Unmanned aerial vehicle carries RIS assisted vehicle network communication method and system

Abstract

The application discloses a kind of unmanned vehicle network communication method and system of carrying RIS assisted, first unmanned vehicle according to prediction model prediction vehicle next time slot position;Then, the phase shift matrix of RIS reflection unit is optimized, while optimizing the trajectory of unmanned vehicle, so that the communication rate of two vehicles in next time slot maximization;Finally, according to the optimization result, the phase shift factor of RIS reflection unit is adjusted, and the position reached by unmanned vehicle in next time slot flies to.This application solves the problem that vehicles cannot directly communicate due to obstacles, under the constraints of vehicle communication delay and unmanned aerial vehicle flight energy consumption, by optimizing the RIS phase shift matrix to enhance the signal quality reflected to the vehicle, by optimizing the trajectory of unmanned vehicle to better serve the vehicle communication, while using DDPG algorithm to solve the optimization problem, to obtain the strategy of maximizing the vehicle communication rate, meet the requirements of vehicle network communication, improve the performance of vehicle communication.

Description

Technical Field

[0001] This invention belongs to the field of vehicle network communication technology, specifically relating to a method and system for UAV-equipped RIS-assisted vehicle network communication. Background Technology

[0002] The main purpose of research on vehicle communication is to make people's daily travel safer and more convenient, thus paving the way for intelligent transportation systems and autonomous driving applications. However, due to the complex propagation environment, especially considering the obstruction of multiple buildings in cities and the rapid changes in channels caused by the high mobility of vehicles, the propagation link of vehicle communication is easily degraded.

[0003] Due to their high mobility, low cost, and line-of-sight transmission capabilities, drones have been used as mobile airborne base stations or airborne relays to improve ground communication performance by deploying them in optimal flight positions. However, traditional communication assistance technologies typically employ auxiliary devices such as relays, which consume a lot of energy and are susceptible to environmental interference. Existing research on drone-assisted vehicle-to-everything (V2X) networks focuses on statically deploying drones to maximize the coverage area for communication between drones and ground users, without considering the mobility of drones. Furthermore, the limited energy carried by drones necessitates energy management. Most existing research combines UAVs and RIS (Radio Reliability System), but only addresses static scenarios involving ground users or targets. In V2X, however, vehicles possess high mobility, introducing significant unknowns to the communication environment. The RSU (Radio Service Unit) needs to continuously exchange various status information with the vehicle and make online decisions based on this real-time information. Solving for optimal decisions using traditional optimization or heuristic algorithms requires substantial computational resources and is difficult to deploy online in real-time. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a method and system for UAV-equipped RIS-assisted vehicle network communication. By optimizing the RIS phase shift matrix and UAV trajectory while satisfying V2V communication delay constraints and UAV energy consumption constraints, the communication rate between vehicles is maximized, thereby improving the communication performance of the vehicle network.

[0005] The present invention adopts the following technical solution:

[0006] A method for RIS-assisted vehicle network communication using a drone, targeting two vehicles in a target area with communication barriers, utilizes a drone equipped with a preset number of RIS reflective units as a communication relay to perform the following steps to maximize the communication rate between the two vehicles in each time slot:

[0007] Step A: Based on the information of the specified vehicle status types of the first vehicle and the second vehicle in the current time slot, predict the vehicle positions of the first vehicle and the second vehicle in the next time slot respectively.

[0008] Step B: Based on the predicted vehicle positions of the first and second vehicles in the next time slot, obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory;

[0009] Step C: Based on the optimized RIS reflector phase shift matrix and the UAV trajectory, adjust the RIS reflector phase shift matrix and the position the UAV will reach in the next time slot.

[0010] Preferably, in step A, based on the information of the specified vehicle state types of the first vehicle and the second vehicle in the current time slot, the vehicle positions of the first vehicle and the second vehicle in the next time slot are predicted using the following formula, that is, the vehicle position of the first vehicle k1 in the next time slot. The vehicle position of the first vehicle k2 in the next time slot

[0011]

[0012]

[0013] in, Indicates the position of the first vehicle; Indicates the position of the second vehicle; This indicates the x-axis coordinate position of the first vehicle k1 in the current time slot n; This indicates the y-axis coordinate position of the first vehicle k1 in the current time slot n; This indicates the x-axis coordinate position of the second vehicle k2 in the current time slot n; The y-coordinate of the second vehicle k2 in the current time slot n is represented by: n+1 represents the next time slot of the current time slot n; v1(n) represents the velocity of the first vehicle in the current time slot n; v2(n) represents the velocity of the second vehicle in the current time slot n; a1(n) represents the acceleration of the first vehicle in the current time slot n; a2(n) represents the acceleration of the second vehicle in the current time slot n; w1(n) represents the angular velocity of the first vehicle in the current time slot n; w2(n) represents the angular velocity of the second vehicle in the current time slot n; θ1(n) represents the deflection angle of the first vehicle in the current time slot n; θ2(n) represents the deflection angle of the second vehicle in the current time slot n.

