, and It can be modeled as and in and

[0017] The cascaded channel between the base station and the user can be represented as: The direct link from the base station to the user is usually long and the communication environment is complex. Assume the direct link channel from the base station to the user... It follows a Rayleigh distribution.

[0018] Step S2: Obtain the lower bound expression for the user average rate:

[0019] Based on the partial channel information from RIS to the user channel given in step S102, namely the distribution of the angle-related direct path component, the distance-related path loss, and the non-direct path component, the lower bound of the user's average rate can be obtained. Parsing the expression:

[0020]

[0021] in, and I is the identity matrix.

[0022] Step S3: Obtain the analytical expression for the average transmit power of RIS:

[0023] Based on the partial channel information of the base station to RIS channel given in step S102, namely the distribution of the angle-related direct path component, the distance-related path loss, and the non-direct path component, the distribution of random channel H can be derived. dr Analytical expression for the average transmit power of RIS

[0024]

[0025] in, and

[0026] Step S4, Beamforming Design Method:

[0027] Based on the analytical expression of the lower bound of the user average rate obtained in step S2 And the analytical expression for the average transmit power of RIS obtained in step S3 The beamforming design of the base station and active RIS is jointly implemented to maximize system performance and data rate while meeting the maximum transmit power requirements of both the base station and the active RIS. The resulting beamforming design problem is as follows:

[0028]

[0029]

[0030]

[0031]

[0032] Among them, P N and P M These are the maximum transmit power of the base station and the RIS, respectively, and constraint 3 is the amplification gain constraint of the RIS reflection coefficient.

[0033] Step S401: Decomplexing the non-concave rate expression:

[0034] Because the objective function The non-concave nature of the objective function makes the modeled beamforming optimization problem difficult to solve. Therefore, a minimax technique is used to construct a simple and easily solvable concave lower bound for the non-concave objective function. Specifically, at the fixed point {F} in the nth iteration... n ,w n At}, construct Lower bound substitution function as follows:

[0035]

[0036] in,

[0037]

[0038]

[0039] Lower bound substitution function It is a quasi-concave function of F and w, so an alternating optimization method is used to jointly design the beamforming of the base station and the active RIS.

[0040] Step S402: Obtain the easily optimized equivalent base station beamforming design problem:

[0041] Combining the quadratic substitution function obtained in step S401, we can obtain the equivalent base station beamforming design subproblem, which is easier to solve through optimization, as follows:

[0042]

[0043]

[0044]

[0045] in, and The optimal solution F for the above problem can be obtained using the gradient descent method.

[0046] Step S403: Obtain the easily optimized equivalent RIS beamforming design problem:

[0047] Combining the quadratic substitution function obtained in step S401, we can obtain the equivalent RIS-end beamforming design subproblem, which is easier to solve through optimization, as follows:

[0048]

[0049] stw H D w w≤P M

[0050]

[0051] The above problem can be solved by using the alternating direction multiplier method to obtain a local optimal solution w.

[0052] The innovation and advantages of this invention compared to traditional solutions are as follows:

[0053] This invention presents a robust beamforming method for active RIS-assisted systems, solving the problem of channel state information uncertainty caused by the imperfect identification of two-hop channels in active RIS-assisted systems. This method achieves green and energy-efficient communication while ensuring user service quality. Compared with traditional solutions, the designed method effectively mitigates performance loss caused by channel uncertainty, ensuring stable and reliable communication with an extremely low interruption probability, and significantly improving user service quality. Attached Figure Description

[0054] Figure 1 A flowchart for designing a beamforming method based on partial channel state information in an active RIS.

[0055] Figure 2 This is a schematic diagram of the active and passive RIS-assisted downlink communication systems studied in this invention.

[0056] Figure 3This is a comparison diagram of the active robust beamforming method proposed in this invention with three comparative methods (active RIS, non-robust active RIS, and passive RIS beamforming schemes based on the lower bound of ergodic rate) in terms of ergodic system and rate.

