User-Centric Network Downlink Communication Transmission Method Based on Reconfigurable Intelligent Surface

Abstract

The present invention relates to a user-centric network downlink communication transmission method based on a reconfigurable intelligent surface, belonging to the field of wireless communication. The method includes: channel estimation: estimating the direct channel between the base station and the user, and setting the reconfigurable intelligent surface to the off state; estimating the cascaded channel formed by the reconfigurable intelligent surface, where the reconfigurable intelligent surface actively sends pilot signals to the base station and the user, and then estimating the corresponding channels at the base station and the user side respectively; joint setting of data downlink transmission parameters: jointly optimizing the access relationship between the base station and the user, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface to maximize the sum rate of the system; and solving the coupling problem of the three parameters by means of alternating iteration. The present invention improves the capacity performance of the system, and at the same time realizes a relatively high system spectral efficiency and energy efficiency.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication, and relates to a downlink communication transmission method for a user-centric network based on a reconfigurable intelligent surface. Background Art

[0002] In the traditional cellular network structure, a serving cell is centered around a base station, which causes users at the cell edge to face poor signal quality, thus affecting their normal communication process. In the B5G / 6G communication scenario, with the access of a large number of user terminals, the edge problem of users has become a bottleneck restricting the improvement of the overall system performance. Due to the cell-driven communication mechanism in the cellular network, the edge problem of users is inevitable. Therefore, the cell-free network structure solves this problem by a way that all base stations jointly serve all users. However, the B5G / 6G communication scenario has the characteristics of large scale, and such a communication method brings huge energy consumption and data transmission overhead to the backhaul link and the base station side, making it unable to be applied in the actual network deployment.

[0003] The reconfigurable intelligent surface is composed of a series of passive reflection units made of special materials. By appropriately adjusting the reflection coefficient of the reflection units, the wireless propagation environment will become controllable and intelligent, and the spectral efficiency of the system is improved. Since the reconfigurable intelligent surface only passively reflects signals by adjusting the on-off of each unit diode in the reconfigurable intelligent surface, it has the characteristic of low power consumption. Therefore, it can be used to replace some base stations to provide better signal quality for users and save the overall energy consumption of the system.

[0004] In the user-centric network, a user only accesses a limited number of base stations and ensures that this access relationship can provide better performance benefits for the system. Under this access relationship, a specific communication service is established between the base station and the user. In the traditional communication scenario, the base station and the user communicate through an uncontrollable direct connection channel, resulting in insufficient or unstable received signal strength at the user end (insufficient signal strength will lead to too low rate of the receiving user to meet the user service requirements; unstable signal strength means that the direct connection channel is likely to be blocked for a short time). Therefore, a reconfigurable intelligent surface (a new controllable cascaded channel) can be added to the communication between the base station and the user to enhance the received signal strength at the user end. Summary of the Invention

[0005] In view of this, the purpose of the present invention is to provide a downlink communication transmission method for a user-centric network based on a reconfigurable intelligent surface, which can improve the capacity performance of the system, and at the same time achieve high system spectral efficiency and energy efficiency.

[0006] To achieve the above object, the present invention provides the following technical solutions:

[0007] A user-centric network downlink communication transmission method based on reconfigurable intelligent surfaces, comprising the following steps:

[0008] S1: Channel estimation, including: estimating the direct channel between the base station and the user, and setting the reconfigurable intelligent surface to the off state; estimating the cascaded channel formed by the reconfigurable intelligent surface, and having the reconfigurable intelligent surface actively send pilot signals to the base station and the user, and then estimating the corresponding channels at the base station and the user ends respectively;

[0009] S2: Determine the parameters of downlink transmission, namely the connection state c of the base station-user link b，k , the power allocation p of the base station b，k and the relationship between the reflection coefficient Θ of the reconfigurable intelligent surface and the system sum rate;

[0010] S3: Joint setting of data downlink transmission parameters: jointly optimize the access relationship between the base station and the user, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface to maximize the system sum rate; and use the alternating iteration method to solve the coupling problem of the three parameters.

