An intelligent reflecting surface selection and phase matrix adjustment method

Abstract

The present invention provides a method for intelligent reflecting surface selection and phase matrix adjustment, belonging to the field of wireless communication. The method of the present invention includes: establishing connections between each user equipment and the base station, calculating the ratio of the average signal power to the average interference-plus-noise signal power of each user equipment, and selecting the intelligent reflecting surface associated with the user equipment according to this ratio; after determining the association mode of the base station-intelligent reflecting surface-user equipment, the user equipment performs 1-bit feedback indication according to the change of the received signal power intensity, and the base station serving this user equipment adjusts the phase matrix of the associated intelligent reflecting surface according to the received feedback indication. The method of the present invention avoids the estimation and storage overhead of the instantaneous channel state information between the base station / access point and the intelligent reflecting surface, does not require the user to provide location information, and solves the problem of selecting an intelligent reflecting surface for users in a distributed multi-base station multi-intelligent reflecting surface multi-user system.

Description

Technical Field

[0001] The present invention belongs to the field of wireless communication, and specifically refers to a method for intelligent reflecting surface selection and phase matrix adjustment based on large-scale fading parameters and 1-bit feedback. Background Art

[0002] Intelligent Reflecting Surface (IRS) technology, also known as reconfigurable intelligent surface technology, has become one of the hot air interface technologies in 6G research due to its low cost, low power consumption, and high reliability. An intelligent reflecting surface is a two-dimensional electromagnetic material surface composed of multiple reflecting units, and the reflecting electromagnetic characteristics of each unit on the reflecting surface can be regulated by means of circuit settings. Deploying an intelligent reflecting surface in a wireless network to assist communication can reconstruct the wireless channel between the transmitter and the receiver, thereby improving channel fading and suppressing interference. Its passive and radio-frequency link-free characteristics help the system reduce costs, complexity, and power consumption.

[0003] Beamforming, also known as beam shaping, spatial domain filtering, etc., uses the principles of constructive interference and destructive interference to generate interference by adjusting the amplitudes and phases of the transmitted signals on different antennas, so that the final signal is transmitted or received in a certain direction or several directions. The transmitted / received signals of multiple antennas or antenna arrays can be regarded as the superposition of multiple signals. By adjusting the amplitudes and phases of each signal, destructive interference can be carried out at some positions to reduce the power of the sum of the signals to zero or even lower, while constructive interference can be carried out at other positions to increase the power of the sum of the signals to a peak or even higher.

[0004] For beamforming in the IRS-assisted wireless communication scenario, in addition to the active beamforming gain of the base station, the reflecting units of the IRS provide passive beamforming gain, and the two need to be jointly designed. To make full use of the passive beamforming gain, the number of reflecting units of the IRS is usually large. If the channel state information (CSI) of both the access point (AP) to IRS channel and the IRS to user equipment (UE) channel is updated in real time, the required channel estimation overhead is unacceptable. Some researchers choose to use the method of random phase shift optimization for design, but it is limited to single IRS systems. At the same time, due to the product distance path loss phenomenon, the deployment location of the IRS will also have a greater impact on the actual effect. However, there is relatively little research on the distributed deployment of multiple IRSs at present.

[0005] The invention patent application with the publication number CN113993180A was published on January 28, 2022, and discloses a method for base station and intelligent reflecting surface selection based on minimizing multiplicative path loss. As Figure 1As shown, the method maintains a base station record table at the base station side for the nearest and the second nearest base stations to each intelligent reflecting surface. The table content includes the base station numbers of the two nearest and the second nearest base stations to each intelligent reflecting surface, and the corresponding base station-intelligent reflecting surface channel state information. The user measures the SINR (signal-to-interference-plus-noise ratio) values of each base station and selects the base station with the maximum SINR as the initial access base station. The user terminal device is equipped with a global positioning system and reports its location information and the measured SINR result through signaling feedback. The initial access base station finds the intelligent reflecting surface closest to the user as the candidate intelligent reflecting surface, and then selects the base station-intelligent reflecting surface-user link for data transmission according to the distance between the candidate intelligent reflecting surface and the user and the SINR threshold. This method realizes the integration of passive intelligent reflecting surfaces into the existing network, sets a service radius for the intelligent reflecting surface, reasonably reduces the search range, and reduces the signaling overhead and delay. However, there are the following problems with this technology:

[0006] (1) The existing technical solution requires the base station to continuously maintain the base station record table. The table content includes the base station numbers of the two nearest and the second nearest base stations to each intelligent reflecting surface, and the corresponding channel state information of the base station-intelligent reflecting surface. Among them, obtaining the channel state information of the base station-intelligent reflecting surface requires a large amount of channel estimation overhead. However, when the user reports that the SINR is higher than the given threshold, the existing technology does not use the intelligent reflecting surface to assist the user's communication, resulting in low resource utilization.