[0057] Figure 4 This is a comparison diagram of the active robust beamforming method proposed in this invention with five comparative methods (active RIS based on the lower bound of ergodic rate, non-robust active RIS, active RIS based on perfect channel, passive RIS, and beamforming scheme without RIS) in terms of ergodic system and rate. Detailed Implementation

[0058] The effectiveness and practicality of the present invention will be demonstrated below with reference to the accompanying drawings and simulation examples. Obviously, the described simulation examples are only some embodiments of the present invention, and not all embodiments:

[0059] The specific implementation method of this invention consists of the following steps:

[0060] Step 1: Determine system parameters, including the number of base station antennas N, the number of RIS reflectors M, the number of users K; the locations of the base stations, RIS, and users, and their corresponding communication distances and path angles; the carrier frequency; and the large-scale fading coefficient β = -PL0-10αlog 10 (d) dB, road loss index α, Rice factor δ0=…=δ K Spatial correlation matrix Σ at base station and RIS B Σ R and Σ r,k ; Maximum amplification gain a for each reflective element of RIS max Circuit power P DC P c and P RF noise power

[0061] Step 2: Calculate the channel from the base station to the user based on the system parameters in Step 1. Channel H from base station to RIS dr and its direct diameter component RIS to user channel h r,k and its direct path components Cascaded Channels Channel matrix Initialize the RIS beamforming vector w.

[0062] Step 3: Given the RIS beamforming matrix w, update the base station beamforming matrix F using the gradient descent method. The update method for the base station beamforming matrix and dual variables in the gradient descent method is as follows:

[0063] In the v-th iteration of the gradient descent method Representing the dual variables, the beamforming matrix can be updated as follows:

[0064]

[0065] Accordingly, the dual variable is updated according to the following rules:

[0066]

[0067]

[0068] in, This is the step size for gradient descent (taken as 0.5).

[0069] Step 4: Based on the beamforming of the F fixed base station obtained in Step 3, update the RIS beamforming vector w using the alternating direction multiplier method. The update method for the RIS-end beamforming vector and dual variables using the alternating direction multiplier method is as follows:

[0070] In the (i+1)th iteration of the alternating multiplier method, the beamforming vector w at the RIS end... [i+1] It can be represented as

[0071]

[0072] Among them, u [i] ,η [i] Let be the dual variable for the i-th iteration, and its update rule is as follows:

[0073]

[0074] η [i+1] =η [i] +w [i+1] -u [i+1]

[0075] Step 5: Alternately repeat steps 3 and 4 until the sequence of objective function values ​​generated by the iteration converges.

[0076] Simulation Example Analysis

[0077] For a fair comparison, the total power consumption of the RIS and base station in the simulation also includes circuit power consumption. That is, the total power consumption of the RIS is P. RIS =P M +M(P c +P DC ), where P c It is the power consumption of the phase shift circuit, P DC It is DC biased; the total power consumption of the base station is P. BS =P N+NP RF , where P RF This refers to the power consumption of the radio frequency circuit. In the simulation, the system parameters are shown in Table 1:

[0078] Table 1 System Simulation Parameters

[0079]

[0080] Figure 3 Performance curves of the traversal reachability of the invented robust beamforming method and three benchmark algorithms as a function of the number of users are presented, where the number of RIS reflection elements M is set to 32.

[0081] Figure 4 Performance curves of the traversal reachability of the invented robust beamforming method and five benchmark algorithms as a function of the number of RIS reflective elements are presented, where the number of users K is set to 4.

[0082] from Figure 3 It can be seen that the proposed traversal rate lower bound closed expression scheme can perfectly fit the traversal rate, and the proposed robust scheme has a significant system rate improvement compared with the non-robust method. This means that the proposed scheme can alleviate the uncertainty caused by some channel state information of the two-hop channel.

[0083] from Figure 4 It can be seen that, compared with a system without RIS assistance, a passive RIS can increase the system rate gain by 59.69%, while an active RIS that consumes the same system power can increase the system rate gain to 688.57%, which greatly improves the quality of user communication services.

