Resource optimization system and method for distributed RIS-assisted millimeter wave communication system

Abstract

The invention discloses a resource optimization system and method for a distributed RIS-assisted millimeter wave communication system, and belongs to the technical field of wireless communication. The system comprises a base station which is provided with MBS antennas to form a uniform linear array and is connected with NRF radio frequency chains; each radio frequency chain is connected with MBS antennas through MBS phase shifters in a full connection mode; the plurality of distributed RISs are used for assisting the terminal equipment to carry out millimeter wave communication in downlink transmission; the plurality of terminal devices are divided into a plurality of NOMA groups, and each NOMA group is provided with a transmission service by a corresponding RIS; wherein joint optimization is carried out on pairing of the RIS and the equipment, RIS phase shift optimization, digital precoding design and base station sending power distribution, so that the total achievable throughput of a communication system in downlink transmission is maximized. The method can solve the problem that millimeter wave signals are easy to block, improves the utilization rate of spectrum resources, and can effectively improve the total throughput of a system.

Description

Technical Field

[0001] The present invention relates to the field of wireless communication technology, and in particular to a resource optimization system and method for a distributed RIS-assisted millimeter wave communication system. Background Art

[0002] With the continuous development of wireless communication technology, people's demand for wireless data services has exploded. Since the millimeter wave (mmWave) band in the frequency range of 30GHz to 300GHz has abundant available spectrum resources, mmWave communication has become one of the most promising technologies in future wireless communication networks. In wireless communication networks, the application of multiple access technology is a necessary way to achieve user reuse. Compared with the traditional orthogonal multiple access (OMA) technology, the application of multi-user non-orthogonal multiple access (NOMA) technology can further improve the system spectrum efficiency, so the combination of mmWave communication and NOMA technology is of great significance. Considering the problem that mmWave signals are easily blocked in actual application scenarios, RIS is introduced as one of the key technologies to solve this problem. Smart reflector (RIS) is a nearly passive device that can change the surrounding communication environment by flexibly adjusting the amplitude and phase of the incident signal. It has the advantages of low hardware cost, low energy consumption, and easy deployment. In actual application scenarios, the deployment of distributed RIS can expand the coverage of mmWave communication and improve the system communication performance.

[0003] In future wireless communication networks, the requirements for large-scale connections and high spectrum efficiency exceed the capabilities of traditional OMA technologies. Therefore, the combination of mmWave communication and NOMA technology has been widely studied. In order to solve the problem that mmWave signals are easily blocked by buildings, the academic community has also carried out extensive research on RIS-assisted millimeter wave NOMA communication systems. However, the current research still has the following problems: 1) At present, most of the research on mmWave systems focuses on the application research of NOMA technology in the mmWave frequency band, or applies RIS to millimeter wave systems. However, the combination of mmWave and NOMA technologies faces the challenges of mmWave signals being easily blocked and having a small coverage range, while RIS-assisted mmWave communication systems are subject to the limitations of traditional OMA technologies and have limited spectrum utilization efficiency. At present, there is a lack of research on combining NOMA technology, RIS and mmWave. 2) At present, most of the research on RIS-assisted mmWave systems focuses on single RIS-assisted mmWave systems. In actual application scenarios, due to the blockage of obstacles such as buildings, mmWave signals face the challenge of limited coverage. 3) At present, the performance optimization research of the mmWave-NOMA system assisted by distributed RIS mainly focuses on optimizing user grouping, RIS phase shift, beamforming and power allocation, and rarely optimizes the pairing between RIS and users. However, in the actual application scenarios of multiple RIS and multiple users, since the number of users that each RIS can serve is limited and the channel conditions of each user are different, it is necessary to match an optimal RIS for each user according to the actual situation. Summary of the invention

[0004] In order to solve at least one of the technical problems existing in the prior art to a certain extent, the object of the present invention is to provide a resource optimization system and method for a distributed RIS-assisted millimeter wave communication system.

[0005] The first technical solution adopted by the present invention is:

[0006] A resource optimization system for a distributed RIS-assisted millimeter wave communication system, comprising:

[0007] Base station, equipped with M BS antennas form a uniform linear array, and connect N RF RF chains; each RF chain is fully connected through M BS phase shifter and M BS Antenna connections, where the number of RF chains is much smaller than the number of antennas;

[0008] Multiple distributed RISs to assist terminal devices in millimeter wave communications in downlink transmission;

[0009] Multiple terminal devices are divided into several NOMA groups, and each NOMA group is provided with transmission services by a corresponding RIS;

[0010] Among them, the pairing of RIS and equipment, RIS phase shift optimization, digital precoding design and base station transmit power allocation are jointly optimized to maximize the total achievable throughput of the communication system in downlink transmission.

[0011] Furthermore, the joint optimization step includes two stages:

[0012] In the first stage, the method based on Gale-Shapley (GS) algorithm and worst connection swapping (WCS) algorithm is used to optimize the pairing of RIS and devices and obtain the best pairing result;

[0013] In the second stage, according to the obtained pairing results, the RIS phase shift, digital precoding and power allocation are jointly optimized using the alternating iterative method.

[0014] The second technical solution adopted by the present invention is:

[0015] A resource optimization method for a distributed RIS-assisted millimeter wave communication system comprises the following steps:

[0016] Design a system model for a distributed RIS-assisted millimeter wave communication system; the communication system includes a base station, multiple distributed RIS and multiple terminal devices;

[0017] Construct mmWave channel model and communication system downlink model based on the system model;

[0018] Determine the optimization problem based on the constructed model;

[0019] Based on the optimization problem, the pairing of RIS and devices, RIS phase shift optimization, digital precoding design and base station transmit power allocation are jointly optimized to maximize the total achievable throughput of the communication system in downlink transmission.

[0020] Furthermore, the system model includes K terminal devices and J RIS, with UE d Indicates the dth terminal device, using RIS j represents the jth RIS;

[0021] The RIS with multiple reflective elements per RIS is modeled as a uniform planar array (UPA), and the distance between adjacent reflective elements of the RIS is d x and d y Separated by horizontal and vertical spacing.