[0011] Further, step S1 specifically includes: the reconfigurable intelligent surface sends pilot signals to all base stations and users, and the base station obtains the CSI of the base station-intelligent surface as g b，r according to the received pilot signals, and the user obtains the CSI of the user-intelligent surface as where CSI represents channel state information; set the reconfigurable intelligent surface to the off state, and each user sends pilot signals and channel estimation values to all base stations The base station obtains the direct link CSI as db,k according to the pilot signals; each base station broadcasts the channel estimation values g b，r and d b，k to the users; thus, both the base station end and the user end obtain the CSI of all links; the total equivalent channel is expressed as follows:

[0012]

[0013] where, Θ H is the transpose matrix of the reflection coefficient Θ of the reconfigurable intelligent surface, which contains the reflection coefficient θ of each unit of the reconfigurable intelligent surface n .

[0014] Further, in step S3, the constructed objective function is the system sum rate f(C, P, Θ), and the expression is:

[0015]

[0016] where, γ k represents the data symbol s of the user kThe signal-to-interference-plus-noise ratio (SINR) at the k-th user is expressed as:

[0017]

[0018] where B is the total number of base stations, K is the total number of users, h b，k is the equivalent channel, c b，k is the activation flag of the base station-user connection (c b，k = 1, the connection is activated; c b，k = 0, the connection is not activated); p b，k is the power allocated by the base station for the data symbol s k of the user when the user connection corresponding to the b-th base station is in the active state, is the white noise power;

[0019] The sum rate of f(C, P, Θ) is transformed into a fractional form by using the Lagrangian dual transform (LDT) technique where, is the adjustment weight factor, α k is the intermediate variable factor in the Lagrangian dual transform.

[0020] Furthermore, in step S3, the joint setting of the data downlink transmission parameters specifically includes the following steps:

[0021] S31: The reflection coefficient of each unit of the reconfigurable intelligent surface randomly takes a value from m preset discrete values as the initial parameter;

[0022] S32: The base station establishes the access relationship C between the base station and the user according to the large-scale fading of each user;

[0023] S33: Perform utility-based base station power allocation;

[0024] S34: Adjust the reflection coefficient Θ of the reconfigurable intelligent surface.

[0025] Furthermore, in step S32, the establishment of the base station-user connection specifically includes the following steps:

[0026] S321: Each base station arranges the large-scale channel fading coefficients between it and all users in ascending order;

[0027] S322: Each base station activates the connections of the first M users with smaller channel fading, and sets the activation flag c b，k of the base station-user connection to 1; if the connection is in the non-activated state, then c b，k = 0; obtain the connection matrix C containing all base station-user connection states.

[0028] Further, in step S33, the power allocation of the base station specifically includes the following steps:

[0029] S331: When only considering the power allocation of the base station, the fractional summation term of the sum rate f1(C, P, Θ) is converted into the form of a quadratic concave function f2(P) only about the power allocation P by means of fractional programming to solve its non-convexity;

[0030] S332: Use the first-order Taylor expansion to obtain an approximate lower bound of the user rate, so that the minimum rate constraint is converted into an approximate convex constraint;

[0031] S333: Use the Lagrange multiplier method to write the Lagrangian function L(P, λ) of the convex quadratic programming problem, and then let The optimal power allocation P including the power of each link can be obtained by solving the system of equations.

[0032] Further, in step S34, adjusting the reflection coefficient Θ of the reconfigurable intelligent surface specifically includes the following steps:

[0033] S341: In the reflection coefficient optimization sub-problem, set the power allocation P of the base station and the access relationship C between the base station and the user to take the results obtained in steps S32 and S33 respectively, and the sum rate f1(C, P, Θ) is converted into a quadratic concave function f3(Θ) only about the reflection coefficient matrix Θ by means of fractional programming;

[0034] S342: Use the method of the first-order Taylor expansion to convert the minimum rate constraint of the user into an approximate convex constraint;

[0035] S343: Use the Lagrange multiplier method to write the Lagrangian function L(Θ, λ) of the programming problem, and then let The reflection coefficient matrix Θ can be obtained;

[0036] S344: According to the optimization result θ of each unit reflection coefficient obtained in S343 n The control unit of the reconfigurable intelligent surface selects the value closest to θ from m preset discrete phase shift coefficients [0, 2Π / m,..., 2(m - 1)Π / m] to control the reflection coefficient of the nth unit of the intelligent surface. n

[0037] The beneficial effects of the present invention are as follows:

[0038] (1) By adding the reconfigurable intelligent surface in the present invention, the received signal strength at the user end is enhanced. Therefore, the communication system assisted by the reconfigurable intelligent surface usually has better capacity performance than the traditional communication system.