[0007] (2) The existing technical solution assumes that the user terminal device is equipped with a global positioning system and requires the user to provide accurate location information to the base station. However, the global positioning system has certain errors, which may cause misjudgment of the nearest location relationship on the base station side. In addition, the existing technical solution only relies on the location relationship for the matching of the base station and the intelligent reflecting surface, and does not fully consider the influence of shadow fading and other factors in the large-scale fading characteristics. Summary of the Invention

[0008] Aiming at the above-mentioned shortcomings of the existing technology, the present invention proposes an intelligent reflecting surface selection and phase matrix adjustment method for the problem of intelligent reflecting surface selection for serving users in a distributed multi-base station, multi-intelligent reflecting surface, multi-user system. Under the premise of phase alignment of active and passive beamforming, the present invention method calculates the ratio of the average signal power to the average interference and noise signal power by using the transmit power of the base station and the large-scale fading parameters of each channel to determine the intelligent reflecting surface-user association scheme; on the basis of determining the intelligent reflecting surface-user association, a phase matrix adjustment scheme for the reflecting units of the intelligent reflecting surface based on 1-bit feedback is proposed.

[0009] Specifically, an intelligent reflecting surface selection and phase matrix adjustment method of the present invention is applied to a scenario where multiple base stations serve multiple user devices with the assistance of multiple intelligent reflecting surfaces. Each base station serves only one user device, and each base station-user device pair is passively beamformed by only one intelligent reflecting surface. The method of the present invention includes the following steps:

[0010] (1) Select the intelligent reflecting surface associated with the user, including:

[0011] (1.1) Each user device receives the reference signals of each base station and establishes a connection between the user device and the base station according to the base station reference signal strength;

[0012] (1.2) Each base station obtains the channel state information of the downlink from the base station to the user device it serves, the large-scale fading coefficient of the channel from the base station to each user device, the large-scale fading coefficient of the channel from the base station to each intelligent reflecting surface, and the large-scale fading coefficient of the channel from each intelligent reflecting surface to each user device; each user device obtains the noise variance when receiving its own signal;

[0013] (1.3) Assume that each base station performs maximum ratio transmission beamforming only according to the downlink channel state information of the user it serves. For each user device, randomly select an association method between the intelligent reflecting surface and the user device, and adjust the phases of the reflection units of each intelligent reflecting surface to align the phases of the signals after passing through the base station-intelligent reflecting surface-user device channel and the base station-user device channel. Calculate the average signal power and average interference power of each user device, and the ratio of the average signal power to the average signal-to-noise-and-interference power (ASAINR); where the ratio of the average signal power to the average signal-to-noise-and-interference power refers to the ratio of the average signal power of the user device to the sum of the average interference power and the noise power;

[0014] Record the minimum value of the ratio of the average signal power to the average signal-to-noise-and-interference power of the user device under each association method between the intelligent reflecting surface and the user device;

[0015] (1.4) Change the association method between the intelligent reflecting surface and the user of each user device to obtain a new set of association methods between the intelligent reflecting surface and the user device, and then go back to step 1.3 to recalculate the ratio of the average signal power to the average signal-to-noise-and-interference power of each user device; traverse all possible association methods between the intelligent reflecting surface and the user device, and select the set of association methods between the intelligent reflecting surface and the user with the largest minimum value of the ratio of the average signal power to the average signal-to-noise-and-interference power of the user device as the final output, and output the obtained set of association methods between the base station-intelligent reflecting surface-user device;

[0016] (2) Adjust the phase matrix of the intelligent reflecting surface based on 1-bit feedback, including:

[0017] (2.1) Determine the association method of each base station-intelligent reflecting surface-user equipment. Each base station randomly initializes the phases of the reflection units of the intelligent reflecting surface associated with it and sets them to the phase shift values corresponding to the current highest received signal power. Each user equipment initializes the received signal power and sets it to the current highest received signal power.

[0018] (2.2) Iteratively update the phases of the reflection units of each intelligent reflecting surface.

[0019] Randomly change the phases of all the reflection units of each intelligent reflecting surface. The associated base station sends signals to the user equipment. Each user equipment updates the received signal power. If the current received signal power is greater than the current highest received signal power, the user equipment feeds back an indication of 1 and updates the current highest received signal power to the current received signal power. Otherwise, the user equipment feeds back an indication of 0 and keeps the current highest received signal power unchanged. When the base station receives a feedback indication of 1 from the user equipment, it keeps the phases of the reflection units of the current intelligent reflecting surface and updates the phase shift values corresponding to the current highest received signal power. When the base station receives a feedback indication of 0 from the user equipment, it restores the phases of the reflection units of the intelligent reflecting surface to the state at the previous iteration and keeps the phase shift corresponding to the current highest received signal power unchanged.