Claims

1. A beamforming method based on partial channel state information in an active RIS, characterized in that, Includes the following steps: Step S1: Perform system configuration, constructing the signal transmission model and channel model; wherein, the system configuration includes: base station configuration. One antenna transmits downlink data to Single-antenna user, active RIS configuration One reflection unit; the base station precoding matrix is... The maximum transmission power of the base station is The beamforming vector of the active RIS is The maximum transmit power of the active RIS is The reflection coefficient amplitude of each reflecting unit satisfies ,in Maximum amplification gain; The signal transmission model is based on and The channel model is constructed, including the channel from the base station to the active RIS. Active RIS to users channel and base station to user direct link channel ; Step S2: Based on partial channel state information of the two-hop channel, derive the analytical expression for the user average rate. Analytical expression for RIS average transmit power ; Step S3: Construct the system and average rate maximization problem based on joint base station and RIS transmit power allocation, and construct optimization sub-problems for base station beamforming and active RIS beamforming respectively based on the alternating optimization method; Step S4: Update the base station beamforming and active RIS beamforming, and obtain the optimal beamforming scheme through iterative optimization. Step S3 includes the following steps: Step S301, Beamforming Design Method for Base Station and RIS End: Based on the analytical expression of the lower bound of the user average rate obtained in step S2 And the analytical expression for the average transmit power of RIS obtained in step S2 The beamforming design of the base station and active RIS was jointly developed to maximize system performance and data rate while meeting the maximum transmit power requirements of both the base station and the active RIS. The beamforming design problem is as follows: ； Step S302: Decomplexing the non-concave rate expression: In the Fixed point of the next iteration Location, Structure Lower bound substitution function ,as follows: ； in, , , , This represents the large-scale fading coefficient. Rice factor, From active RIS to users The direct diameter component, This represents the noise power at the active RIS. For users Noise power at the location, To and Related extended vectors; ； Lower bound substitution function It is about and The quasi-concave function is used, therefore an alternating optimization method is adopted to jointly design the beamforming of the base station and the active RIS; Step S303: Obtain the easily optimized equivalent base station beamforming design problem: Combining the quadratic substitution function obtained in step S302, the equivalent base station beamforming design subproblem, which is easier to solve through optimization, is obtained as follows: ； in, and ; The large-scale fading coefficient from the base station to the active RIS is given. Rice factor from base station to active RIS For the direct path component from the base station to the active RIS, and These are the spatial correlation matrices at the base station and the active RIS, respectively; Step S304: Obtain the easily optimized equivalent RIS beamforming design problem: Combining the quadratic substitution function obtained in step S302, the equivalent RIS-end beamforming design subproblem, which is easier to solve through optimization, is obtained as follows: 。 2. A beamforming method based on partial channel state information in an active RIS according to claim 1, characterized in that, Step S1 includes the following steps: Step S101: Construct a signal transmission model: Base station configuration One antenna, transmitting downlink data Give There are single-antenna users, whose transmitted signals satisfy... ; The base station and the user are configured with... A RIS consisting of active reflective units is used to build additional communication links and increase spatial diversity gain. The base station precoding matrix is And the base station's transmission power meets the requirements. ,in This is the maximum transmit power of the base station; , and These represent the connection from the base station to the active RIS, and from the base station and active RIS to the user, respectively. The channel; then the first The received signal for each user is ； in, For the RIS reflection matrix, and These are additive white Gaussian noise from the active RIS and the user receiver, respectively, and they follow a distribution. and , and This refers to the corresponding noise power; the emitter power of an active RISC. ;in For the beamforming vector of the active RIS end; the first The achievable rate for each user is ； Step S102: Construct the channel model: Base station to RIS channel and RIS to users channel Follows Rice distribution: , ； in, It is the Rice factor. It is a distance-dependent large-scale fading coefficient, a direct path component. and The angle of the direct path is determined; the non-direct path component follows a Rayleigh distribution, i.e. and ,in It is the spatial correlation matrix of the antenna at the base station, with all diagonal elements being 1. and These are RIS reflector arrays and and The relevant spatial correlation matrix has all-1 diagonal elements; therefore, the non-direct path and It can be modeled as and ,in and ; The cascaded channel between the base station and the user is represented as: Assuming a direct link channel from the base station to the user. It follows a Rayleigh distribution.

3. A beamforming method based on partial channel state information in an active RIS according to claim 2, characterized in that, Step S2 includes the following steps: Step S201: Obtain the lower bound expression for the user average rate: Based on the partial channel information from RIS to the user channel given in step S102, namely the distribution of the angle-related direct path component, the distance-related path loss, and the non-direct path component, the lower bound of the user average rate is obtained. Parsing the expression: ； in, and , It is the identity matrix; Step S202: Obtain the analytical expression for the average transmit power of RIS: Based on the partial channel information of the base station to RIS channel given in step S102, namely the distribution of the angle-related direct path component, the distance-related path loss, and the non-direct path component, the distribution based on random channels is derived. Analytical expression for the average transmit power of RIS : ； in, and .

4. A beamforming method based on partial channel state information in an active RIS according to claim 1, characterized in that, Gradient descent and alternating directional multiplier techniques are used to update base station beamforming and active RIS beamforming, respectively.

Images