A communication system based on smart surfaces

Abstract

The application discloses a communication system based on an intelligent reflecting surface, and the system realizes wireless communication based on a D2D communication system, and the system comprises a UAV base station, cellular users, an intelligent reflecting surface comprising a plurality of reflecting elements, and a plurality of D2D user pairs; wherein a D2D user power allocation is expressed as a non-convex optimization problem; and the non-convex optimization problem is solved based on an SD3 algorithm. The application can learn the properties of the problem itself to improve the calculation efficiency, learn the common features of the problem in a training mode to directly deploy and test the trained model, and then with the aid of the parallel computing capability of a neural network, the model can give an optimization scheme in a very short time, and solve the large-dimension optimization problem of the D2D system assisted by the intelligent reflecting surface with a relatively low complexity.

Description

Technical Field

[0001] This invention belongs to the field of wireless communication technology, and in particular relates to a communication system based on an intelligent reflective surface. Background Technology

[0002] Device-to-Device (D2D) communication allows two nearby users to transmit signals directly without going through a base station (BS), and is considered a potential solution for improving the utilization of limited spectrum. In traditional cellular communication, long transmission distances require significant transmission power, leading to greater path loss. D2D communication drastically reduces these distances, allowing users to achieve the required transmission rates with lower transmit power. D2D communication offers even greater advantages when users are located at the cell edge or when channel attenuation is severe in the environment.

[0003] In recent years, Reconfigurable Intelligent Surface (RIS) technology has become an effective method for improving channel quality. Especially with the help of a RIS controller, by rationally designing the RIS phase matrix, the desired signal and the reflected signal can be constructively combined to increase the power gain of the desired signal, or destructively reduce malicious interference, thereby establishing a communication channel that is conducive to improving system performance.

[0004] Currently, for RIS-assisted D2D communication systems, the main approaches combine alternating maximization and majorization-minimization algorithms, as well as semidefinite relaxation alternating optimization techniques. These traditional algorithms require global information, leading to significant signaling overhead. Furthermore, since traditional optimization methods only consider resource variables within a specific time interval, formulate optimization schemes based on resource allocation problems, and solve them using algorithms, they may only be applicable to small-scale networks and static wireless environments. Summary of the Invention

[0005] The purpose of this invention is to overcome the shortcomings of existing technologies and provide a communication system based on an intelligent reflective surface. This system can learn the properties of the problem itself to improve computational efficiency. It learns the common features of the problem through training and then directly deploys and tests the trained model. With the help of the parallel computing capabilities of neural networks, the model can provide an optimization solution in a very short time, thus solving the large-dimensional optimization problem of the D2D system assisted by the intelligent reflective surface with low complexity.

[0006] The objective of this invention is achieved through the following technical solution:

[0007] A communication system based on a smart reflective surface, wherein the system realizes wireless communication based on a D2D communication system, characterized in that:

[0008] The system includes a drone base station, cellular users, a smart reflective surface containing several reflective elements, and several D2D user pairs;

[0009] The D2D user power allocation problem is described as a non-convex optimization problem:

[0010]

[0011] in, , This represents the ratio of signal received by the j-th D2D receiver to interference plus noise. Let j be the transmit power of the j-th D2D transmitter. Let be the noise variance of the j-th D2D receiver. The transmission power from the drone base station to the cellular user. This represents the emission coefficient matrix of the smart reflector. , and Let be the phase shift and amplitude reflection coefficient of the m-th reflecting element, respectively. This represents the channel gain from the drone base station to the smart reflector.

[0012] The non-convex optimization problem is solved using the SD3 algorithm.

[0013] Furthermore, the SD3 algorithm takes current channel information as input, and the environmental status includes the current UAV location, and the signal-to-interference-plus-noise ratio for cellular users and D2D users;

[0014] The intelligent agent executes actions, including the phase matrix of the intelligent reflector, the power of the D2D user, and the transmission power of the UAV base station. After the intelligent agent completes the action, it enters the next state.

[0015] Using D2D users and rates as rewards, and given the current channel information and actions, the rewards are calculated based on the actor network.

[0016] Furthermore, the drone base station to the user i Channel gain is , among which, among which It is the path loss coefficient between the user and the drone link. It is the correlation coefficient with the NLOS link. A and B are constants related to the environment. , d represents the distance from the drone base station to the user.

[0017] Furthermore, the channel gain from the drone base station to the cellular user is:

[0018] ;

[0019] in, This indicates the distance from the drone base station to the cellular user. This represents the Rayleigh distribution coefficient generated randomly.