[0020] (2.3) Repeat step 2.3 to continuously update the phases of the reflection units of each intelligent reflecting surface and the highest received signal power of each user equipment until the preset number of iterations or the preset signal-to-interference-plus-noise ratio threshold of the user equipment is reached, and output the final highest received signal power of each user equipment and the phase of the reflection units of each intelligent reflecting surface.

[0021] The advantages and positive effects of the present invention are as follows:

[0022] Compared with the existing solutions, the present invention avoids the estimation and storage overhead of the instantaneous channel state information between the base station / access point and the intelligent reflecting surface, and at the same time does not require the user to provide location information, solving the problem of intelligent reflecting surface selection for serving users in a distributed multi-base station, multi-intelligent reflecting surface, multi-user system. The prior art requires the base station side to establish and maintain a record table of the nearest and the second nearest base stations to each intelligent reflecting surface, and the table content includes the base station numbers of the two nearest and the second nearest base stations to each intelligent reflecting surface, and the corresponding base station-intelligent reflecting surface channel state information, so as to perform base station-intelligent reflecting surface matching and design the beamforming matrix at the base station end and the phase matrix of each intelligent reflecting surface. However, the method of the present invention does not require real-time estimation of the base station-intelligent reflecting surface channel state information for the design of the intelligent reflecting surface phase matrix, and only uses the 1-bit feedback information interacted between the user equipment and the base station to randomly adjust the phase matrix of the intelligent reflecting surface associated therewith, greatly saving the channel estimation overhead of the base station-intelligent reflecting surface channel and the intelligent reflecting surface-user equipment channel. In addition, the method of the present invention does not require obtaining the respective specific location information of the intelligent reflecting surface and the user equipment, and only needs to estimate the large-scale fading coefficients from each base station to each user, from each base station to each intelligent reflecting surface, and from each intelligent reflecting surface to each user equipment. Compared with the prior art, it avoids the path loss calculation error caused by only considering the position relationship. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] Figure 1 is a schematic flowchart of an existing method for base station and intelligent reflecting surface selection;

[0024] Figure 2 is a schematic architecture diagram of a multi-base station serving multi-user equipment system assisted by multiple intelligent reflecting surfaces;

[0025] Figure 3 is a schematic flowchart of the intelligent reflecting surface-user association process implemented by the method of the present invention;

[0026] Figure 4 is a flowchart of the intelligent reflecting surface phase matrix adjustment scheme based on 1-bit feedback implemented by the method of the present invention;

[0027] Figure 5 is a diagram of a multi-base station serving multi-user equipment assisted by multiple intelligent reflecting surfaces in the simulation experiment of the embodiment of the present invention;

[0028] Figure 6 is a graph showing the relationship between the ratio of the average signal power of the user to the average dry noise signal power and the number of IRS reflection units in the simulation experiment of the embodiment of the present invention;

[0029] Figure 7 is an experimental graph of the convergence of the 1-bit feedback algorithm in the present invention in the embodiment of the present invention. DETAILED DESCRIPTION OF THE EMBODIMENTS

[0030] The present invention will be further described in detail below in conjunction with the accompanying drawings and embodiments.

[0031] An intelligent reflecting surface selection and phase matrix adjustment method of the present invention. The overall architecture of the multi-base station serving multi-user equipment system assisted by multiple intelligent reflecting surfaces is as Figure 2 shown, including base stations, intelligent reflecting surfaces, and user equipment. Among them, the base station can be any type of base station, including but not limited to macro base stations, pico base stations, micro base stations, etc., or an access point with certain signal processing capabilities; the user equipment can be any device accessing the wireless network, including but not limited to mobile phones, computers, smart tablets, etc.

[0032] In the prior art, it is necessary to continuously maintain the base station-intelligent reflecting surface channel state information; the present invention considers the scenario where each base station serves only one user, and uses the ratio of the average signal power calculated by the large-scale fading parameters to the average dry-noise signal power as the judgment basis for the intelligent reflecting surface-user association scheme. As Figure 2 shown, the association relationship between each base station and each user equipment is determined by the strength of the reference signal power of each base station received by each user equipment. Each base station serves only one user equipment. Each base station only transmits valid information to the single user equipment it serves, but there is co-channel interference from other base stations in the system. Each pair of base station-user equipment is only subjected to passive beamforming by one intelligent reflecting surface. As Figure 3 shown, in the method of the present invention, the implementation steps of the scheme for selecting the intelligent reflecting surface associated with the user are as follows:

[0033] (1.1) Each user equipment receives the reference signals of each base station and establishes a connection with the base station according to the strength of the reference signals of each base station.