[0022] Furthermore, the mmWave channel model is constructed as follows:

[0023] The mmWave channel is characterized by a line-of-sight (LoS) link and L non-line-of-sight (NLoS) links. j The channels between b,j Indicates that, and RIS j and UE d The channel between j,d Represented, and all mmWave channels are modeled using the Saleh-Valenzuela channel model as follows:

[0024]

[0025]

[0026] In the formula, and represent the complex gain of the LoS channel, and and They represent the complex gains of the NLoS channels respectively; and Represents BS and RIS respectively j Between and RIS j and UE d The LoS component of the channel between and They represent the NLoS components of the corresponding channels respectively.

[0027] Furthermore, the communication system downlink model is constructed as follows:

[0028] Get the effective channel matrix from the base station to all terminal devices: From base station to user UE d The effective channel vector is expressed as in Indicates base station to RIS j Then to the user UE d The cascaded channel is represented as Θ j Represents RIS j The phase shift matrix, w j Indicates that the base station sends to RIS j The simulated precoding vector is expressed as:

[0029]

[0030] Based on the obtained effective channel, consider the pairing problem between RIS and terminal equipment and introduce μ j,dTo indicate the pairing relationship between RIS and terminal devices, when RIS j and UE d When pairing is complete, μ j,d =1; otherwise, μ j,d =0;

[0031] Set the base station to transmit to RIS j The signal representation of the NOMA group of a service is: p d For UE d Distributed power, s d Sent by the base station to the user UE sdd The target signal, then RIS j UE of service d The received signal is represented as:

[0032]

[0033] The first part of the formula represents UE d The received target expected signal, the second part indicates the UE d The third part represents the UE receiving the interference signal from the terminal device in the same NOMA group. d Interference received from other terminal devices served by other RIS; From base station to user UE d The effective channel vector, D is the digital coding matrix, expressed as D = [t1, t2, ..., t J ], where t j Represents RIS j The number and coding vector corresponding to the service NOMA group, x is the transmission signal vector, expressed as x = [x1, x2, ..., x J ], J is the number of RIS, K is the number of terminal devices, H is the conjugate transpose, n k is the additive white Gaussian noise received by the kth terminal device;

[0034] For the terminal devices in the area, the NOMA mode is proposed, that is, each RIS paired terminal device forms a NOMA group, and the serial interference cancellation (SIC) technology is used to eliminate the interference within the group;

[0035] According to the serial interference elimination principle, RIS j UE of service d The downlink data rate is:

[0036]

[0037] in, Indicates UE d The inter-group interference, μr,n For RIS r and user UE n The pairing relationship coefficient, if μ r,n =1, representing RIS r and user UE n Complete the pairing, otherwise, μ r,n =0. n Assigned to the user UE by the base station n The power, Indicates UE d The total throughput of the terminal devices in the area is expressed as

[0038] Furthermore, determining the optimization problem according to the constructed model includes:

[0039] By jointly optimizing the pairing relationship between RIS and terminal devices μ j,d , the phase shift matrix Θ of all RIS, the digital precoding matrix D and the power allocation vector P to maximize the total achievable throughput; the mathematical expression of the optimization problem is (P1):

[0040]

[0041] In the optimization problem (P1), C1 is the pairing variable μ between RIS and equipment. j,d C2 means that each RIS can only serve two terminal devices at most; C3 means that each terminal device can only be served by one RIS; C4 is the unit modulus constraint of the RIS phase shift; C5 is the non-negative constraint on the power allocated to each terminal device; C6 represents the maximum transmission power constraint of the base station; C7 is the minimum data rate constraint for each terminal device; and C8 is the SIC decoding rate constraint.

[0042] Furthermore, the joint optimization of pairing of RIS and equipment, RIS phase shift optimization, digital precoding design and base station transmission power allocation includes:

[0043] In the first stage, the method based on Gale-Shapley (GS) algorithm and worst connection swapping (WCS) algorithm is used to optimize the pairing of RIS and devices and obtain the best pairing result;

[0044] In the second stage, according to the obtained pairing results, the RIS phase shift, digital precoding and power allocation are jointly optimized using the alternating iterative method.

[0045] Furthermore, the first stage specifically includes:

[0046] Given the RIS phase shift matrix Θ j=I N , the digital precoding matrix is ​​D = I J , and the average power distribution p d =P BS / K premise, P BS For the maximum transmission power of the base station, the optimization problem of pairing RIS and terminal devices is transformed into (P2.1):

[0047]

[0048] Based on the cascade channel information of the terminal device, the initial pairing result of the RIS and the terminal device is obtained through the Gale-Shapley algorithm;

[0049] According to the initial pairing results obtained, the worst connection swapping algorithm is used to perform refined swapping and pairing to obtain the final approximate optimal solution.

[0050] Furthermore, the second stage specifically includes:

[0051] According to the pairing results of RIS and equipment obtained in phase 1, the optimization of the three parameters of RIS phase shift, digital precoding and power allocation is divided into three sub-phases for optimization;

[0052] In the first sub-stage, given the digital precoding matrix and any feasible power allocation, the RIS phase shift matrix is ​​optimized by using the convex difference algorithm and the semi-positive definite relaxation algorithm.

[0053] In the second sub-stage, after obtaining the optimized RIS phase shift matrix, the digital precoding matrix is ​​designed using the MMSE-based digital precoding design;

[0054] In the third sub-stage, based on the obtained RIS phase shift matrix and digital precoding matrix, the convex difference algorithm is used to solve the optimization problem of power allocation.