[0014] Preferably, the information specifying each vehicle state type includes the vehicle's position, speed, yaw angle, and angular acceleration in the current time slot.

[0015] Preferably, in step B, the following steps are specifically performed to obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory:

[0016] Step B1: Based on the UAV equipped with a RIS reflector unit as a communication relay, and combined with the communication rate model of the two vehicles, construct the optimization problem of communication between the two vehicles in the next time slot in the current time slot.

[0017] Step B2: Based on the predicted vehicle positions of the first and second vehicles in the next time slot, solve the optimization problem of communication between the two vehicles in the next time slot of the current time slot to obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory.

[0018] Preferably, in step B1, based on the UAV equipped with a RIS reflector unit as a communication relay, the communication rate model between the two vehicles is expressed as follows:

[0019]

[0020] in,

[0021]

[0022]

[0023]

[0024] In the formula, R b [i] represents the communication rate between the two vehicles in time slot i, i.e., the communication rate model between the two vehicles; B w For bandwidth; h1[i] represents the transmit power of the first vehicle k1; h2[i] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot i; Θ[i] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot i; σ[i] represents the phase shift matrix of the RIS reflection unit in time slot i; 2 ρ represents noise power; α represents path loss at a reference distance of 1m; α represents the path loss exponent; λ represents the carrier wavelength. This represents the distance between the first vehicle k1 and the drone in time slot i; This represents the distance between the second vehicle k2 and the UAV in time slot i; M represents the total number of RIS reflection units; This represents the cosine of the signal angle from the first vehicle k1 to the drone; The x represents the cosine of the signal angle from the drone to the second vehicle k2; A [i] represents the x-axis coordinate of the UAV in time slot i, y A [i] represents the y-axis coordinate position of the UAV in time slot i, and H represents the z-axis coordinate position of the UAV in time slot i, which is a fixed altitude; This represents the z-axis coordinate position of the first vehicle k1 in time slot i; This indicates the z-axis coordinate position of the second vehicle k2 in time slot i.

[0025] Preferably, in step B1, the optimization problem of communication between the two vehicles in the next time slot of the current time slot is to maximize the communication rate between the two vehicles in the next time slot of the current time slot, as follows:

[0026]

[0027] st

[0028]

[0029] E UAV ≤E MAX

[0030] ||l[n+1]-l[n]|| 2 ≤D l 2 n = 1, 2, ..., N-1

[0031] ||l[1]-l0|| 2 ≤D l 2

[0032] Among them, R b [n+1] represents the communication rate between the two vehicles in the next time slot n+1 of the current time slot; Φ represents the phase shift matrix in the RIS reflection unit; Θ[n+1] represents the phase shift matrix in the RIS reflection unit of time slot n+1; B w For bandwidth; L represents the drone's location; h1[n+1] represents the transmit power of the first vehicle k1; h2[n+1] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot n+1; h2[n+1] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot n+1; σ 2 Indicates noise power; P represents the delivery probability; ΔT is the communication duration between the two vehicles using the drone as a communication relay; B represents the total amount of data generated by the two vehicles during the time interval ΔT; N represents the total number of communication time slots; P th E represents the preset delivery probability threshold; UAV E represents the energy consumption of the drone in the current time slot. MAX The maximum energy consumption of the drone is represented by l[n]; l[n+1] represents the position of the drone in the current time slot n; l[n+1] represents the position of the drone in time slot n+1; l[1] represents the position of the drone in the first time slot; l0 represents the initial position of the drone; D l =V max τ, V max τ is the maximum speed of the drone; τ is the time slot duration.

[0033] Preferably, in step B2, based on the predicted vehicle positions of the first and second vehicles in the next time slot, the DDPG algorithm is used to solve the optimization problem of communication between the two vehicles in the next time slot of the current time slot, so as to obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory.

[0034] Preferably, for the optimization problem of communication between two vehicles in the next time slot of the current time slot, the DDPG algorithm uses a drone as the intelligent agent, and the communication environment between the drone and the two vehicles is used as the DDPG algorithm environment:

[0035] The state of the environment is set as: s(t)={h1(t),h2(t),R b (t),B k}

[0036] Where h1[t] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot t; h2[t] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot t; R b (t) represents the communication rate between the two vehicles in time slot t; B k s(t) represents the remaining effective load in the environmental state space at time slot t; s(t) represents the state in the environmental state space at time slot t.