[0034] Each base station serves only one user equipment. Each user equipment preferentially selects the base station with the highest reference signal, and at the same time each base station preferentially selects the user with the highest reference signal power. For the user equipment, if there is already another better user equipment connected to the base station, the user equipment will continue to try to connect to the next base station in the order of the reference signal power strength until all user equipment are connected to the base station.

[0035] (1.2) Each base station obtains the downlink channel state information of the user equipment it serves, and obtains the large-scale fading coefficients of the channels from each base station to each user equipment, from each base station to each intelligent reflecting surface, and from each intelligent reflecting surface to each user equipment. Each user equipment obtains the noise variance when receiving its own signal.

[0036] The methods for obtaining the above uplink and downlink channel state information, large-scale fading coefficients of each channel, and noise variances of each user equipment include, but are not limited to, each base station sending reference signals for channel estimation or channel quality measurement to each user terminal in the downlink, directly obtaining the downlink channel state information according to channel reciprocity, or a specific hybrid method of the above two methods.

[0037] Let the large-scale fading coefficient of the direct channel from the nth base station to the kth user equipment be The large-scale fading coefficient of the channel from the nth base station to the jth intelligent reflecting surface is The large-scale fading coefficient of the channel from the jth intelligent reflecting surface to the kth user equipment is

[0038] (1.3) Assume that each base station performs maximum ratio transmission (MRT) beamforming only based on the downlink channel state information of its served users. For each user equipment, randomly select a possible intelligent reflecting surface-user equipment association method, and adjust the phases of the reflection units of each intelligent reflecting surface to align the phases of the signals after passing through the base station-intelligent reflecting surface-user equipment channel and the base station-user equipment channel. Calculate the average signal power and average interference power at each user equipment, as well as the ratio of the average signal power to the average signal-to-noise power. The average signal power and average interference power are calculated from the base station transmission power, large-scale fading coefficients of each channel, and noise variances of each user equipment.

[0039] The average signal power E{y k 2} is:

[0040]

[0041] where y k is the original signal received by the kth user equipment from the kth base station; P is the transmission power of the base station; L is the number of antennas at the base station; M is the number of reflection units of the intelligent reflecting surface; is the large-scale fading coefficient of the channel from the kth base station to the kth user equipment; a j,k indicates whether the jth intelligent reflecting surface serves the kth user equipment. When a j,k = 1, it means the jth intelligent reflecting surface serves the kth user equipment. When a j,k = 0, it means the jth intelligent reflecting surface does not serve the kth user equipment; q k,j,k represents the average large-scale channel gain from the kth base station to the kth user equipment via the jth intelligent reflecting surface, q k,j,k = β k,j η j,k ; Γ(·) represents the Gamma function; The set of numbers for intelligent reflecting surfaces.

[0042] When calculating the base station-intelligent reflecting surface-user equipment cascaded channel, it is assumed that the phase matrices of each intelligent reflecting surface have been adjusted so that the phase of the reflected signal is aligned with that of the direct signal transmitted by the base station. Therefore, the large-scale fading of this cascaded channel is the product of the large-scale fading of the base station-intelligent reflecting surface channel and the large-scale fading of the intelligent reflecting surface-user equipment channel.

[0043] The average interference power E{|I k | 2} for the k-th user equipment is:

[0044]

[0045] where I k is the interference received by the k-th user equipment from other base stations; represents the set of numbers for base station-user equipment pairs; is the large-scale fading coefficient of the channel from the n-th base station to the k-th user equipment; q n,j,k represents the large-scale channel gain from the n-th base station to the k-th user equipment via the j-th intelligent reflecting surface.

[0046] The ratio of the average signal power to the average interference plus noise power (Average Signal to Average Interference plus Noise Ratio, ASAINR) refers to the ratio of the average signal power of the user equipment to the sum of the average interference power and the noise power. The ratio γ of the average signal power to the average interference plus noise power for the k-th user equipment k is calculated as follows:

[0047]

[0048] where σ 2 is the average noise power at the k-th user equipment, i.e., the user equipment noise variance obtained in step 1.2.

[0049] Record the current intelligent reflecting surface-user equipment association method for each user, find the minimum value of the ratio of the average signal power to the average interference plus noise power of the user equipment and record it.

[0050] (1.4) Change the intelligent reflecting surface-user association method for each user equipment to obtain a new set of intelligent reflecting surface-user equipment association methods, and then continue to step 1.3 to recalculate the ratio of the average signal power to the average interference-plus-noise signal power of the user equipment until all possible intelligent reflecting surface-user equipment association methods are traversed. Compare the ratios of the average signal power to the average interference-plus-noise signal power of all user equipment with the smallest values among all association methods, and select the set of intelligent reflecting surface-user associations with the largest ratio as the final association method.