[0039] ​(2) Based on the large-scale fading coefficients of each base station-user link, the base station selects the top M users with stronger service channel state information to optimize the base station-user access relationship, which can effectively reduce the overhead of the fronthaul network, the computational complexity at the base station side, and the signaling overhead of communication between the base station and users in a large-scale network, thereby enhancing the feasibility of network practical deployment.

[0040] (3) By jointly optimizing the access relationship between the base station and users, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface, the sum rate of the system gradually approaches a stable and better value.

[0041] (4) In the downlink communication mechanism of the present invention, the reconfigurable intelligent surface can change the signal propagation environment by setting different reflection coefficients. Combining with the optimization of the base station-user access relationship and the optimization of the base station power allocation, the capacity performance of the system can approach the capacity performance in the full-connected scenario of the base station and users.

[0042] (5) Since the downlink communication mechanism in the present invention achieves better system capacity performance with a sparser base station-user connection, by setting appropriate parameters, this mechanism can further optimize the energy efficiency performance of the system on the premise of achieving better system spectral efficiency performance, which is also the core of this patent.

[0043] Other advantages, objectives, and features of the present invention will be described to some extent in the subsequent specification, and to some extent, will be obvious to those skilled in the art based on the study of the following text, or can be taught from the practice of the present invention. The objectives and other advantages of the present invention can be achieved and obtained through the following specification. BRIEF DESCRIPTION OF THE DRAWINGS

[0044] In order to make the objectives, technical solutions, and advantages of the present invention clearer, the present invention will be described in detail preferably with reference to the accompanying drawings, where:

[0045] Figure 1 is the communication system scenario of the present invention;

[0046] Figure 2 is the information interaction diagram of users, base stations, reconfigurable intelligent surfaces, and central control units;

[0047] Figure 3 is the overall flowchart of downlink communication parameter setting;

[0048] Figure 4 is the base station-user interaction mode;

[0049] Figure 5 is the scenario schematic diagram of the embodiment of the present invention;

[0050] Figure 6 Schematic diagram of the connection result of base station users;

[0051] Figure 7 System spectrum efficiency simulation result;

[0052] Figure 8 System energy efficiency simulation result. Specific implementation manners

[0053] The following uses specific specific examples to illustrate the implementation manners of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific implementation manners. Various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that the diagrams provided in the following embodiments only illustrate the basic concept of the present invention in a schematic manner. Without conflict, the following embodiments and the features in the embodiments can be combined with each other.

[0054] Please refer to Figures 1 to 8 , Figure 1 which is the communication system scenario diagram of the present invention. As Figure 1 shown, multiple users communicate with multiple base stations with the assistance of a reconfigurable intelligent surface. In a user-centric network, a user will only access a limited number of base stations and ensure that this access relationship can provide better performance benefits for the system. Under this access relationship, a specific communication service is established between the base station and the user. In a traditional communication scenario, the base station and the user communicate through an uncontrollable direct connection channel, and the reconfigurable intelligent surface adds a new controllable cascaded channel to the communication between the base station and the user. By appropriately adjusting the reflection coefficient of the intelligent surface, the received signal strength at the user end can be significantly enhanced.

[0055] Under the above scenario, by jointly designing the access relationship of base station users, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface, the capacity performance of the system can be improved. At the same time, the determination of these three parameters ensures the integrity of the downlink communication process. The setting of the three parameters determines the performance of the system, and at the same time, they will affect each other. Therefore, it is difficult to reasonably set the three parameters. The present invention proposes a downlink communication mechanism that can effectively solve this problem and at the same time achieve high system spectrum efficiency and energy efficiency.