[0055] The beneficial effects of the present invention are as follows: the present invention considers a system framework combining mmWave, RIS and NOMA technologies. Introducing RIS in the NOMA system of the mmWave frequency band can solve the problem that the millimeter wave signal is easily blocked; introducing NOMA technology in the RIS-assisted mmWave system can effectively improve the utilization of spectrum resources. In addition, the present invention proposes a two-stage multi-dimensional resource joint optimization method for RIS and user pairing, RIS phase shift, digital precoding and power allocation to improve the total throughput of the system. BRIEF DESCRIPTION OF THE DRAWINGS

[0056] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the embodiments of the present invention or the drawings of related technical solutions in the prior art are introduced below. It should be understood that the drawings introduced below are only for the convenience of clearly describing some embodiments of the technical solutions of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0057] Figure 1 Schematic diagram of the structure (system transmission model) of the distributed RIS-assisted mmWave band NOMA system in an embodiment of the present invention;

[0058] Figure 2 A schematic diagram showing the influence of the maximum transmission power of the base station, the RIS and the number of users on the total throughput of the system when the pairing scheme proposed by the present invention, the initialization pairing scheme based on the GS algorithm and the random pairing scheme are respectively adopted in the distributed RIS-assisted mmWave band NOMA system considered in the embodiment of the present invention;

[0059] Figure 3 A schematic diagram showing the influence of the number of RIS reflection units and the minimum horizontal distance from the base station to the user on the total throughput of the system in a distributed RIS-assisted mmWave band NOMA system considered in an embodiment of the present invention, when the scheme proposed in the present invention, the initialization pairing scheme based on the GS algorithm, and the random pairing scheme are respectively adopted;

[0060] Figure 4 A schematic diagram showing the influence of the maximum transmission power of the base station, the RIS and the number of users on the total throughput of the system when the power allocation scheme and the water injection algorithm and the average power allocation proposed in the present invention are respectively adopted in the distributed RIS-assisted mmWave band NOMA system considered in the embodiment of the present invention;

[0061] Figure 5 In the distributed RIS-assisted mmWave band NOMA system considered in an embodiment of the present invention, a schematic diagram showing the influence of the total throughput of the system on the minimum horizontal distance from the base station to the user, the RIS and the number of users when the power allocation scheme and water injection algorithm proposed in the present invention and the average power allocation are respectively adopted. DETAILED DESCRIPTION

[0062] The embodiments of the present invention are described in detail below, and examples of the embodiments are shown in the accompanying drawings, wherein the same or similar reference numerals throughout represent the same or similar elements or elements having the same or similar functions. The embodiments described below with reference to the accompanying drawings are exemplary and are only used to explain the present invention, and are not to be construed as limitations of the present invention. For the step numbers in the following embodiments, they are only provided for the convenience of explanation, and the order between the steps is not limited in any way, and the execution order of each step in the embodiment can be adaptively adjusted according to the understanding of those skilled in the art.

[0063] In the description of the present invention, it should be understood that descriptions involving orientations, such as up, down, front, back, left, right, etc., and orientations or positional relationships indicated are based on the orientations or positional relationships shown in the accompanying drawings, and are only for the convenience of describing the present invention and simplifying the description, and do not indicate or imply that the device or element referred to must have a specific orientation, be constructed and operated in a specific orientation, and therefore should not be understood as a limitation on the present invention.

[0064] In the description of the present invention, "several" means one or more, "more" means more than two, "greater than", "less than", "exceed" etc. are understood as not including the number itself, and "above", "below", "within" etc. are understood as including the number itself. If there is a description of "first" or "second", it is only used for the purpose of distinguishing the technical features, and cannot be understood as indicating or implying the relative importance or implicitly indicating the number of the indicated technical features or implicitly indicating the order of the indicated technical features.

[0065] In the description of the present invention, unless otherwise clearly defined, terms such as setting, installing, connecting, etc. should be understood in a broad sense, and technicians in the relevant technical field can reasonably determine the specific meanings of the above terms in the present invention based on the specific content of the technical solution.

[0066] In order to overcome the problems of limited propagation range and shortage of spectrum resources faced by the NOMA system based on the mmWave frequency band and the RIS-assisted mmWave system respectively, and considering the shortcomings of single RIS in actual application scenarios, a distributed RIS-assisted NOMA system in the mmWave frequency band is designed, in which the distributed RIS assists multiple users in mmWave communication in downlink transmission, and the users are divided into several NOMA groups. Each NOMA group is provided with transmission services by a corresponding RIS, and the target information of each user is decoded by the NOMA technology.

[0067] In order to improve the total throughput of the system, the present invention proposes a two-stage multi-dimensional resource joint optimization method for RIS and user pairing, RIS phase shift, digital precoding and power allocation. Specifically, the joint optimization method can be divided into two stages. In the first stage, the pairing of RIS and the device is divided into two steps, namely initialization pairing and refinement pairing. Among them, considering the channel information of the cascade channel between RIS and the user, a method based on the GS algorithm is used for initialization pairing. In order to obtain a more effective pairing result, the idea of ​​exchange pairing is adopted, and a method based on the WCS algorithm is proposed for refined pairing. In the second stage, the RIS phase shift, digital precoding and power allocation are jointly optimized using an alternating iteration method. Among them, under the premise of given digital precoding matrix and power allocation, the RIS phase shift is optimized using the convex difference algorithm and the semi-positive relaxation method. After obtaining the optimized RIS phase shift matrix, a digital precoding matrix design based on MMSE is adopted to eliminate inter-group interference. Based on the obtained RIS phase shift matrix and digital precoding matrix, considering the decoding rate constraints between users in the same NOMA group, the convex difference algorithm is used to solve the optimization problem of power allocation. The simulation results show that, compared with other optimization schemes, the multi-dimensional resource joint optimization method proposed in the present invention can effectively improve the total system throughput.

[0068] The technical solution of the present invention is explained in detail below in conjunction with the accompanying drawings and specific embodiments.

[0069] (1) System model

[0070] In the system model considered in this embodiment, K users are randomly and evenly distributed in a circle with a radius of 30 meters and a base station as the center. J RIS are randomly and evenly distributed in a circle with a radius of 10 meters and a base station as the center. d and RIS j denote the dth user and the jth RIS respectively, and BS denotes the base station. Each RIS has N = N x N y The RIS with 1 reflective element can be modeled as a uniform planar array (UPA), and the distance between adjacent reflective elements of the RIS is d x and d y Separated for horizontal and vertical spacing. j The phase shift matrix is ​​set to The base station is equipped with M BS The antennas form a uniform linear array (ULA) and connect N RF Each RF chain is fully connected through M BS phase shifter and M BS The number of RF chains is much smaller than the number of antennas, that is, M BS>>N RF In order to be closer to the actual application scenario, this embodiment considers K>J, that is, the number of users is greater than the number of RISs.