[0051] In the method of the present invention, after determining the final intelligent reflecting surface-user association method, the user equipment performs 1-bit feedback indication transmission according to the change in the received signal power intensity. The base station serving the user equipment makes a random phase change within a certain range for the phase shifts of the reflection units of the associated intelligent reflecting surface according to the received feedback indication until the stoppable condition is reached. As Figure 4 shown, the implementation steps of the intelligent reflecting surface phase matrix adjustment scheme based on 1-bit feedback in the method of the present invention are as follows.

[0052] (2.1) Determine the association method of the base station-intelligent reflecting surface-user equipment. Each base station performs maximum ratio transmission beamforming only according to the downlink channel state information of the users it serves.

[0053] (2.2) In the first iteration process, each base station randomly initializes the phases of the reflection units of the associated intelligent reflecting surface. Let θ t,k,i represent the phase shift of the i-th reflection unit of the k-th intelligent reflecting surface at the t-th iteration. When initializing, θ 1,k,i ∈[0,2π), i = 1, 2,..., M; Let θ 0,k,i be the phase shift value of the reflection unit corresponding to the current highest received signal power. Initially, θ 0,k,i = θ 1,k,i , i = 1, 2,..., M. In the embodiments of the present invention, each base station serves only one user equipment, and each pair of base station-user equipment is passively beamformed by only one intelligent reflecting surface. The numbers of the associated base stations, user equipment, and intelligent reflecting surfaces can be marked as the same, for example, all numbered k.

[0054] In the first iteration process, each base station can also calculate the phase difference between channels according to the known cascaded channel and direct channel, and set the initial phase matrix of the associated intelligent reflecting surface.

[0055] After the initialization is completed, each base station records the phase matrix of the associated intelligent reflecting surface. For the convenience of representation, let the number of reflection units of each intelligent reflecting surface be M. The protection scope of the present invention includes but is not limited to the case where the number of reflection units of each intelligent reflecting surface is the same.

[0056] The user equipment initializes the received signal power and sets it as the current highest received signal power. Let the received signal power at the k-th user equipment be r 1,k , and initially set the current highest received signal power r of the k-th user 0,k = r 1,k . The user equipment can obtain the received signal power through methods including but not limited to code division, etc.

[0057] (2.3) In the second and subsequent iteration processes, for all reflection units of each intelligent reflecting surface, the phase is randomly changed, that is, θ t,k,i = θ 0,k,i + δ t,k,i , i = 1, 2, …, M, where δ t,k,i is uniformly distributed in [-Δ, Δ], and is the phase shift change of the i-th reflection unit of the k-th intelligent reflecting surface at the t-th iteration. Δ is the maximum step size of each random phase shift change, and its value range is (0, π]. When the phase value range of the intelligent reflecting surface is discrete, let the discrete phase interval be δ, and the maximum step size of the random phase shift change of the 1-bit feedback algorithm is nδ, and the phase change range of δ t,k,i is discrete values from -nδ to nδ. n is a positive integer.

[0058] (2.4) In the second and subsequent iteration processes, after the random change of the phase of the reflection unit of the intelligent reflecting surface in step 2.3, each user equipment updates the received signal power. At the t-th iteration, the received signal power at the k-th user equipment is r t,k . If r t,k > r 0,k , the k-th user equipment feeds back an indication of "1" and sets r 0,k = r t,k ; if r t,k ≤ r 0,k , the k-th user equipment feeds back an indication of "0" and keeps r 0,k unchanged. For the k-th base station, if the feedback indication received from the user equipment is "1", then change the phase shift value of the reflection unit of the associated intelligent reflecting surface, that is, update the phase matrix of the associated intelligent reflecting surface, and let θ 0,k,i = θ t,k,i , i = 1, 2, …, M; if the feedback indication received is "0", then keep the phase shift value θ of the reflection unit of the associated intelligent reflecting surface 0,k,i , i = 1, 2, …, M unchanged, that is, restore the phase matrix of the associated intelligent reflecting surface to the state at the previous iteration. For the sake of easy representation, denote the feedback indication of the user equipment as "1" or "0", and the protection scope of the present invention includes but is not limited to any feedback indication that can reflect the situation of the user equipment. Introduce a binary variable b t,k, when the UE feeds back an indication of "1", b t,k = 1, otherwise b t,k = 0. For t ≥ 2, the phase of each reflecting element of the IRS can be expressed as

[0059]