[0056] The present invention designs a user-driven downlink communication mechanism based on a reconfigurable intelligent surface. As Figure 2 shown, it mainly includes the parameter setting process of channel estimation and data downlink transmission.

[0057] 1) The process of channel estimation.

[0058] 2) The downlink data transmission control needs to complete three operations, namely: determining the access relationship between the base station and users, allocating the base station power, and setting the reflection coefficient of the reconfigurable intelligent surface. There is a coupling relationship among these three, and they need to be jointly designed to improve the system capacity. In this downlink communication mechanism, we use an alternating iteration method to solve this problem. Once the access relationship is established, the establishment of communication services will only exist between the connected base station and users.

[0059] The information interaction process among users, base stations, reconfigurable intelligent surfaces, and the central control unit is as Figure 3 shown:

[0060] Channel estimation process:

[0061] (1) The intelligent surface sends pilot signals to users and base stations.

[0062] (2) Users and base stations respectively estimate the channel state information (CSI) of user-intelligent surface and base station-intelligent surface according to the received pilot signals.

[0063] (3) Set the intelligent surface to the off state, and all users send pilot signals and the estimated user-intelligent surface CSI through the direct link.

[0064] (4) The base station receives the pilot signals sent by users and estimates the direct link CSI. At this time, the base station obtains the CSI of the entire communication link.

[0065] (5) All base stations broadcast the direct link CSI and the base station-intelligent surface CSI to users, and users obtain the CSI of all links.

[0066] Data downlink transmission process:

[0067] (6) The base station sorts the large-scale fading values of the channels with each user in ascending order. According to the sorting result, the base station selects the access relationships of the top M active users. Among them, the large-scale fading coefficient related to the spatial positions of the base station and users is used for sorting.

[0068] (7) The base station sends the labels of the connected users to the central control unit, and the central control unit transmits the data symbols s of these users k to the base station. For users without established connections, the central control unit will not transmit data symbols s for them k .

[0069] (8) The base station allocates transmission power to the connected users and sends downlink data symbols s to users through the direct link and the cascaded link k . The base station only allocates power to the connected users, and the signal received by the users is the total signal strength passing through the direct channel and the cascaded channel.

[0070] (9) The reconfigurable intelligent surface adjusts the reflection coefficient to improve the signal strength received by the user.

[0071] The specific content of the channel estimation process is:

[0072] The specific channel estimation process of the present invention is as follows: the reconfigurable intelligent surface sends a pilot signal to all base stations and users, and the base station obtains the channel state information (CSI) of the base station-intelligent surface according to the received pilot signal: b，r , the user obtains the CSI of the user-smart surface according to the received pilot signal: The reconfigurable smart surface is set to the off state, and each user sends a pilot signal and channel estimation value to all base stations. The base station obtains the direct link CSI based on the pilot signal: b，k Each base station broadcasts the channel estimation value g to the user b，r and d b，k At this point, both the base station and the user end obtain the CSI of all links. The total equivalent channel can be expressed as follows:

[0073]

[0074] Among them, Θ H is the transposed matrix of the reflection coefficient θ of the reconfigurable smart surface, which contains the reflection coefficient θ of each unit n .

[0075] Specific content of the data downlink transmission process:

[0076] The downlink transmission process of user data mainly involves the setting of three parameters, namely: base station-user link connection state c b，k , power allocation of base station p b，k and the reflectance Θ of the reconfigurable smart surface.

[0077] a) The relationship between the three parameters and the system and rate

[0078] When the user connection corresponding to the b-th base station is in an activated state, the central control unit will k The data symbol is transmitted to the base station, and then the base station allocates power p to the data symbol. b，k Sent to user, c b，k The activation flag of the base station-user connection (c b，k =1, connection activated; c b，k = 0, the connection is not activated). After the b-th base station establishes a connection relationship with all users and allocates the transmission power of the users, it can obtain the signal x sent by the base station. b .