[0020] Furthermore, the channel gain from the UAV base station to the smart reflector is:

[0021] ;

[0022] in, This indicates the distance from the drone base station to the smart reflector. This represents the Rayleigh distribution coefficient generated randomly.

[0023] Furthermore, the drone base station to the first j The channel gain for each D2D user is:

[0024] ;

[0025] in, Indicates the drone base station to the j The distance between D2D users This represents the Rayleigh distribution coefficient generated randomly.

[0026] Furthermore, the gain from the smart reflector to the cellular user channel is:

[0027] ;

[0028] in, Indicates the distance from the smart reflector to the cellular user. Indicates path loss. d It is the distance between nodes. This represents the path loss at a distance of 1m. This represents the path loss coefficient at a distance of 1m.

[0029] Furthermore, the intelligent reflective surface to the first j The channel gain for each D2D user is:

[0030] ;

[0031] in, Indicates the intelligent reflective surface to the first j The distance between D2D users This represents the Rayleigh distribution coefficient generated randomly.

[0032] Furthermore, the cellular user to the j The channel gain for each D2D user is:

[0033] ;

[0034] in, Indicates cellular users up to the j The distance between D2D users This represents the Rayleigh distribution coefficient generated randomly.

[0035] Furthermore, the first j The D2D transmitter to the first The channel gain of each D2D receiver is:

[0036] ;

[0037] in, Indicates the first j The D2D transmitter to the first The distance between two D2D receivers This represents the Rayleigh distribution coefficient generated randomly.

[0038] The beneficial effects of this invention are as follows:

[0039] (1) This invention proposes a RIS-assisted D2D wireless communication system model. Due to the flexibility and portability of unmanned aerial vehicles, the use of UAVs as aerial base stations can improve the limitations of traditional base stations. The use of intelligent reflective surfaces can reduce interference caused by LOS links, improve channel quality, and solve the problems of signal attenuation and low data transmission rate caused by NLOS links.

[0040] (2) This invention applies the SD3 algorithm to the proposed optimization problem. The SD3 algorithm uses a soft maxima operator based on double estimation to solve the problem of large underestimation bias, which can smooth the optimization environment and thus help with empirical learning. The proposed RIS-SD3 algorithm has low implementation complexity and can solve the large-dimensional optimization problem involved in RIS-assisted D2D communication systems.

[0041] (3) Unlike previous studies that used alternating optimization techniques to optimize power allocation and phase, the RIS-SD3 algorithm proposed in this invention simultaneously obtains the power of D2D user pairs and the phase of the smart reflector. Specifically, the algorithm is trained using the sum rate as an immediate reward, and the algorithm parameters are iteratively adjusted by observing the reward. Attached Figure Description

[0042] Figure 1 This is a model diagram of a communication system based on a smart reflective surface according to an embodiment of the present invention;

[0043] Figure 2 This is a schematic diagram of the SD3 algorithm used in the embodiments of the present invention;

[0044] Figure 3 This is a comparison chart showing the impact of the learning rate on the convergence of the SD3 algorithm in embodiments of the present invention. Detailed Implementation

[0045] The following specific examples illustrate the implementation of the present invention. Those skilled in the art can easily understand other advantages and effects of the present invention from the content disclosed in this specification. The present invention can also be implemented or applied through other different specific embodiments, and various details in this specification can also be modified or changed based on different viewpoints and applications without departing from the spirit of the present invention. It should be noted that, unless otherwise specified, the following embodiments and features described therein can be combined with each other.

[0046] Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0047] Currently, for RIS-assisted D2D communication systems, the main approaches combine alternating maximization and majorization-minimization algorithms, as well as semidefinite relaxation alternating optimization techniques. These traditional algorithms require global information, leading to significant signaling overhead. Furthermore, since traditional optimization methods only consider resource variables within a specific time interval, formulate optimization schemes based on resource allocation problems, and solve them using algorithms, they may only be applicable to small-scale networks and static wireless environments.

[0048] To address the aforementioned technical problems, the present invention provides the following embodiments of a communication system based on an intelligent reflective surface.

[0049] Reference Figure 1 ,like Figure 1 The diagram shown is a model of a communication system based on a smart reflector in this embodiment. The system includes a drone base station, one CUE (cell user), a RIS (smart reflector) containing four reflective elements, and two D2D user pairs.