[0071] a) mmWave channel modeling

[0072] The mmWave channel can be characterized by one line-of-sight (LoS) link and L non-line-of-sight (NLoS) links. j The channels between b,j Indicates that, and RIS j and UE d The channel between j,d Represented, and all mmWave channels are modeled using the Saleh-Valenzuela channel model as follows:

[0073]

[0074]

[0075] in, and represent the complex gain of the LoS channel, and and represent the complex gains of the NLoS channel respectively. and Represents BS and RIS respectively j Between and RIS j and UE d The LoS component of the channel between them. and Respectively represent the NLoS components of the corresponding channels.

[0076]

[0077]

[0078] Assuming that the base station antennas are deployed at half-wavelength intervals, the BS and RIS j The ULA array response of the lth outgoing path can be expressed as:

[0079]

[0080] in, is the exit elevation angle of the lth path. Let λ represent the signal wavelength, and introduce and BS and RIS j The UPA array response of the lth arrival path between j and UE dThe UPA array response of the lth outgoing path between can be expressed as:

[0081]

[0082]

[0083] in, and BS and RIS respectively j The arrival azimuth and elevation of the lth path between and RIS j and UE d The exit azimuth and elevation of the lth path between them.

[0084] b) NOMA system downlink in mmWave band assisted by distributed RIS

[0085] The maximum transmission power of the base station is P BS , user UE d The allocated power is denoted as p d Consider base station to RIS j The simulated precoding vector is:

[0086]

[0087] in, Then the analog beamforming matrix sent by the base station is In addition, we consider using digital precoding technology to eliminate inter-group interference and introduce a digital precoding matrix D = [t1, t2, …, t J Considering BS-RIS j -UE d The cascaded channel can be expressed as Then BS to UE d The effective channel vector can be obtained by concatenating the channel and simulating the coding vector Therefore, the effective channel matrix from the base station to all users is

[0088] According to the obtained effective channel, consider the pairing problem between RIS and users and introduce μ j,d To indicate the pairing relationship between RIS and the device, when RIS j and UE d When pairing is complete, μ j,d =1; otherwise, μ j,d = 0. In addition, we assume that each RIS can serve at most two users, that is And each user can only be served by one RIS, that is Without loss of generality, we consider J<K≤2J.

[0089] Set the base station to transmit to RIS j The signal of the NOMA group of the service is represented as represents the transmission signal vector. Then the user RIS j UE of service d The received signal can be expressed as:

[0090]

[0091] The first part of the above formula represents UE d The received target expected signal, the second part indicates the UE d The third part represents the UE receiving the interference signal from the users in the same NOMA group. d The fourth part represents the interference received from other users of other RIS services. d The received additive white Gaussian noise (AWGN) is σ 2 Represents the noise power.

[0092] For users in the area, the NOMA mode is proposed, that is, each RIS paired user forms a NOMA group, and the intra-group interference is eliminated through the serial interference cancellation (SIC) technology. Without loss of generality, we arrange the users according to the size of the effective channel gain, that is, According to the SIC principle, RIS j UE of service d The downlink data rate is:

[0093]

[0094] in, Indicates UE d The inter-group interference Indicates UE d The total throughput of users in the area can be expressed as

[0095] When performing SIC decoding, consider the UE d and UE k In the same NOMA group and with RIS j In the case of matching, assuming d < k, then according to the SIC rule, UE d Decode the UE before decoding its own signal k The corresponding data rate of the signal is:

[0096]

[0097] in, and Respectively represent the inter-group interference and intra-group interference received during the decoding process. In order to ensure the smooth implementation of SIC, the decoding rate needs to meet

[0098] (2) Maximizing the total system throughput

[0099] To complete the above definition, the present invention aims to optimize the pairing relationship μ between RIS and equipment by joint optimization. j,d 、The phase shift matrix of all RIS Θ={Θ1,…,Θ J}、Digital precoding matrix D=[t1,t2,…,t J ] and the power allocation vector To maximize the total achievable throughput. The mathematical expression of the optimization problem is (P1):

[0100]

[0101] In the optimization problem (P1), C1 is the pairing variable μ between RIS and equipment. j,d C2 means that each RIS can only serve two users at most. C3 means that each user can only be served by one RIS. C4 is the unit modulus constraint of the RIS phase shift. C5 is the non-negative constraint on the power allocated to each user. C6 represents the maximum transmission power constraint of the base station. C7 is the minimum data rate constraint for each user. C8 is the SIC decoding rate constraint.

[0102] (3) Joint resource optimization allocation method

[0103] Due to the above constraint C1, μ j,d is a binary variable, and constraints C2 and C3 are non-convex constraints. In addition, the J RIS phase shift matrices are highly correlated, and the optimization parameter μ j,d, Θ, D and P are highly coupled to each other. In summary, the optimization problem (P1) is a highly complex non-convex optimization problem. In order to effectively optimize these parameters, an embodiment of the present invention proposes a two-stage resource optimization design method. Specifically, in stage one, a method based on the Gale-Shapley (GS) algorithm and the worst connection swapping (WCS) algorithm is proposed to optimize the pairing of RIS and equipment, and obtain approximately optimal results. In the second stage, for the remaining parameters Θ, D and P, the alternating iteration (AO) algorithm is adopted for iterative optimization. Among them, we approximately transform the joint optimization problem of the phase shift matrices of J RIS into J independent RIS phase shift optimization problems. In each iteration process, given the digital precoding matrix D and any feasible power allocation P, the convex difference algorithm and the semi-positive definite relaxation method are applied to solve the RIS phase shift optimization problem. After obtaining the optimized RIS phase shift matrix, we adopt a digital precoding design based on the minimum mean square error (MMSE) criterion. Based on the RIS phase shift matrix and the digital precoding matrix obtained after the above optimization, the convex difference algorithm is used to solve the optimization problem of power allocation. The following is a detailed description of the two stages:

[0104] a) Phase 1: Optimizing the pairing of RIS and devices

[0105] In practical communication networks, the number of RIS is often smaller than the number of users, and each RIS can only serve a limited number of users. Therefore, it is necessary to pair RIS with users to find the best pairing solution to improve the communication performance of the system. Represent the set of all RIS and all users respectively. Given the RIS phase shift matrix is The digital precoding matrix is ​​D = I J , and the average power distribution Under the premise of , the optimization problem of RIS and user pairing is transformed into (P2.1):

[0106]

[0107] Since constraint C1 is a binary variable constraint, and constraints C2 and C3 are both non-convex constraints, the above optimization problem is a non-convex integer programming problem. It is very difficult to find the global optimal solution to this problem. Although the exhaustive method can solve this problem, its computational complexity is very high, especially when the number of RIS and the number of users are large. In order to obtain an effective but low-complexity method, the present invention proposes a pairing algorithm that combines the GS algorithm and the WCS algorithm to solve the pairing problem of RIS and devices. Specifically, the grouping algorithm can be divided into two steps. In the first step, based on the user's cascade channel information, the GS algorithm is used to obtain the initial pairing result of RIS and the user. In the second step, based on the pairing result obtained in the first step, the WCS algorithm is used to perform refined exchange pairing to obtain the final approximate optimal solution.