[0079] After passing through the equivalent channel h b，k , the signal received by the k-th user is

[0080]

[0081] where z k is the additive white Gaussian noise (AWGN) at the k-th user, B is the total number of base stations, and K is the total number of users. The data symbol s of the user k The signal-to-interference-plus-noise ratio (SINR) at the k-th user is expressed as The sum rate of the system is expressed as:

[0082]

[0083] The sum rate is an important indicator to measure the system capacity performance. Therefore, the problem of improving the communication system performance can be equivalent to the problem of finding the maximum sum rate of the system. In the downlink communication mechanism of the present invention, the base station power allocation and the reflection coefficient adjustment of the intelligent surface are both based on the utility method, that is, the optimal solution of this programming problem is used as the result of each update. In this programming problem, the objective function is the sum rate of the system To ensure the QoS of each user, the user needs to satisfy a minimum rate constraint. Considering the limited connection ability of the base station, each base station needs to send data to the user with a limited power size. At the same time, the number of users connected to each base station also needs to be limited. The intelligent surface needs to satisfy the set phase shift constraint conditions. To simplify the objective function f(C, P, Θ) of the original programming problem, the present invention uses the technical means of Lagrangian dual transformation (LDT) to transform it into a fractional sum form

[0084] The sum rate f1(C, P, Θ) of the system depends on the base station-user access relationship, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface. At the same time, the settings of these three parameters will affect each other. To solve this problem, the downlink communication mechanism in the present invention decomposes the original problem into three sub-problems: the establishment of base station-user connection, the power allocation of the base station, and the reflection coefficient adjustment of the reconfigurable intelligent surface, and uses the idea of alternating iteration to obtain a better parameter setting. To solve the difficulties brought by the objective function f1(C, P, Θ) and the user minimum rate constraint to the parameter setting, the present invention also uses the technical means of fractional programming and first-order Taylor expansion.

[0085] b), The overall process of parameter setting

[0086] The overall process of downlink communication parameter setting in the present invention is as Figure 3As shown in the figure, in the first step, the reflection coefficient of each unit of the reconfigurable intelligent surface randomly takes a value from m preset discrete values as the initial parameter; in the second step, the base station establishes the access relationship C between the base station and the user according to the large-scale fading of each user; in the third step, utility-based base station power allocation is performed; the fourth step is the adjustment of the reflection coefficient Θ of the reconfigurable intelligent surface. After the updates in the first to fourth steps are completed, the base station monitors the rate of the user and then uploads it to the central control unit to calculate the sum rate of the system. The central control unit records the sum rate of the system for two adjacent times. If the change in the two results is less than a preset difference δ th Then the update process for this cycle ends.

[0087] To ensure the real-time performance of this downlink communication mechanism, this process will be repeated in a cycle T. In this way, even if the user moves, the communication system can track the real-time and reasonable parameter settings. The specific processes of each parameter setting are given in the following parts respectively.

[0088] ① Establishment of base station-user connection:

[0089] Both the power allocation of the base station and the optimization of the reflection coefficient of the reconfigurable intelligent surface need to be carried out on the premise of determining the base station-user connection. Therefore, the optimization of the base station-user connection is the sub-problem to be solved first. The present invention proposes a strategy for establishing a connection relationship based on channel state information. The small-scale fading between the base station and the user changes relatively fast, and the reflection coefficient of the reconfigurable intelligent surface is optimized to offset the total channel fading and will also be updated relatively fast. In contrast, the large-scale fading of the direct connection channel only depends on the spatial positions of the user and the base station, and its change frequency will be much lower than the previous two. To ensure the stability and feasibility of the actual system communication, the frequency of base station-user connection switching cannot be too high. Therefore, in the present invention, the large-scale fading between the base station and the user is considered as the basis for establishing the connection. The specific steps are as follows:

[0090] In the first step, each base station arranges the large-scale channel fading coefficients between it and all users in ascending order;

[0091] In the third step, each base station activates the connections of the first M users with smaller channel fading, and makes c b，k = 1; if the connection is in an inactive state, then c b，k = 0. A connection matrix C containing all base station-user connection states is obtained.