[0060] At the t-th iteration, the received signal y at the k-th UE t,k is:

[0061]

[0062] where, x t,k represents the original signal sent from the k-th AP to the k-th UE at the t-th iteration, represents the equivalent direct channel from the k-th AP to the k-th UE after using MRT beamforming, a j,k represents whether the corresponding IRS serves the UE: when a j,k = 1, it means the j-th IRS serves the k-th UE; when a j,k = 0, it means the j-th IRS does not serve the k-th UE; h k,j represents the equivalent channel from the k-th AP to the j-th IRS after using MRT beamforming; g j,k = [g j,k,1 , g j,k,2 , …, g j,k,M represents the channel from the j-th IRS to the k-th UE; the reflection matrix of the j-th IRS can be expressed as a diagonal matrix Φ t,j represents the reflection matrix of the j-th IRS at the t-th iteration.

[0063] Furthermore, the SINR of the k-th UE at the t-th iteration and the corresponding achievable rate calculated according to the Shannon formula can be calculated.

[0064] (2.5) Repeat step 2.4, continuously update the phase of the reflecting elements of each intelligent reflecting surface and the maximum received signal power value of each user equipment until the preset number of iterations or the SINR threshold at the user equipment is reached. Finally, output the phase of the reflecting elements of each intelligent reflecting surface and the maximum received signal power of each user equipment.

[0065] The intelligent reflecting surface phase matrix adjustment scheme based on 1-bit feedback of the present invention does not design the intelligent reflecting surface phase matrix depending on the channel state information from the base station to the intelligent reflecting surface, iteratively updates the phase of the reflecting elements of each intelligent reflecting surface, determines whether to retain the randomly changed intelligent reflecting surface phase matrix each time through 1-bit feedback indication, and at the same time each user equipment records the current maximum received signal power, greatly saving the channel estimation overhead of the base station-intelligent reflecting surface channel and the intelligent reflecting surface-user equipment channel.

[0066] The simulation experiment of the method of the present invention is carried out, and some experimental results are as Figures 5 - 7 shown.

[0067] As Figure 5 shown, in this scenario, 4 APs are configured with single antennas, and the transmit power is 40 dBm. 4 IRSs serve 4 single-antenna UEs. The association connections of AP-IRS-UE with corresponding numbers are established by the method of the present invention. 2000 different channels are taken by the Monte Carlo method, and the number of reflection units of each IRS is gradually increased from 0 to 1000. It can be seen from Figure 6 that the ASA INR of each UE increases with the increase of the number M of IRS reflection units. For example, for UE1, its distance from the AP and IRS serving it is relatively close, and it is subject to strong interference from other UEs. When M increases, the beamforming gain it obtains can effectively compensate for the increased interference from other UEs. UE2 is the closest to the AP serving it. When M is small, the gain of active beamforming makes its ASA INR the highest. However, when M increases, affected by the large-scale fading coefficient, the passive beamforming gain it obtains is weaker than that of UE1, and the performance improvement is weaker than that of UE1. It can be seen from this that to achieve a higher passive beamforming gain of the IRS, the number of reflection units needs to be increased. At the same time, the distance between the UE and the AP and IRS serving it will also affect its performance. To enable UEs farther from the AP to have a better user experience, IRSs with more reflection units should be equipped for them or the number of IRSs should be increased.

[0068] As Figure 7As shown, when M of the IRS is 50, the convergence of the 1-bit feedback algorithm with step sizes of π / 30, π / 20, and π / 10 respectively, and the convergence rates of the D-MRT and E-MRT algorithms under the corresponding channel states are presented. The D-MRT algorithm refers to the MRT beamforming obtained based on the CSI of the direct link channel from the AP to the UE. The E-MRT algorithm refers to the joint beamforming that adjusts the phases of the IRS reflection elements to align with the phases of the signals on the direct link from the AP to the UE. The experimental scenario is a single-IRS-assisted single-AP serving single-UE system. The large-scale fading means of the channels from the AP to the UE, from the AP to the IRS, and from the IRS to the UE are all set to 0.1, and the small-scale fades all follow the Rayleigh distribution. It can be seen from the figure that D-MRT only utilizes the gain of the active beamforming of the AP, ignoring the passive beamforming effect of the IRS. The IRS randomly reflects the signals, and the corresponding received signal strength is the weakest overall. E-MRT achieves the alignment of the signals after the passive beamforming of the IRS with the active beamforming. Its corresponding received signal strength is the strongest. However, phase alignment means that the accurate CSI of the channels from the AP to the IRS and from the IRS to the UE needs to be obtained. As M increases, the corresponding channel estimation overhead also increases. For different step size values, after sufficient feedback, convergence can be achieved. The figure shows the results corresponding to integer multiples of 100 on the abscissa. The smaller the step size, the closer the converged value is to the effect of E-MRT. The 1-bit feedback algorithm adopted by the method of the present invention does not require CSI information related to the IRS channel and has the effect of saving channel estimation overhead when M increases.