[0108] Step 1: Initialize pairing. Initialize pairing mainly consists of two steps: selecting the head user and pairing the head user with the remaining users. When selecting the head user, we use the channel gain of the cascade channel between each RIS and each user as the preference value between them, for example, RIS j and UE d The preference relationship value between is defined as:

[0109]

[0110] in, Represents RIS j Select UE d The preference value of Indicates UE d Choose RIS j In order to better characterize the mutual preference relationship between each RIS and each user, define Q j and Q d For RIS j Preference lists for all users and UE d A list of preferences for all RIS. And Q j By adding RIS j Arrange the preference values ​​of all users from large to small. Similarly, Q d You can use UE d The preference values ​​of all RIS are arranged from large to small. After obtaining the preference lists of all RIS and user pairs, the set of all preference lists in the pairing problem is defined as Based on the preference relationship obtained above, the head user selection problem in the pairing problem can be solved by the proposed method based on the GS algorithm. The Proposal Rule and Rejection Rule of the algorithm are defined as follows:

[0111] Proposal Rule: Any RIS j It makes an invitation to the user who ranks highest in its preference list. If the invitation is rejected, the user who rejected it is removed from the preference list.

[0112] Rejection Rule: For any user UE that receives an invitation d If the RIS that makes the invitation is not the highest ranking RIS in the user's preference list, the invitation is rejected.

[0113] According to the rules defined above, through a series of iterative processes, the GS algorithm is used to select the head user of each RIS pairing to complete the first sub-step. For the pairing of the remaining users with RIS, it can be transformed into the grouping problem of the remaining users and the head users matched with each RIS, and this pairing is also completed based on the GS algorithm. The difference is that when defining the preference values ​​of RIS and the remaining users, the channel correlation between the head users of each RIS and the remaining users is used as an indicator. For the convenience of description, it is better to define any RIS j The head user is UE sj , then the preference value can be defined as:

[0114]

[0115] Similarly, define Q′ j Represents RIS j Preference lists for all remaining users, Q′ d Represents the remaining user UE d A list of preferences for all RIS. Similarly, Q′ j By adding RIS j Arrange the preference values ​​of all remaining users from large to small to obtain Q′ d The remaining user UE d The preference values ​​of all RIS are arranged from large to small. After obtaining the preference list between all RIS and the remaining users, the GS algorithm is used to complete the pairing of RIS and the remaining users based on the Proposal Rule and Rejection Rule defined above. After the selection of the head user and the pairing of the remaining users, the initial pairing result between RIS and the user can be obtained. And the pairing method adopted can make the users served by the same RIS have a higher channel correlation. However, since the total system throughput is also affected by factors such as interference, the initial pairing result still needs refined pairing to further improve the optimization performance.

[0116] Step 2: Refined pairing. Since the initial pairing result obtained in the first step only considers the cascade channel gain and channel correlation between the RIS and the user, and the total system throughput is also affected by factors such as interference, further refined pairing is required to obtain a near-optimal pairing result.

[0117] In order to better describe the iterative idea of ​​the proposed pairing algorithm based on the WCS idea, we define As a pairing strategy between RIS and devices, Representation and RIS j The set of users that complete the pairing is given with the following definition.

[0118] Definition 1: In each iteration, we define the user with the lowest data rate among the available users as the worst user and select it, that is:

[0119]

[0120] in Indicates that under the pairing strategy Π, and fixed D=I J and When UE d data rate.

[0121] Definition 2: For the worst user selected We consider the following process to j Matched and RIS r Matching UE k To exchange, that is:

[0122]

[0123] In each iteration loop, first find the worst user in that iteration based on Definition 1 Then discuss each situation.

[0124] Case 1: If the worst user is the only user in the NOMA group, then It is necessary to exchange with all other K-1 users and obtain K-1 new pairing strategies, that is, Ξ={Π1,Π2,…,Π B}, where B = K-1.

[0125] Case 2: If the worst user is not the only user in the NOMA group. It is necessary to exchange with all K-2 users matched with other RISs and obtain K-2 new pairing strategies, that is, Ξ={Π1,Π2,…,Π B}, where B = K-2.

[0126] Among all the new optional pairing strategies obtained in the i-th iteration, find the pairing strategy that makes the total system throughput the highest as the candidate pairing strategy for the i+1-th iteration, that is,

[0127]

[0128] in In the pairing strategy Π (i) After selecting the candidate pairing strategy, we need to further decide whether to adopt the candidate strategy, so we give the following definition.

[0129] Definition 3: A candidate strategy can be adopted if and only if the data rate of each user does not decrease and the data rate of at least one user increases, that is:

[0130]

[0131]

[0132] If the candidate strategy If the conditions of Definition 3 are met, it will be adopted and enter the next iteration. If the condition is not met, then the worst user in this iteration will no longer be selected as the worst user in subsequent iterations, i.e.

[0133] Based on the above definition, the second step of refined exchange pairing will be carried out in an iterative way. When no user can be selected as the worst user, the iteration stops, thus obtaining a refined pairing result.

[0134] b) Phase 2: Jointly optimizing RIS phase shift, digital precoding and power allocation using the AO algorithm

[0135] According to the pairing results of RIS and equipment obtained in stage one, the optimization of other parameters can be divided into three sub-stages. In the first sub-stage, given the digital precoding matrix and any feasible power allocation, the RIS phase shift matrix is ​​optimized by the convex difference algorithm and the semi-positive definite relaxation algorithm. In the second sub-stage, after obtaining the optimized RIS phase shift matrix, we use the MMSE-based digital precoding design to design the digital precoding matrix. In the third sub-stage, based on the RIS phase shift matrix and digital precoding matrix obtained above, the optimization problem of power allocation can be solved by applying the convex difference algorithm.