[0092] The interaction relationship between the base station and the user during the base station-user connection establishment is as Figure 4As shown. When a user does not establish a service relationship with any base station, the user will be in an idle state. At this time, the user only consumes a very small amount of power to maintain this state, generally less than 15 mW. Once the user needs to access the network, the base station will establish a connection with the relevant user according to the above connection establishment method based on the channel state. After the connection is established, the base station will page the relevant users connected to it. At this time, the user is in a connected state and can establish services with the base station. When the user is in a connected state, since it needs to constantly monitor whether there is data transmission, it will consume a relatively high power, usually 1000 - 3500 mW. In Figure 2 For the connected state of a user, the more base stations that establish services with it, the more power it needs to consume to maintain this connected state. At this time, if a certain base station does not page the user and establish services with it, then the user is actually in an idle state for this base station. In short, the power required for each user to maintain the connected state depends on the number of base stations that establish services with it.

[0093] ② Power allocation of the base station:

[0094] The power allocation of the base station proposed in the downlink communication mechanism of the present invention is a utility-based scheme. Taking the sum rate of the system as a reference index, the power allocation result of each base station to its served users is obtained when the sum rate reaches the maximum value. That is, this power allocation scheme can effectively optimize the capacity performance of the system. To solve the problem of the mutual influence of three parameters in this downlink communication mechanism, when optimizing the power allocation, the access relationship between the base station and the user can be obtained through the previous process, and the reflection coefficient of each unit of the reconfigurable intelligent surface randomly takes a value among the preset m phase shift control coefficients. The specific method is as follows:

[0095] In the first step, when only considering the power allocation of the base station, the fractional summation term of the objective function f1(C, P, Θ) can be transformed into the form of a quadratic concave function f2(P) only about the power allocation P by means of fractional programming to solve its non-convexity.

[0096] In the second step, each achievable user rate γ k is also in the form of a fraction. The downlink communication mechanism in the present invention uses the first-order Taylor expansion to obtain an approximate lower bound of the user rate, so that the minimum rate constraint is converted into an approximate convex constraint. After solving the above two problems, the original programming problem becomes a convex quadratic programming problem with inequality constraints.

[0097] In the third step, the Lagrangian function L(P, λ) of the convex quadratic programming problem is written using the Lagrange multiplier method, and then let The optimal power allocation matrix P including the power of each link can be obtained by solving the system of equations.

[0098] ③Adjustment of the reflection coefficient of the reconfigurable surface:

[0099] The reconfigurable intelligent surface can change the signal propagation environment by adjusting the reflection coefficient, thereby affecting the strength of the signal received by the user terminal. Therefore, the setting of the reflection coefficient of the intelligent surface directly affects the capacity performance of the entire system. The reflection coefficient matrix Θ of the intelligent surface composed of N reflection units is a diagonal matrix diag(θ1,..., θ N ), and each element is the reflection coefficient of each unit where the amplitude coefficient and the phase shift coefficient are β n ∈[0, 1] and Φ n ∈[0, 2Π). In practical applications, in order to simplify the control complexity of the intelligent surface, we take the amplitude coefficient β n of each unit to be 1, and Φ n takes m discrete phase shift control coefficients in [0, 2Π) with a step of 2Π / m. The reflection coefficient optimization in the downlink communication mechanism of the present invention also adopts a utility-based optimization scheme, and the goal is still to maximize the system sum rate. The algorithm is completed in the control unit (FPGA, Field Programmable Gate Array) of the reconfigurable intelligent surface, without consuming the computing power of the base station side. The specific steps of the algorithm are as follows:

[0100] First step, in the reflection coefficient optimization sub-problem, set the power allocation P of the base station and the access relationship C between the base station and the user to the results obtained in the previous two steps. The original programming problem f1(C, P, Θ) can also be converted into a quadratic concave function f3(Θ) only about the reflection coefficient matrix Θ by means of fractional programming;

[0101] Second step, use the first-order Taylor expansion method to convert the minimum rate constraint of the user into an approximate convex constraint;

[0102] Third step, use the Lagrange multiplier method to write the Lagrangian function L(Θ, λ) of the programming problem, and then let The reflection coefficient matrix Θ can be obtained;