[0069] The above experimental results prove that the method of the present invention solves the problem of providing intelligent reflector selection for user services in a distributed multi-base station, multi-intelligent reflector, multi-user setup system. By appropriately selecting the step size of phase adjustment, the convergence rate of the 1-bit feedback algorithm is improved. The converged 1-bit feedback algorithm can approach the performance when the phases of the active and passive beamformings are aligned, while reducing the requirement for channel estimation accuracy.

[0070] Except for the technical features described in the specification, they are all well-known technologies to those skilled in the art. The present invention omits the description of well-known components and well-known technologies to avoid redundancy and unnecessarily limit the present invention. The embodiments described in the above examples do not represent all embodiments consistent with the present application. Based on the technical solution of the present invention, various modifications or deformations that can be made by those skilled in the art without creative efforts are still within the protection scope of the present invention.

Claims

1. An intelligent reflecting surface selection and phase matrix adjustment method, which is used for the scenario of multiple base stations serving multiple user equipments assisted by multiple intelligent reflecting surfaces, is characterized in that The method includes the following steps: (1) Select an intelligent reflecting surface associated with the user, including: (1.1) Each user equipment receives the reference signals of each base station, and establishes a connection between the user equipment and the base station according to the base station reference signal strength; (1.2) Each base station obtains the channel state information of the downlink from the base station to the user equipment it serves, the large-scale fading coefficient of the channel from the base station to each user equipment, the large-scale fading coefficient of the channel from the base station to each intelligent reflecting surface, and the large-scale fading coefficient of the channel from each intelligent reflecting surface to each user equipment; each user equipment obtains the noise variance when receiving its own signal; (1.3) Assume that each base station performs maximum ratio transmission beamforming only according to the downlink channel state information of the user it serves. For each user equipment, randomly select an association method between the intelligent reflecting surface and the user equipment, and adjust the phases of the reflection units of each intelligent reflecting surface to align the phases of the signals after passing through the base station-intelligent reflecting surface-user equipment channel and the base station-user equipment channel, and calculate the average signal power and average interference power of each user equipment, as well as the ratio of the average signal power to the average signal-to-noise-and-interference power; where the ratio of the average signal power to the average signal-to-noise-and-interference power refers to the ratio of the average signal power of the user equipment to the sum of the average interference power and the noise power; Record the minimum value of the ratio of the average signal power to the average signal-to-noise-and-interference power of the user equipment under each association method between the intelligent reflecting surface and the user equipment; (1.4) Change the association method between the intelligent reflecting surface and the user of each user equipment to obtain a new set of association methods between the intelligent reflecting surface and the user equipment, and then go back to step 1.3 to recalculate the ratio of the average signal power to the average signal-to-noise-and-interference power of each user equipment; traverse all possible association methods between the intelligent reflecting surface and the user equipment, and select a set of association methods between the intelligent reflecting surface and the user with the largest minimum value of the ratio of the average signal power to the average signal-to-noise-and-interference power of the user equipment as the final output, and output the obtained set of association methods between the base station-intelligent reflecting surface-user equipment; (2) Adjust the phase matrix of the intelligent reflecting surface based on 1-bit feedback, including: (2.1) Determine the association method between each base station-intelligent reflecting surface-user equipment. Each base station randomly initializes the phases of the reflection units of the intelligent reflecting surface associated with the base station, and sets them to the phase shift values corresponding to the current highest received signal power; each user equipment initializes the received signal power and sets it to the current highest received signal power; (2.2) Iteratively update the phases of the reflection units of each intelligent reflecting surface; Randomly vary the phases of all reflection elements of each intelligent reflecting surface. The associated base station sends signals to the user equipment, and each user equipment updates the received signal power. If the current received signal power is greater than the current highest received signal power, the user equipment feeds back an indication of 1 and updates the current highest received signal power to the current received signal power; otherwise, the user equipment feeds back an indication of 0 and keeps the current highest received signal power unchanged. When the base station receives a feedback indication of 1 from the user equipment, it keeps the phases of the reflection elements of the current intelligent reflecting surface and updates the phase shift value of the reflection element corresponding to the current highest received signal power. When the base station receives a feedback indication of 0 from the user equipment, it restores the phases of the reflection elements of the intelligent reflecting surface to the state at the previous iteration while keeping the phase shift of the reflection element corresponding to the current highest received signal power unchanged. (2.3) Repeat step 2.2, continuously updating the phases of the reflection elements of each intelligent reflecting surface and the highest received signal power of each user equipment until a preset number of iterations or a preset user equipment signal-to-interference-plus-noise ratio threshold is reached, and output the final highest received signal power of each user equipment and the phase of the reflection element of each intelligent reflecting surface.