[0136] The first sub-stage: Optimization of the RIS phase shift matrix. The optimization problem of this stage is expressed as (P3.1):

[0137]

[0138] Since the phase shift matrices of the J RIS are highly correlated, the complexity of the joint optimization problem (P3.1) is extremely high. In order to obtain a low-complexity solution, we decompose the joint phase shift optimization problem of J RIS into J independent sub-problems. j The typical sub-problem can be expressed as (P3.2):

[0139]

[0140] Where, given D = I J Sometimes, there is Introduction in Representing RIS j Phase shift matrix Θ j To further transform the above problem into a solvable form, we have:

[0141]

[0142] in Let f j,d =Ψ j,d w j ,have Then, the objective function of the optimization problem (P3.2) is continuously approximated and the logarithmic expression in the objective function is converted into the form of logarithmic difference, that is:

[0143]

[0144] in, And there are:

[0145]

[0146] make and Where V j ≥0, rank(V j )=1. Then we have and And the optimization problem can be equivalent to (P3.3):

[0147]

[0148] in,

[0149]

[0150] The above problem (P3.3) is still a non-convex problem, because the objective function is the difference between two convex functions, and the constraint C11 is a non-convex rank 1 constraint. In order to transform the problem (P3.3) into a form that can be solved by the convex difference algorithm, we first use the first-order linear approximation to find The upper bound of is:

[0151]

[0152] in, is V in the tth iteration j The value of yes The first-order derivative at the tth iteration is:

[0153]

[0154] Then, by using the semi-definite relaxation method to relax the rank 1 constraint of C11, problem (P3.3) can be transformed into (P3.4):

[0155]

[0156] Obviously, the optimization problem (P3.4) is now in a standard semidefinite programming form and can be solved by CVX. In addition, in order to ensure that the optimization problem can still obtain a rank 1 result after relaxing the rank 1 constraint, we perform an additional Gaussian randomization process to transform the high-rank solution into a rank 1 solution. By solving the above J independent RIS phase shift optimization problems, we can obtain the set of phase shift matrices of all RIS

[0157] The second sub-stage: Design of digital precoding matrix. After obtaining the optimized RIS phase shift matrix, we use MMSE-based digital precoding design to eliminate inter-group interference. Since each NOMA group may have more than one user, we perform singular value decomposition (SVD) on the effective channel matrix of each NOMA group. For example, if the user UE d and user UE kBelong to the same NOMA group and have the same j Complete pairing, represents the effective channel matrix of the NOMA group. Then, the effective channel matrix is ​​subjected to SVD to obtain:

[0158]

[0159] Among them, U j =[u 1,j ,u 2,j ]and Represent the left singular matrix and the right singular matrix respectively. j represents the singular value matrix, and its diagonal elements are arranged from large to small. Then, for the above user UE d and user UE k The NOMA group composed of , its equivalent channel vector can be expressed as For only one user UE k When forming a NOMA group, the equivalent channel vector of the group can be directly expressed as Based on the above, the equivalent channel matrix of all NOMA groups can be expressed as

[0160] After obtaining the equivalent channel matrix, the MMSE-based digital precoding matrix can be expressed as:

[0161]

[0162] The third sub-stage: Optimize power allocation. After obtaining the optimized RIS phase shift matrix and digital precoding matrix, the power allocation optimization problem can be expressed as (P4.1):

[0163]

[0164] By applying continuous convex approximation, the objective function of problem (P4.1) can be transformed into the form of logarithmic difference, as follows:

[0165]

[0166] Then, we define and as follows:

[0167]

[0168] Therefore, problem (P4.1) can be rewritten in the standard convex programming form of (P4.2):

[0169]

[0170] in,

[0171]

[0172]

[0173]

[0174] At iteration t, we define The gradient of and get The lower bound of is:

[0175]

[0176] in, and

[0177]

[0178] Substitution Into the objective function of problem (P4.2), problem (P4.2) can be transformed into (P4.3):

[0179]

[0180] Problem (P4.3) is a standard convex programming problem and can be solved using CVX.

[0181] Iterative optimization: Finally, the RIS phase shift matrix, digital precoding matrix, and power allocation are iterated alternately until convergence.

[0182] (4) Simulation Experiment Results

[0183] In this embodiment, the effectiveness of the proposed two-stage resource optimization method in improving the achievable throughput of the system is verified through simulation analysis, and a comparative analysis is performed in two dimensions. One dimension is to optimize the remaining parameters using the optimization method proposed in stage two, and compare the effects of different RIS and user pairing algorithms on the total system throughput. The comparison algorithms are the initialization pairing algorithm based on the GS algorithm and the random pairing algorithm of the pairing algorithm proposed in the present invention. Another dimension is to compare the effects of different power allocation algorithms on the total system throughput while keeping the optimization methods of other parameters unchanged. The comparison algorithms are the classic water injection algorithm and average power allocation.

[0184] The relevant parameters of the simulation experiment are set as follows: The mmWave simulation channel is set with one LoS path and ten NLoS paths. The path loss of mmWave at a reference distance of 1m is set to -30dB. The minimum horizontal distance from the base station to K users is set to d min= 30m, and J RIS are randomly and evenly deployed in a circle with a radius of 10m and the base station as the center. The maximum transmit power is set to 30dBm. The noise power is σ 2 =-80dBm. The number of base station antennas is M BS =64, the number of each RIS reflection unit is N=16.

[0185] Figure 2 It is shown that in the NOMA system under the mmWave frequency band under the consideration of distributed RIS assistance, when the two-stage resource optimization method proposed in this invention is adopted, the total system throughput is affected by the maximum transmission power P of the base station. BS The changes in RIS and the number of users are analyzed, and the optimization algorithm in stage one is compared with the initialization pairing algorithm based on the GS algorithm and the random allocation algorithm. It can be seen from the simulation diagram that as the maximum transmission power of the base station increases, the total system throughput of the three schemes increases. Compared with the scheme that only implements the initialization pairing, the proposed scheme can make the total system throughput higher. This is because the proposed scheme includes further refined pairing and can obtain better pairing results. In addition, the total system throughput obtained by the scheme that only performs the initialization pairing based on the GS algorithm is still higher than that of the scheme that adopts random allocation, which further illustrates that optimizing the pairing of RIS and devices can greatly improve the system transmission rate. With the increase of RIS and the number of users, it can be observed that the proposed algorithm is still better than the other two comparison schemes, which shows that the pairing algorithm we proposed can also obtain effective pairing results in scenarios with a large number of users.