[0037] Based on the state of the environment, the actions of the agent in the environment include determining the RIS phase shift matrix and the UAV trajectory, i.e., a(t)={Φ,L};

[0038] in, This represents the phase shift matrix of the RIS reflection unit in time slot t;

[0039] a(t) represents the drone trajectory in time slot t; a(t) represents the agent's action in the environmental state space in time slot t;

[0040] The reward function is obtained based on the agent's actions in the environment:

[0041] Where r(t) represents the reward in time slot t;

[0042] Based on the environmental state, the agent's actions within the environment, and the reward function, the environment's evaluation function Q is:

[0043]

[0044] Where π represents the policy, i.e., the agent's action in the environment; E represents the expectation; γ∈[0,1] represents the discount factor of r(t), and r(s) t ,a t) represents the immediate reward for taking action a in environmental state s under time slot t.

[0045] Preferably, in the DDPG algorithm, the optimal strategy π * Through the objective function π * =argmaxQ π (s,a) is obtained, where π represents the policy, s represents the state of the environment, and a represents the action of the agent in the environment.

[0046] A UAV equipped with a RIS-assisted vehicle network communication system, applied to the UAV equipped with a RIS-assisted vehicle network communication method as described in claim 1, is characterized by comprising a position prediction module and an optimization module; the position prediction module predicts the vehicle positions of the first vehicle and the second vehicle in the next time slot based on information about the specified vehicle state types of the first vehicle and the second vehicle in the current time slot; the optimization module optimizes the phase shift matrix of the RIS reflection unit and the UAV trajectory based on the predicted vehicle positions of the first vehicle and the second vehicle in the next time slot.

[0047] The beneficial effects of this invention are as follows: This invention provides a method and system for UAV-equipped RIS-assisted vehicle network communication. In a scenario where a UAV is used for communication within a RIS-assisted vehicle network, the quality of the predicted position signal reflected to the vehicle is enhanced by optimizing the phase shift matrix of the RIS. Simultaneously, the UAV trajectory is optimized based on a vehicle position prediction model to determine the UAV's position in each time slot, thereby better serving vehicle communication within the coverage area. Under the constraints of vehicle communication delay, reliability, and UAV energy consumption, the DDPG algorithm is used to quickly solve the optimization problem of maximizing the communication rate between vehicles, obtaining the optimal strategy for inter-vehicle communication. This method maximizes the communication rate between vehicles by optimizing the RIS phase shift matrix and the UAV trajectory while satisfying V2V communication delay and UAV energy consumption constraints, thus improving the communication performance of the vehicle network. Attached Figure Description

[0048] Figure 1 A schematic diagram of the overall process of a UAV-equipped RIS-assisted vehicle network communication method provided in an embodiment of the present invention;

[0049] Figure 2 This is a schematic diagram of a UAV equipped with a RIS-assisted vehicle network communication scenario provided in an embodiment of the present invention. Detailed Implementation

[0050] The present invention will be further described below with reference to the accompanying drawings. The following embodiments will enable those skilled in the art to more fully understand the present invention, but do not limit the present invention in any way.

[0051] A method for RIS-assisted vehicle network communication using a drone (UAV) is proposed. For two vehicles (k1 and k2) with communication barriers within a target area, a UAV equipped with a predetermined number of RIS reflector units acts as a communication relay. Specifically, each UAV carries a RIS element with M reflector units. Figure 1 As shown, the following steps are performed to maximize the communication rate between the two vehicles in each time slot. This embodiment considers a crossroads scenario, such as... Figure 2 As shown, vehicles k1 and k2 are traveling in their respective lanes, and buildings obstruct direct communication between the two vehicles.

[0052] Step A: Based on the information of the first vehicle and the second vehicle in the current time slot, predict the vehicle positions of the first vehicle and the second vehicle in the next time slot respectively; the information of the specified vehicle state types includes the vehicle's position, speed, yaw angle and angular acceleration in the current time slot.

[0053] Based on the information regarding the specified vehicle state types of the first and second vehicles in the current time slot, the vehicle positions of the first and second vehicles in the next time slot are predicted using the following formula, i.e., the vehicle position of the first vehicle k1 in the next time slot. The vehicle position of the first vehicle k2 in the next time slot

[0054]