[0103] Fourth step, according to the optimized result θ n of each unit reflection coefficient obtained in the third step, the control unit of the reconfigurable intelligent surface selects the value closest to θ n from the m preset discrete phase shift coefficients [0, 2Π / m,..., 2(m - 1)Π / m] to control the reflection coefficient of the nth unit of the intelligent surface. Specific embodiment:

[0105] The present invention lists a typical embodiment, such as Figure 5As shown in the figure, 4 base stations are evenly distributed on a semi-circular arc with a radius of 100m centered at (100m, 0). 4 users are distributed within a circular area with a radius of 2m centered at the center of the semi-circle, and their specific coordinates are (98.62m, -0.80m), (99.62m, 1.56m), (99.93m, -0.05m), and (101.3m, -0.79m). The reconfigurable intelligent surface is located 5m directly below the center of the semi-circle, with the center coordinates of (100m, -5m).

[0106] According to the propagation environment proposed by 3GPP, the path losses of the cascaded link and the direct link can be calculated by two formulas 35.6 + 22lg(d) and 32.6 + 36.7lg(d) respectively, where d is the distance between two communication nodes. For small-scale fading, the cascaded link is modeled as a Rice fading model, which consists of the LOS part and the non-LOS part. The LOS part is determined by the angle of arrival / departure of the signal. Here, we adopt a reconfigurable intelligent surface with a uniform linear array (ULA), and the response formula of the ULA with N elements can be directly substituted to calculate the value of this part. Since the LOS part of the direct link may be blocked, it is modeled as a Rayleigh fading model.

[0107] Based on the above formulas, we can calculate the large-scale and small-scale fading values corresponding to each base station-user link, and thus obtain the equivalent channel coefficient of each link. Each base station establishes connections with M users with relatively small large-scale fading in the direct link, and a general base station-user access relationship can be obtained. Under the above base station-user access relationship, the base station side executes the power allocation scheme proposed in the present invention to determine the specific power allocation for each user's data. Then, the control unit of the reconfigurable intelligent surface executes the optimization scheme of the reflection coefficient. The period of the process of establishing the above base station-user access relationship, allocating the base station power, and adjusting the reflection coefficient is T, and the process of each period is as Figure 4 shown.

[0108] In this example, the result of the base station-user access relationship is as Figure 6 shown, where the base station-users in the same area indicate that the access relationship is in an active state. It can be seen from the figure that each base station serves at most three users simultaneously. Figure 7 and Figure 8 show that under the maximum power limit P max = 18dBm of each base station and the phase shift limit |θ n|≤1, simulation results of the system spectral efficiency and energy efficiency with the number of reflecting elements of the reconfigurable intelligent surface ranging from 10 to 40. We respectively compare the results of the user-centric network and the base station-user fully connected network, where the base station-user fully connected network refers to the connection paradigm in which each user is connected to all base stations. The rate value of the user-centric network is similar to the sum rate performance of the base station-user fully connected network. It is worth noting that after optimizing the access relationship in the user-centric network, the overhead of the fronthaul network and the signaling between the base station and the user can be saved, and the data processing complexity of the central control unit and the base station also decreases to a certain extent. In addition, the simulation results show that the downlink communication mechanism of the user-centric network in the present invention has better energy efficiency performance compared with the base station-user fully connected network. As the number of reflecting elements of the reconfigurable intelligent surface increases, both the spectral efficiency and energy efficiency of the system increase to a certain extent, indicating that within a certain range, the increase in the scale of the reflecting elements of the reconfigurable intelligent surface can bring greater benefits to the system performance.

[0109] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the preferred embodiments, those of ordinary skill in the art should understand that the technical solutions of the present invention can be modified or equivalently replaced without departing from the spirit and scope of the present technical solution, and they should all be covered within the scope of the claims of the present invention.