[0186] Figure 3 It is shown that in the NOMA system under the mmWave frequency band under the consideration of distributed RIS assistance, when the two-stage resource optimization method proposed in this invention is adopted, the total system throughput is affected by the number of RIS reflection units N and the minimum distance d between the base station and the user. min The settings of the comparison scheme are the same as the above figure. It can be seen from the figure that with the increase in the number of RIS reflection units, the total system throughput of the three schemes increases accordingly, because the increase in the number of RIS reflection units brings a greater UPA array response gain. Among them, it can be observed that the resource optimization scheme proposed in the present invention can still obtain a higher total system throughput than the other two comparison schemes. When the minimum distance from the base station to the user increases, the proposed scheme can still obtain better system performance, which shows that the proposed scheme is still applicable to communication scenarios with a larger coverage area.

[0187] Figure 4 It is shown that in the NOMA system under the mmWave frequency band under the consideration of distributed RIS assistance, when the two-stage resource optimization method proposed in this invention is adopted, the total system throughput is affected by the maximum transmission power P of the base station.BS The changes in the power allocation algorithm and the impact of the number of RIS and users are analyzed, and the power allocation algorithm in stage two is compared with the water injection algorithm and the random allocation algorithm. It can be observed from the simulation diagram that as the maximum transmission power of the base station increases, the total system throughput of the three schemes increases accordingly, and the resource optimization scheme proposed in the present invention can still obtain a higher total system throughput. This shows that the power allocation algorithm proposed by us can obtain a more effective power allocation result, thereby obtaining better system performance. With the increase in the number of RIS and the number of users, the proposed scheme is still better than the other two comparison schemes, which further illustrates that effective resource optimization is crucial to improving the total system throughput.

[0188] Figure 5 It is shown that in the NOMA system under the mmWave frequency band under the consideration of distributed RIS assistance, when the two-stage resource optimization method proposed in this invention is adopted, the total system throughput is affected by the minimum horizontal distance d from the base station to the user. min As well as the impact of RIS and the number of users. The settings of the comparison scheme are the same as the above figure. It can be observed from the figure that as the minimum horizontal distance from the base station to the user increases, the total system throughput obtained by the three schemes decreases accordingly, and the proposed scheme is still better than the other two schemes. Moreover, when the number of RIS and the number of users increase, the resource optimization scheme proposed by the present invention can still obtain a higher total system throughput, which shows the effectiveness of the proposed two-stage resource optimization method in improving the total system throughput.

[0189] (5) Advantages and beneficial effects

[0190] At present, most of the research on RIS-assisted NOMA systems in the mmWave frequency band focuses on the application scenarios of a single RIS. In the research on system performance optimization, existing research mainly focuses on optimizing user grouping, RIS phase shift, beamforming and power allocation, and rarely optimizes the pairing between RIS and users. Considering that the existing technology is not in-depth enough in the research on RIS and user pairing and the joint optimization of multiple parameters, the present invention considers a distributed RIS-assisted mmWave frequency band NOMA system, and performs two-stage optimization on the optimization problems involved in RIS and user pairing, RIS phase shift optimization, digital precoding and power allocation. The present invention mainly has the following advantages:

[0191] 1) The present invention considers a system framework that combines mmWave, RIS and NOMA technologies. Introducing RIS in the NOMA system of the mmWave frequency band can solve the problem that the millimeter wave signal is easily blocked; introducing NOMA technology in the mmWave system assisted by RIS can effectively improve the utilization rate of spectrum resources.

[0192] 2) At the same time, in actual application scenarios, since the number of users that each RIS can serve is limited, the introduction of multiple RIS can effectively increase the coverage of mmWave signals, provide high-quality services to more users and bring higher total system throughput.

[0193] 3) Based on the proposed system, in order to maximize the total throughput of the system, the present invention proposes a multi-dimensional resource joint optimization method for RIS and user pairing, RIS phase shift optimization, digital precoding design and power allocation. The simulation results show that the proposed scheme can effectively improve the total throughput of the system.

[0194] In the above description of this specification, the description with reference to the terms "one embodiment / example", "another embodiment / example" or "certain embodiments / examples" etc. means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present invention. In this specification, the schematic representation of the above terms does not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials or characteristics described may be combined in any one or more embodiments or examples in a suitable manner.

[0195] Although embodiments of the present invention have been shown and described, it will be appreciated by those skilled in the art that various changes, modifications, substitutions and variations may be made to the embodiments without departing from the principles and spirit of the present invention, and that the scope of the present invention is defined by the claims and their equivalents.

[0196] The above is a specific description of the preferred implementation of the present invention, but the present invention is not limited to the above embodiments. Those skilled in the art may make various equivalent modifications or substitutions without violating the spirit of the present invention. These equivalent modifications or substitutions are all included in the scope defined by the claims of this application.

Claims

1. A distributed RIS-assisted millimeter wave communication system resource optimization system, characterized in that: include: Base station, equipped with M BS antennas form a uniform linear array, and connect N RF RF chains; each RF chain is fully connected through M BS phase shifter and M BS Antenna connections, where the number of RF chains is much smaller than the number of antennas; Multiple distributed RISs to assist terminal devices in millimeter wave communications in downlink transmission; Multiple terminal devices are divided into several NOMA groups, and each NOMA group is provided with transmission services by a corresponding RIS; Among them, the pairing of RIS and equipment, RIS phase shift optimization, digital precoding design and base station transmit power allocation are jointly optimized to maximize the total achievable throughput of the communication system in downlink transmission.

2. The resource optimization system of a distributed RIS-assisted millimeter wave communication system according to claim 1, characterized in that: The joint optimization process consists of two stages: In the first stage, the Gale-Shapley algorithm and the worst connection swapping algorithm are used to optimize the pairing of RIS and devices to obtain the best pairing result. In the second stage, according to the obtained pairing results, the RIS phase shift, digital precoding and power allocation are jointly optimized using the alternating iterative method.