[0055] in, Indicates the position of the first vehicle; Indicates the position of the second vehicle; This indicates the x-axis coordinate position of the first vehicle k1 in the current time slot n; This indicates the y-axis coordinate position of the first vehicle k1 in the current time slot n; This indicates the x-axis coordinate position of the second vehicle k2 in the current time slot n; Let v1(n) represent the y-coordinate position of the second vehicle k2 in the current time slot n; n+1 represents the next time slot of the current time slot n; v1(n) represents the velocity of the first vehicle in the current time slot n; v2(n) represents the velocity of the second vehicle in the current time slot n; a1(n) represents the acceleration of the first vehicle in the current time slot n; a2(n) represents the acceleration of the second vehicle in the current time slot n; w1(n) represents the angular velocity of the first vehicle in the current time slot n; w2(n) represents the angular velocity of the second vehicle in the current time slot n; θ1(n) represents the deflection angle of the first vehicle in the current time slot n; θ2(n) represents the deflection angle of the second vehicle in the current time slot n. Considering a three-dimensional coordinate system, in the current time slot n, the UAV coordinates are (x... A [n],y A[n],H), the UAV senses the current state information of the vehicle, including the position coordinates of vehicle k1 and k2, respectively. and The flight period T of the drone is divided into N identical time slots, where the time slot duration is a sufficiently small constant τ. This means that the data corresponding to each time slot approximates the data at a single moment in time. And the flight altitude H remains constant. Within the constant τ, the drone and vehicle are assumed to be stationary.

[0056] After predicting the vehicle's next position, the UAV feeds it back to the RIS. By optimizing the phase shift matrix of the RIS reflection unit, the quality of the reflected signal is enhanced, and the position information is transmitted to the vehicle, facilitating highly reliable communication between vehicles. At the same time, the UAV optimizes its own trajectory L based on the predicted position of the vehicle to determine the position to fly to the next time slot, so as to better serve vehicle communication within the coverage area.

[0057] Step B: Based on the predicted vehicle positions of the first and second vehicles in the next time slot, obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory;

[0058] In step B, the following steps are specifically performed to obtain the optimized RIS reflection unit phase shift matrix and UAV trajectory:

[0059] Step B1: Based on the UAV equipped with a RIS reflector unit as a communication relay, and combined with the communication rate model of the two vehicles, construct the optimization problem of communication between the two vehicles in the next time slot in the current time slot.

[0060] Since the current time slot is the nth time slot, the channel gain between vehicle k1 and RIS is expressed as:

[0061]

[0062] in, This represents the distance between vehicle k1 and RIS. ρ is the path loss when the reference distance is 1m, α is the path loss exponent, and λ represents the carrier wavelength. This represents the cosine of the signal angle from vehicle k1 to RIS.

[0063] Similarly, in the nth time slot, the channel gain between RIS and vehicle k2 is expressed as:

[0064]

[0065] in This represents the distance between vehicle k2 and RIS. This represents the cosine of the signal angle from RIS to vehicle k2.

[0066] Therefore, in the nth time slot, the communication rate between the two vehicles is expressed as:

[0067]

[0068] Among them B w For bandwidth, This represents the transmission power of vehicle k1. In this scheme, the vehicle's transmission power is set to remain constant in each time slot. σ 2 This represents noise power.

[0069] Therefore, in the (n+1)th time slot, the communication rate between the two vehicles is expressed as:

[0070]

[0071] Therefore, based on the UAV equipped with a RIS reflector unit acting as a communication relay, the communication rate model between the two vehicles is expressed as follows:

[0072]

[0073] in,

[0074]

[0075]

[0076]

[0077] In the formula, R b [i] represents the communication rate between the two vehicles in time slot i, i.e., the communication rate model between the two vehicles; B w For bandwidth; h1[i] represents the transmit power of the first vehicle k1; h2[i] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot i; Θ[i] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot i; σ[i] represents the phase shift matrix of the RIS reflection unit in time slot i; 2 ρ represents noise power; α represents path loss at a reference distance of 1m; α represents the path loss exponent; λ represents the carrier wavelength. This represents the distance between the first vehicle k1 and the drone in time slot i; This represents the distance between the second vehicle k2 and the RIS reflection unit A in time slot i; M represents the total number of RIS reflection units. This represents the cosine of the signal angle from the first vehicle k1 to the drone; The x represents the cosine of the signal angle from the drone to the second vehicle k2; A [i] represents the x-axis coordinate of the UAV in time slot i, y A[i] represents the y-axis coordinate position of the drone in time slot i, and H represents the z-axis coordinate position of the drone in time slot i, which is a preset fixed height; This represents the z-axis coordinate position of the first vehicle k1 in time slot i; This represents the z-axis coordinate position of the second vehicle k2 in time slot i, which is set to 0 in the scheme.