[0050] The drone base station, CUE, and D2D are all equipped with a single antenna. The drone base station is located at... m, CUE position is m, RIS position is m, D2D users are randomly distributed within a circle with a radius of 10m, and the location of the first D2D transmitter (DT) is denoted as m. The distance between the DT and its corresponding receiver DR is 5m. The maximum transmit power of the base station is 30W, the maximum transmit power of the DT is 10W, and the noise variance is... Path loss index between user and drone link Correlation coefficient with NLOS link Path loss coefficient at 1m in other links The path loss index under this link .

[0051] The processing steps at the drone base station transmitter are as follows:

[0052] The drone base station generates 128 bits of information The generated information is modulated, and the modulated signal is And satisfy The generated signal s Send into the channel.

[0053] The signals transmitted by the drone base station will form LOS and NLOS links. The drone-to-user channel construction process is as follows:

[0054] LOS and NLOS in the channel gain occur with a certain probability. Assume the ground node coordinates are... The coordinates of the UAV base station are The distance between the drone and the node is Then the LOS component probability of the node is:

[0055]

[0056] in The NLOS component probabilities of a node are: .

[0057] Path loss coefficient between user and drone link Correlation coefficient with NLOS link Then the drone will reach the user. i The channel gain is:

[0058]

[0059] The distance between the drone and the CUE user is ,but , , ,

[0060] Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0061] The distance between the drone and the RIS is If we ignore the differences in RIS position, then , , Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0062] Drones and the j The distance between D2D users is ,but , , Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0063] Other link channel construction processes are as follows:

[0064] Path loss coefficient at 1m The path loss index under this link For other link channel gains, considering a random Rayleigh distribution and path loss, the path loss is:

[0065]

[0066] The distance between RIS and CUE users is ,but Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0067] RIS and the j The distance between D2D users is ,but Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0068] CUE and the j The distance between D2D users is ,but Randomly generate Rayleigh distribution coefficients Then the channel gain is: .

[0069] No. j The first DT (D2D sender) and the second The distance between each DR (D2D receiver) user is ,but Randomly generate Rayleigh distribution coefficients Then the channel gain is: , of whichj The channel gain between DT and DR is denoted as , No. j The first DT and the first The channel gain between the DRs is denoted as .

[0070] The processing steps at the receiving end are as follows:

[0071] The SINR received by the CUE is:

[0072]

[0073] No. j The SINR received by each DR is:

[0074]

[0075] No. j The achievable rate of each DR is:

[0076]

[0077] The sum rate of all DRs is:

[0078]

[0079] As one implementation method, this embodiment aims to improve the achievable rate of the DUE while limiting the amount of interference to the CUE. An optimization algorithm is needed to perform optimal phase design for the RIS and power allocation for D2D users. Since strong transmission always exists between D2D links, it is necessary to guarantee the QoS of the CUE and ensure that the CUE's SINR is greater than the minimum SINR threshold. Therefore, resource allocation is formulated as a nonconvex optimization problem as follows:

[0080]

[0081] in .

[0082] Reference Figure 2 ,like Figure 2The diagram shows the principle of the SD3 algorithm used in this embodiment. The entire training process consists of 10,000 episodes, each iterating 60 steps, with each step representing the UAV's position. The algorithm inputs current channel information, and the environmental state includes the current UAV position, SINR of CUE users, and SINR of D2D users. The agent's current action includes the phase matrix of the RIS, the power of D2D users, and the base station transmit power. After the agent completes its action, it enters the next state. The optimization objective, namely D2D users and rate, is used as the reward. Given the current channel information and actions, the reward can be calculated based on the actor network. Furthermore, to meet the minimum SNR and maximum interference requirements of CUE users and the QoS requirements of D2D users, the reward can be set as follows:

[0083]

[0084] The reward for each episode is: .

[0085] Reference Figure 3 ,like Figure 3 The image shown is a comparison of the impact of the learning rate on the convergence of the SD3 algorithm in this embodiment. From... Figure 3 The impact of different learning rates on the convergence of the proposed algorithm can be observed. A large learning rate leads to poor stability, affecting later performance and even causing the model to fail to converge. A small learning rate results in slow convergence or even failure to learn, wasting training time. As shown in the figure, the RIS-SD3 algorithm used in this embodiment exhibits the best convergence when the actor and critic learning rates are 1e-6.