3. A resource optimization method for a distributed RIS-assisted millimeter wave communication system, characterized in that: The following steps are involved: Design a system model for distributed RIS-assisted mmWave communication systems; The communication system includes a base station, multiple distributed RIS and multiple terminal devices; Construct mmWave channel model and communication system downlink model based on the system model; Determine the optimization problem based on the constructed model; Based on the optimization problem, the pairing of RIS and devices, RIS phase shift optimization, digital precoding design and base station transmit power allocation are jointly optimized to maximize the total achievable throughput of the communication system in downlink transmission.

4. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 3, characterized in that: The system model includes K terminal devices and J RIS. d Indicates the dth terminal device, using RIS j represents the jth RIS; Each RIS with multiple reflective elements is modeled as a uniform planar array, and the distance between adjacent reflective elements of the RIS is d x and d y Separated by horizontal and vertical spacing.

5. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 3, characterized in that: The mmWave channel model is constructed as follows: The mmWave channel is characterized by a line-of-sight link and L non-line-of-sight links, the base station BS and RIS j The channels between b,j Indicates that, and RIS j and UE d The channel between j,d Represented, and all mmWave channels are modeled using the Saleh-Valenzuela channel model as follows: In the formula, and represent the complex gain of the LoS channel, and and They represent the complex gains of the NLoS channels respectively; and Represents BS and RIS respectively j Between and RIS j and UE d The LoS component of the channel between and They represent the NLoS components of the corresponding channels respectively.

6. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 3, characterized in that: The communication system downlink model is constructed as follows: Get the effective channel matrix from the base station to all terminal devices From base station to UE d The effective channel vector is expressed as in Indicates base station to RIS j Then to UE d The cascaded channel is represented as Θ j Represents RIS j The phase shift matrix, w j Indicates that the base station sends to RIS j The simulated precoding vector is expressed as: Based on the obtained effective channel, consider the pairing problem between RIS and terminal equipment and introduce μ j,d To indicate the pairing relationship between RIS and terminal devices, when RIS j and UE d When pairing is complete, μ j,d =1; otherwise, μ j,d =0; Set the base station to transmit to RIS j The signal representation of the NOMA group of a service is: p d Indicates that the base station is UE d Distributed power, s d Sent by the base station to the UE d The target signal, then RIS j UE of service d The received signal is represented as: The first part of the formula represents UE d The received target expected signal, the second part indicates the UE d The third part represents the UE receiving the interference signal from the terminal device in the same NOMA group. d Interference received from other terminal devices served by other RIS; D is a digital coding matrix, expressed as D = [t1, t2, ..., t J ], where t j Represents RIS j The number and coding vector corresponding to the service NOMA group, x is the transmission signal vector, expressed as x = [x1, x2, ..., x J ], J is the number of RIS, K is the number of terminal devices, H is the conjugate transpose, n k is the additive white Gaussian noise received by the kth terminal device; For the terminal devices in the area, the NOMA mode is adopted, that is, each RIS paired terminal device forms a NOMA group, and the serial interference cancellation technology is used to eliminate the interference within the group; According to the serial interference elimination principle, RIS j UE of service d The downlink data rate is: in, Indicates UE d The inter-group interference, μ r,n For RIS r and user UE n The pairing relationship coefficient, if μ r,n =1, representing RIS r and user UE n Complete the pairing, otherwise, μ r,n =0; p n Assigned to the user UE by the base station n The power, Indicates UE d The total throughput of the terminal devices in the area is expressed as 7. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 3, characterized in that: Determining the optimization problem according to the constructed model includes: By jointly optimizing the pairing relationship between RIS and terminal devices μ j,d , the phase shift matrix Θ of all RIS, the digital precoding matrix D and the power allocation vector P to maximize the total achievable throughput; the mathematical expression of the optimization problem is (P1): In the optimization problem (P1), C1 is the pairing variable μ between RIS and equipment. j,d C2 means that each RIS can only serve two terminal devices at most; C3 means that each terminal device can only be served by one RIS; C4 is the unit modulus constraint of the RIS phase shift; C5 is the non-negative constraint on the power allocated to each terminal device; C6 represents the maximum transmission power constraint of the base station; C7 is the minimum data rate constraint for each terminal device; and C8 is the SIC decoding rate constraint.

8. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 7, characterized in that: The joint optimization of pairing of RIS and equipment, RIS phase shift optimization, digital precoding design and base station transmission power allocation includes: In the first stage, the Gale-Shapley algorithm and the worst connection swapping algorithm are used to optimize the pairing of RIS and devices to obtain the best pairing result. In the second stage, according to the obtained pairing results, the RIS phase shift, digital precoding and power allocation are jointly optimized using the alternating iterative method.

9. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 8, characterized in that: The first stage specifically includes: Given the RIS phase shift matrix Θ j =I N , the digital precoding matrix is ​​D = I J , and the average power distribution p d =P BS / K premise, P BS For the maximum transmission power of the base station, the optimization problem of pairing RIS and terminal devices is transformed into (P2.1): stC1,C2,C3. Based on the cascade channel information of the terminal device, the initial pairing result of the RIS and the terminal device is obtained through the Gale-Shapley algorithm; According to the initial pairing results obtained, the worst connection swapping algorithm is used to perform refined swapping and pairing to obtain the final approximate optimal solution.

10. The resource optimization method of a distributed RIS-assisted millimeter wave communication system according to claim 8, characterized in that: The second stage specifically includes: According to the pairing results of RIS and equipment obtained in phase 1, the optimization of the three parameters of RIS phase shift, digital precoding and power allocation is divided into three sub-phases for optimization; In the first sub-stage, given the digital precoding matrix and any feasible power allocation, the RIS phase shift matrix is ​​optimized by using the convex difference algorithm and the semi-positive definite relaxation algorithm. In the second sub-stage, after obtaining the optimized RIS phase shift matrix, the digital precoding matrix is ​​designed using the MMSE-based digital precoding design; In the third sub-stage, based on the obtained RIS phase shift matrix and digital precoding matrix, the convex difference algorithm is used to solve the optimization problem of power allocation.

Images