[0078] For vehicle network systems to communicate, they need to meet the latency and reliability constraints of V2V communication, as well as the energy consumption and trajectory constraints of UAVs. In step B1, the optimization problem of communication between two vehicles in the next time slot of the current time slot is to maximize the communication rate of the two vehicles in the next time slot of the current time slot, that is, to take the optimization problem of maximizing the V2V communication rate of the (n+1)th time slot, and under the constraints of V2V communication latency and UAV energy consumption and trajectory, it is expressed as follows:

[0079]

[0080] st

[0081]

[0082] E UAV ≤E MAX

[0083] ||l[n+1]-l[n]|| 2 ≤D l 2 n = 1, 2, ..., N-1

[0084] ||l[1]-l0|| 2 ≤D l 2

[0085] Among them, R b [n+1] represents the communication rate between the two vehicles in the next time slot n+1 of the current time slot; Φ represents the phase shift matrix in the RIS reflection unit; Θ[n+1] represents the phase shift matrix in the RIS reflection unit of time slot n+1; B w For bandwidth; L represents the drone's location; h1[n+1] represents the transmit power of the first vehicle k1; h2[n+1] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot n+1; h2[n+1] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot n+1; σ 2 Indicates noise power; P represents the delivery probability; ΔT is the communication duration between the two vehicles using the drone as a communication relay, i.e., T; B represents the total data size required for communication between the two vehicles within ΔT, i.e., the payload size that the V2V link needs to deliver; N represents the total number of communication time slots; Pth E represents the preset delivery probability threshold; UAV E represents the energy consumption of the drone in the current time slot. MAX The maximum energy consumption of the drone is represented by l[n]; l[n+1] represents the position of the drone in the current time slot n; l[n+1] represents the position of the drone in time slot n+1; l[1] represents the position of the drone in the first time slot; l0 represents the initial position of the drone; D l =V max τ, V max τ is the maximum speed of the drone; τ is the time slot duration.

[0086] The constraints represent the following: the latency and reliability constraints of V2V communication are expressed as a measure of the probability of successful delivery, defined as the probability of successfully delivering a V2V communication payload of size B bits within an acceptable time T. This means that the probability of successfully delivering a B-bit V2V communication payload within an acceptable time ΔT is not less than its delivery threshold P. th E UAV ≤E MAX This means that the current energy consumption of the UAV must not exceed the maximum energy consumption of the UAV; the distance the UAV flies in each time slot must not exceed D. l =V max τ, V max It is the maximum speed of flight.

[0087] Since optimization problems are non-convex functions with high complexity, DDPG is used to solve them. The agent inputs the current state into the action network based on the current state of the environment and the existing decision-making strategy, outputs the Q value corresponding to each action, evaluates the network to estimate the state-action value, and substitutes the selected action and the next state into the target network for solving.

[0088] Step B2: Based on the predicted vehicle positions of the first and second vehicles in the next time slot, solve the optimization problem of communication between the two vehicles in the next time slot from the current time slot using the DDPG algorithm to obtain the optimized RIS reflector phase shift matrix and UAV trajectory. That is, obtain the optimal action for the optimal RIS phase shift matrix and optimal UAV trajectory to maximize vehicle communication rate.

[0089] To address the optimization problem of communication between two vehicles in the next time slot from the current time slot, the DDPG algorithm uses a drone as an agent to interact with the vehicle communication environment. The communication environment between the drone and the two vehicles serves as the DDPG algorithm environment.

[0090] The state of the environment is set as: s(t)={h1(t),h2(t),R b (t),B k}

[0091] Where h1[t] represents the channel gain between the first vehicle k1 and the RIS reflection unit in time slot t; h2[t] represents the channel gain between the second vehicle k2 and the RIS reflection unit in time slot t; R b (t) represents the communication rate between the two vehicles in time slot t; B k s(t) represents the remaining effective load in the environmental state space at time slot t; s(t) represents the state in the environmental state space at time slot t.

[0092] Based on the state of the environment, the actions of the agent in the environment include determining the RIS phase shift matrix and the UAV trajectory, i.e., a(t)={Θ,L};

[0093] in, This represents the phase shift matrix of the RIS reflection unit in time slot t; a(t) represents the drone trajectory in time slot t; a(t) represents the agent's action in the environmental state space in time slot t;

[0094] The reward function is obtained based on the agent's actions in the environment: After performing action a(t) in the environmental state space, a reward is returned to the vehicle communication system, which is defined as maximizing the V2V communication rate;

[0095] Where r(t) represents the reward in time slot t;

[0096] Based on the environmental state, the agent's actions within the environment, and the reward function, the environment's evaluation function Q is:

[0097]

[0098] Where π represents the policy, i.e., the agent's action in the environment; E represents the expectation; γ∈[0,1] represents the discount factor of r(t), and r(s) t ,a t ) represents the immediate reward for taking action a in environmental state s at time slot t+i.