[0086] This embodiment proposes a RIS-assisted D2D wireless communication system model. Due to the flexibility and portability of unmanned aerial vehicles, using drones as aerial base stations can improve the limitations of traditional base stations. The use of intelligent reflective surfaces can reduce interference caused by LOS links, improve channel quality, and solve the problems of signal attenuation and low data transmission rate caused by NLOS links.

[0087] This embodiment applies the SD3 algorithm to the proposed optimization problem. The SD3 algorithm uses a soft maxima operator based on double estimation to address the problem of large underestimation bias, which can smooth the optimization environment and thus facilitate empirical learning. The proposed RIS-SD3 algorithm has low implementation complexity and can solve large-dimensional optimization problems involved in RIS-assisted D2D communication systems.

[0088] Unlike previous studies that used alternating optimization techniques to optimize power allocation and phase, this embodiment proposes a RIS-SD3 algorithm that simultaneously obtains the power of D2D user pairs and the phase of the smart reflector. Specifically, the algorithm is trained using the sum rate as an immediate reward, and the parameters are iteratively adjusted by observing the reward.

[0089] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A smart reflector-based communication system, the system implements wireless communication based on D2D communication system, characterized in that: The system includes a UAV base station, a cellular user, a smart reflector containing several reflecting elements and several D2D user pairs; Wherein, the D2D user power allocation is expressed as a non-convex optimization problem: 0 < P c ≤ P max , SINR c ≥SINR thr , where P = {P1, P2,..., P D}, denotes the signal-to-interference-plus-noise ratio of the jth D2D receiver, P j′ is the transmit power of the jth D2D transmitter, σ j 2 is the noise variance of the jth D2D receiver, P c is the transmit power of the UAV base station to cellular users, denotes the transmit coefficient matrix of the intelligent reflecting surface, θ m ∈ [0, 2π) and β m are the phase shift and amplitude reflection coefficient of the mth reflecting element, respectively, h r denotes the channel gain from the UAV base station to the intelligent reflecting surface; The non-convex optimization problem is solved based on the SD3 algorithm.

2. The intelligent surface-based communication system of claim 1, wherein, The SD3 algorithm inputs the current channel information, the environmental state includes the current UAV position, the signal-to-interference-plus-noise ratio of the cellular user and the D2D user; The agent performs an action, the action includes the phase matrix of the smart reflector, the power of the D2D user and the transmission power of the UAV base station, after the agent performs the action, it enters the next state; The D2D user and the rate are taken as the reward, given the current channel information and the action, the reward is calculated according to the actor network.

3. The intelligent surface-based communication system of claim 1, wherein, The drone base station-to-user i channel gain is wherein, wherein a0is a path loss coefficient between a user and a drone link, η is a correlation coefficient with an NLOS link, A, B are environment-related constants, P NLoS = 1 - P LoS , d is the distance from the drone base station to the user.

4. The intelligent surface-based communication system of claim 3, wherein, The channel gain from the UAV base station to the cellular user is: wherein, denotes the distance of the drone base station to the cellular user, denotes a randomly generated Rayleigh distribution coefficient.

5. The intelligent surface-based communication system of claim 3, wherein, The channel gain from the UAV base station to the smart reflector is: wherein, represents the distance from the drone base station to the smart reflector, represents a randomly generated Rayleigh distribution coefficient.

6. The intelligent surface-based communication system of claim 3, wherein, The channel gain from the UAV base station to the jth D2D user is: wherein, denotes the distance from the drone base station to the jth D2D user, denotes a randomly generated Rayleigh distribution coefficient.

7. The intelligent surface-based communication system of claim 3, wherein, The channel gain from the smart reflector to the cellular user is: wherein, denotes the distance of the intelligent reflecting surface to the cellular user, denotes the path loss, d is the distance between nodes, p denotes the path loss at 1 m, and v denotes the path loss coefficient at 1 m.

8. The intelligent surface-based communication system of claim 7, wherein, The channel gain from the smart reflector to the jth D2D user is: wherein, denotes the distance of the intelligent reflecting surface to the jth D2D user, denotes a randomly generated Rayleigh distribution coefficient.

9. The intelligent surface-based communication system of claim 7, wherein, The channel gain from the cellular user to the jth D2D user is: wherein, denotes the distance from the cellular user to the jth D2D user, denotes a randomly generated Rayleigh distribution coefficient.

10. The intelligent surface-based communication system of claim 7, wherein, The channel gain from the jth D2D transmitter to the j'th D2D receiver is: wherein, represents a distance from the jth D2D transmitter to the j'th D2D receiver, represents a randomly generated Rayleigh distribution coefficient.

Images