2. The method according to claim 1, wherein In step 1.3 described above, in each association mode of the intelligent reflecting surface - user equipment, the average signal power E{|y k | 2} of the k-th user equipment is calculated as follows: where y k is the original signal received by the k-th user equipment from the k-th base station; P is the transmission power of the base station; L is the number of antennas at the base station; M is the number of reflecting elements of the intelligent reflecting surface; is the large-scale fading coefficient of the channel from the k-th base station to the k-th user equipment; a j,k marks whether the j-th intelligent reflecting surface serves the k-th user equipment. When a j,k = 1, it means that the j-th intelligent reflecting surface serves the k-th user equipment. When a j,k = 0, it means that the j-th intelligent reflecting surface does not serve the k-th user equipment; q k,j,k represents the average large-scale channel gain from the k-th base station to the k-th user equipment via the j-th intelligent reflecting surface. q k,j,k = β k,j η j,k , is the large-scale fading coefficient of the channel from the k-th base station to the j-th intelligent reflecting surface, is the large-scale fading coefficient of the channel from the j-th intelligent reflecting surface to the k-th user equipment; Γ(·) represents the Gamma function; is the set of numbers of the intelligent reflecting surfaces; The average interference power E{|I k | 2} for the k-th user equipment is: Among them, I k is the interference received by the k-th user equipment from other base stations; represents the set of numbers of base station-user equipment pairs; is the large-scale fading coefficient of the channel from the n-th base station to the k-th user equipment; q n,j,k represents the average large-scale channel gain from the n-th base station to the k-th user equipment via the j-th intelligent reflecting surface; The ratio γ of the average signal power to the average dry-noise signal power of the k-th user equipment k is calculated as follows: where σ 2 is the average noise power of the k-th user equipment.

3. The method according to claim 1, characterized in that, The above step 2.2 includes: Let the numbers of the intelligent reflecting surface and the base station associated with the user equipment be both k, and let the current maximum received signal power of the k-th user equipment be r 0,k The phase shift value of the reflection unit of the intelligent reflecting surface corresponding to the current maximum received signal power is θ 0,k,i , i = 1, 2, …, M, where M is the number of reflection units of the intelligent reflecting surface; At the t-th iteration, randomly vary the phases of all reflection elements of the k-th intelligent reflecting surface, which is expressed as: θ t,k,i = θ 0,k,i + δ t,k,i , i = 1, 2, …, M; where, δ t,k,i is the phase shift change of the i-th reflection element of the k-th intelligent reflecting surface at the t-th iteration; δ t,k,i is uniformly distributed in [-Δ, Δ], where Δ is the maximum step size of each random phase shift change, and the value range is (0, π]; When the phase value range of the intelligent reflecting surface is discrete, assume the discrete phase interval is δ, the maximum step size of the random phase shift change is nδ, the phase change range is discrete values from -nδ to nδ, and n is a positive integer.

4. The method according to claim 1 or 3, characterized in that, The above step 2.2 includes: At the t-th iteration, after randomly varying the phases of all the reflecting elements of the intelligent reflecting surface, the received signal of the k-th user equipment is expressed as where x t,k represents the original signal sent by the k-th base station to the k-th user equipment at the t-th iteration; represents the equivalent direct channel from the k-th base station to the k-th user equipment after maximum ratio transmission beamforming; is the set of numbers of the intelligent reflecting surfaces; a j,k marks whether the j-th intelligent reflecting surface serves the k-th user equipment. When a j,k = 1, it means that the j-th intelligent reflecting surface serves the k-th user equipment. When a j,k = 0, it means that the j-th intelligent reflecting surface does not serve the k-th user equipment; g j,k represents the channel from the j-th intelligent reflecting surface to the k-th user equipment; Φ t,j represents the reflection matrix of the j-th intelligent reflecting surface at the t-th iteration; h k,j represents the equivalent channel from the k-th base station after maximum ratio transmission beamforming to the j-th intelligent reflecting surface; Furthermore, calculate the signal-to-interference-plus-noise ratio of the k-th user equipment at the t-th iteration.

5. The method according to claim 1, characterized in that, In the above step 2.1, when initializing the phase matrix of the intelligent reflecting surface, each base station calculates the phase difference between channels according to the determined cascaded channel of base station-intelligent reflecting surface-user equipment and the direct channel of base station-user equipment, and sets it as the initial phase matrix of the intelligent reflecting surface associated with the base station.

Images