[0099] In the DDPG algorithm, the optimal strategy π at the current time... * pass We obtain the following: where π represents the policy, s represents the state of the environment, and a represents the agent's action in the environment. The intelligent agent takes actions according to the policy π, which is a mapping from the state space S to the action space A, denoted as s(t)∈S→a(t)∈A. During the learning process, the agent learns to optimize its policy to achieve a better policy through experience transformation, with the optimal policy π being the optimal policy. * It can be done We obtain the result. During the DDPG algorithm solution process, the Q value will converge to the optimal Q with probability 1.*

[0100] Step C: Based on the optimized RIS reflector phase shift matrix and the UAV trajectory, adjust the RIS reflector phase shift matrix and the UAV's arrival position in the next time slot. Based on the optimization results Θ obtained using the DDPG algorithm... * ,L * The phase shift factor of the RIS reflection unit is adjusted, and the position reached by the UAV in the next time slot is determined, thereby maximizing the communication rate between the two vehicles in the next time slot. During the communication time between the two vehicles, the UAV trajectory and RIS phase shift matrix variables in each time slot are optimized to maximize the vehicle communication rate in the next time slot and ensure the quality of vehicle communication until the communication between the two vehicles ends or there is no communication obstruction.

[0101] The DDPG algorithm is used to solve optimization problems. Its core components are the policy function and the Q-value function, which can be approximated as an Actor network and a Critic network, respectively. During the DDPG algorithm's solution process, the Q-value converges to the optimal Q-value with probability 1. * It can exhibit stability and high convergence in the optimization of a series of continuous action spaces, including the following specific steps:

[0102] 1) Initialize the experience buffer D and the total number of time slots N;

[0103] 2) Initialize the Actor network μ, the Critic network Q, and the network parameters θ μ and θ Q ;

[0104] 3) Initialize the Target-Actor network μ′ and the Target-Critic network Q′, as well as the network parameters θ. μ′ and θ Q′ ;

[0105] 4) For each episode, repeat the following steps for training:

[0106] (1) Initialize the vehicle communication network environment and select the initial state s1.

[0107] (2) For each step in the round, repeat the following steps:

[0108] ①Based on the current input state s t The Actor network executes the output action a t Receive instant reward r t and the next state s t+1 To further obtain empirical data (s t ,a t ,r t ,s t+1);

[0109] ② Empirical data (s) t ,a t ,r t ,s t+1 Stored in the experience cache D;

[0110] ③ When the experience buffer D is full, a small batch of experience data (s) is randomly sampled from the experience buffer D. i ,a i ,r i ,s i+1 );

[0111] ④ Calculate the expected reward y of the current action using the Target-Critic network. i :

[0112] y i =r i +γQ′(s i+1 ,max Q(s i+1 |θ Q )|θ Q′ )

[0113] r i Q′(s) represents the instantaneous reward obtained in time slot i. i+1 ,maxQ(s i+1 |θ Q )θ Q′ This indicates that, based on the action corresponding to the maximum Q value obtained in time slot i, the next state s is entered. i+1 The corresponding Q value.

[0114] ⑤ Define a loss function to minimize the Critic network to update the Critic network parameters θ. Q :

[0115]

[0116] ⑥ Update the Actor network parameters using the following gradient function:

[0117]

[0118] ⑦ Update the parameters of the Target-Actor network and the Target-Critic network using the following formula:

[0119]

[0120] Where κ << 1.

[0121] (3) End the step loop.

[0122] 5) End the episode loop.

[0123] 6) After training, obtain the optimal strategy for optimizing the UAV trajectory and RIS phase shift matrix to maximize the communication rate.

[0124] The DDPG algorithm is used to obtain the optimization results, and the phase shift matrix of the optimized RIS reflection unit and the UAV trajectory variables are obtained for each time slot. The RIS phase shift factor is then adjusted to determine the position to be reached by the UAV in the next time slot, thereby maximizing the vehicle communication rate in the next time slot and ensuring the quality of vehicle communication.

[0125] A UAV-equipped RIS-assisted vehicle network communication system, applied to the aforementioned UAV-equipped RIS-assisted vehicle network communication method, includes a position prediction module and an optimization module. The position prediction module predicts the positions of the first and second vehicles in the next time slot based on information about the specified vehicle state types of the first and second vehicles in the current time slot. The prediction model is used to determine the positions of vehicles k1 and k2 when they reach time slot n+1. and The optimization module optimizes the phase shift matrix of the RIS reflection unit and the UAV trajectory based on the predicted vehicle positions of the first and second vehicles in the next time slot.

[0126] Reconfigurable Smart Surfaces (RIS) have become an important transmission technology for improving spectrum or energy efficiency in next-generation (5G (B5G) and 6G) ​​wireless communication networks, and have numerous applications in the fields of IoT and vehicle communication systems. RIS, composed of a large number of reflective elements, can passively reflect incident signals by intelligently adjusting its phase shift through a RIS controller, improving link quality and significantly enhancing coverage. Compared to traditional communication aids (such as relays), RIS consumes less energy due to passive reflection and can operate in full-duplex mode without self-interference. Unmanned aerial vehicles (UAVs), with their advantages of high maneuverability, low cost, and line-of-sight transmission, have been used as mobile airborne base stations or airborne relays to improve ground communication performance by deploying at optimal flight locations. The Deep Deterministic Policy Gradient (DDPG) algorithm has higher convergence than the DQN algorithm. It has two networks: an Actor network to generate the current policy and a Critic network to evaluate the policy's merits in the current state. It can perform stably in the optimization of a series of continuous action spaces. By using drones as intelligent agents, actions are selected and executed based on the current state to obtain a reward function. The optimal strategy is then updated through feedback to obtain the optimal RIS phase shift matrix and drone trajectory variables, thereby maximizing the communication rate between vehicles and improving the performance of vehicle-to-everything (V2X) communication.