Claims

1. A downlink communication transmission method for a user-centric network based on reconfigurable intelligent surfaces, characterized in that The method includes the following steps: S1: Channel estimation, including: estimating the direct channel between the base station and the user, and setting the reconfigurable intelligent surface to the off state; estimating the cascaded channel formed by the reconfigurable intelligent surface, and having the reconfigurable intelligent surface actively send pilot signals to the base station and the user, and then estimating the corresponding channels at the base station and the user ends respectively; S2: Determine the parameters for downlink transmission, namely the base station-user link connection status , the power allocation of the base station and the reflection coefficient of the reconfigurable intelligent surface and their relationship with the system sum rate; S3: Joint setting of data downlink transmission parameters: jointly optimizing the access relationship between the base station and the user, the power allocation of the base station, and the reflection coefficient of the reconfigurable intelligent surface to maximize the sum rate of the system; and solving the coupling problem of the three parameters by means of alternating iteration; Step S1 specifically includes: The reconfigurable intelligent surface sends pilot signals to all base stations and users. The base stations obtain the CSI of the base station-intelligent surface based on the received pilot signals as , and the users obtain the CSI of the user-intelligent surface based on the received pilot signals as , where CSI represents the channel state information; The reconfigurable intelligent surface is set to the off state, and each user sends pilot signals and channel estimates to all base stations , and the base stations obtain the direct link CSI as ; Each base station broadcasts the channel estimates to the users and ; At this point, both the base station side and the user side obtain the CSI of all links; The total equivalent channel is expressed as follows: wherein, is the reflection coefficient of the reconfigurable intelligent surface and its transposed matrix, which contains the reflection coefficients of each unit of the reconfigurable intelligent surface ; In step S3, the constructed objective function is the sum rate of the system , and the expression is as follows: Among them, represents the data symbol of the user The signal-to-interference-plus-noise ratio expression at the k-th user is as follows: wherein, B is the total number of base stations, K is the total number of users, is the equivalent channel, is the activation flag of the base station-user connection; is the power allocated by the base station for the data symbol of the user when the user connection corresponding to the b-th base station is in the active state, and is the white noise power; Use the Lagrangian dual transformation technique to transform into the sum rate in fractional form , where is the adjustment weight factor, , is the intermediate variable factor in the Lagrangian dual transformation; The joint setting of data downlink transmission parameters specifically includes the following steps: S31: Randomly select a value from m preset discrete values as the initial parameter for the reflection coefficient of each unit of the reconfigurable intelligent surface; S32: The base station establishes the access relationship C between the base station and the user according to the large-scale fading of each user; S33: Perform utility-based base station power allocation; S34: Adjust the reflection coefficient of the reconfigurable intelligent surface ; In step S32, the establishment of the connection between the base station and the user specifically includes the following steps: S321: Each base station arranges the large-scale channel fading coefficients between it and all users in ascending order; S322: Before each base station activates the connections of the M users with smaller channel fading, set the activation flag of the base station-user connections ; If the connection is in an inactive state, then ; Obtain the connection matrix C containing the connection status of all base station-user connections; In step S33, the power allocation of the base station specifically includes the following steps: S331: When only considering the base station power allocation, the sum rate of the fractional summation terms is transformed into a quadratic concave function only about the power allocation by means of fractional programming to solve the non-convexity it brings; S332: Use the first-order Taylor expansion to obtain an approximate lower bound of the user rate, so that the minimum rate constraint is converted into an approximate convex constraint; S333: Write the Lagrangian function of the convex quadratic programming problem using the Lagrange multiplier method , and then let , , and obtain the optimal power allocation P including the power of each link by solving the system of equations; In step S34, the reflection coefficient of the reconfigurable intelligent surface is adjusted , which specifically includes the following steps: S341: In the reflection coefficient optimization sub-problem, set the power allocation P of the base station and the access relationship C between the base station and the user to the results obtained in steps S32 and S33 respectively, and the sum rate Use the method of fractional programming to transform it into a quadratic concave function that only depends on the reflection coefficient matrix ; ; S342: Use the method of first-order Taylor expansion to convert the minimum rate constraint of the user into an approximate convex constraint; S343: Write the Lagrangian function of the programming problem using the Lagrange multiplier method , and then let , to obtain the reflection coefficient matrix ; S344: Optimize the results of the reflection coefficient of each unit obtained in S343 , the control unit of the reconfigurable intelligent surface selects from m preset discrete phase shift coefficients the one closest to to control the reflection coefficient of the nth unit of the intelligent surface.

Images