[0127] Therefore, this invention designs a method and system for UAV-equipped RIS-assisted vehicle network communication. In a scenario where a UAV-equipped RIS-assisted vehicle network communicates, the quality of the predicted position signal reflected to the vehicle is enhanced by optimizing the phase shift matrix of the RIS. Simultaneously, the UAV trajectory is optimized based on the vehicle position prediction model to determine the UAV's position in each time slot, thereby better serving vehicle communication within the coverage area. Under the constraints of vehicle communication delay, reliability, and UAV energy consumption, the DDPG algorithm is used to quickly solve the optimization problem of maximizing the communication rate between vehicles, obtaining the optimal strategy for inter-vehicle communication. This method maximizes the communication rate between vehicles by optimizing the RIS phase shift matrix and the UAV trajectory while satisfying V2V communication delay and UAV energy consumption constraints, thus improving the communication performance of the vehicle network.

[0128] The above are merely preferred embodiments of the present invention, but do not limit the patent scope of the present invention. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art can still modify the technical solutions described in the foregoing specific embodiments, or make equivalent substitutions for some of the technical features. Any equivalent structures made using the content of the present invention specification and drawings, directly or indirectly applied to other related technical fields, are similarly within the patent protection scope of the present invention.

Claims

1. A method for UAV-equipped RIS-assisted vehicle network communication, characterized in that: For two vehicles in the target area, in the presence of communication barriers, a UAV equipped with a preset number of RIS reflecting units is used as a communication relay to perform the following steps to maximize the communication rate of the two vehicles in each time slot: Step A: Based on the preset vehicle state information of the first vehicle and the second vehicle in the current time slot, the vehicle positions of the first vehicle and the second vehicle in the next time slot are predicted respectively; Step B: Based on the predicted vehicle positions of the first vehicle and the second vehicle in the next time slot, an optimized RIS reflecting unit phase shift matrix and UAV trajectory are obtained; Step C: Based on the optimized RIS reflecting unit phase shift matrix and UAV trajectory, the phase shift matrix of the RIS reflecting unit is adjusted, and the position reached by the UAV in the next time slot is adjusted; In the above step B, the following steps are performed to obtain the optimized RIS reflecting unit phase shift matrix and UAV trajectory: Step B1: Based on the UAV equipped with RIS reflecting units as a communication relay, combined with the communication rate model of the two vehicles, an optimization problem of the communication of the two vehicles in the next time slot in the current time slot is constructed; Step B2: Based on the predicted vehicle positions of the first vehicle and the second vehicle in the next time slot, the optimization problem of the communication of the two vehicles in the next time slot in the current time slot is solved to obtain the optimized RIS reflecting unit phase shift matrix and UAV trajectory; In step B1, based on the UAV equipped with RIS reflecting units as a communication relay, the communication rate model of the two vehicles is represented as follows: ； wherein , , , , In the formula, Indicates that the two vehicles are in the time slot The communication rate, i.e., the communication rate model between two vehicles; For bandwidth; Indicates the first vehicle The transmission power; Indicates the first vehicle In time slots between the RIS reflection unit and the RIS reflection unit Channel gain; Indicates the second vehicle In time slots between the RIS reflection unit and the RIS reflection unit Channel gain; Indicates in time slot The phase shift matrix of the RIS reflector unit; Indicates noise power; This represents the path loss when the reference distance is 1m. Indicates the path loss index; Indicates the carrier wavelength; Indicates the first vehicle With drones in time slots The distance; Indicates the second vehicle With drones in time slots The distance; M represents the total number of RIS reflector units; Indicates from the first vehicle The signal angle cosine of the drone; Indicating a transition from drones to second vehicles The signal angle cosine; Indicates that the drone is in the time slot x-axis coordinate position Indicates that the drone is in the time slot The y-axis coordinate position, Indicates that the drone is in the time slot The z-axis coordinate position is at a fixed height; Indicates the first vehicle In the time slot z-axis coordinate position; Indicates the second vehicle In the time slot The z-axis coordinate position.

2. The unmanned aerial vehicle-mounted RIS-assisted vehicle network communication method according to claim 1, characterized in that: In the step A, based on the preset vehicle state information of the first vehicle and the second vehicle in the current time slot, the vehicle positions of the first vehicle and the second vehicle in the next time slot are predicted by the following formula, i.e. the vehicle position of the first vehicle in the next time slot the vehicle position of the second vehicle in the next time slot ； ； wherein represents a position of the first vehicle; represents a position of the second vehicle; represents the first vehicle an x-axis coordinate position in a current time slot n; represents the first vehicle a y-axis coordinate position in a current time slot n; represents the second vehicle an x-axis coordinate position in a current time slot n; represents the second vehicle a y-axis coordinate position in a current time slot n; represents a next time slot of the current time slot n; represents a speed of the first vehicle in a current time slot n; represents a speed of the second vehicle in a current time slot n; represents an acceleration of the first vehicle in a current time slot n; an acceleration of the second vehicle in a current time slot n; represents an angular velocity of the first vehicle in a current time slot n; represents an angular velocity of the second vehicle in a current time slot n; represents a yaw angle of the first vehicle in a current time slot n; represents a yaw angle of the second vehicle in a current time slot n.

3. The method of claim 1, wherein: The preset vehicle state information includes the position, speed, bias angle, and angular acceleration of the vehicle in the current time slot.

4. The unmanned aerial vehicle-mounted RIS-assisted vehicle network communication method of claim 1, wherein: In step B1, the optimization problem of the communication of the two vehicles in the next time slot in the current time slot is to maximize the communication rate of the two vehicles in the next time slot in the current time slot, which is represented as follows: ； ； wherein, denotes the communication rate of the two vehicles in the next time slot of the current time slot ; denotes the RIS reflecting unit phase shift matrix ; denotes the RIS reflecting unit phase shift matrix in time slot ; is the bandwidth; denotes the UAV position; denotes the transmit power of the first vehicle ; denotes the channel gain between the first vehicle and the RIS reflecting unit in time slot ; denotes the channel gain between the second vehicle and the RIS reflecting unit in time slot ; denotes the noise power; denotes the delivery probability; is the communication duration of the two vehicles based on the UAV as a communication relay; denotes the total data size generated by the two vehicles for communication within the time ; N denotes the total number of communication time slots; denotes the preset delivery probability threshold; denotes the energy consumption of the UAV in the current time slot; denotes the maximum energy consumption of the UAV; denotes the position of the UAV in the current time slot n; denotes the position of the UAV in time slot ; denotes the position of the UAV in the first time slot; denotes the initial position of the UAV; , is the maximum speed of the UAV flight; 5. The method of claim 1, wherein: In step B2, based on the predicted vehicle positions of the first vehicle and the second vehicle in the next time slot, the optimization problem of the communication of the two vehicles in the next time slot in the current time slot is solved using the DDPG algorithm to obtain the optimized RIS reflecting unit phase shift matrix and UAV trajectory.

6. The unmanned aerial vehicle-mounted RIS-assisted vehicle network communication method according to claim 5, characterized in that: For the optimization problem of the communication of the two vehicles in the next time slot in the current time slot, in the DDPG algorithm, the UAV is used as an agent, and the communication environment of the UAV and the two vehicles is used as the DDPG algorithm environment: The state of the environment is set as: ; wherein, denotes the first vehicle channel gain between the first vehicle and the RIS reflecting unit at time slot denotes the second vehicle channel gain between the second vehicle and the RIS reflecting unit at time slot denotes the communication rate of the two vehicles at time slot denotes the remaining payload of the environment state space at time slot denotes the state of the environment state space at time slot t; Based on the state of the environment, the actions of the agent in the environment include determining a RIS phase shift matrix and a drone trajectory, i.e. ; wherein, denotes the RIS reflecting unit phase shift matrix at time slot t; denotes the drone trajectory at time slot t; denotes the action of the agent in the environment state space at time slot t; Based on the action of the agent in the environment, the reward function is obtained: ; wherein, represents the reward at time slot t; based on the state of the environment, the actions of the agent in the environment, a reward function, an evaluation function of the environment is: ； wherein, represents a policy, i.e. the actions of the agent in the environment; E represents the expectation; represents a discount factor of represents the immediate reward of taking action in time slot from the environment state .

7. The method of claim 6, wherein: In the DDPG algorithm, the optimal policy is obtained by the objective function where represents the policy, represents the state of the environment, represents the action of the agent in the environment.

8. An unmanned aerial vehicle-mounted RIS-assisted vehicle network communication system, applied to the unmanned aerial vehicle-mounted RIS-assisted vehicle network communication method of claim 1, characterized in that: It includes a position prediction module and an optimization module. The position prediction module predicts the vehicle positions of the first vehicle and the second vehicle in the next time slot based on the preset vehicle state information of the first vehicle and the second vehicle in the current time slot. The optimization module optimizes the phase shift matrix of the RIS reflecting unit and the UAV trajectory based on the predicted vehicle positions of the first vehicle and the second vehicle in the next time slot.

Images