A vehicle network symbiotic communication system and its secondary link capacity optimization method

Abstract

The application discloses a kind of Internet of Vehicles symbiosis communication systems and its secondary link capacity optimization method, it is related to 6G wireless communication network technical field.System includes: base station, intelligent omnidirectional auxiliary surface, passive Internet of Things equipment and vehicle user cluster, intelligent omnidirectional auxiliary surface realizes all-around communication link enhancement in spatial range by the way of providing additional link gain, while passive Internet of Things equipment can be through capturing radio frequency signal in environment to carry out energy collection, by the energy collected is used for the circuit energy supply of self device and the transmission of information, to this reaches the improvement system energy efficiency, by enhancing communication link and the effective cooperation of improving system energy efficiency, effectively alleviate the influence of secondary link double-fading effect, ensure that in the system capacity of main link is not disturbed, significantly improve the transmission performance of system secondary link, with very strong application value and development potential.

Description

Technical Field

[0001] The present invention relates to the field of 6G wireless communication network technology, and in particular to a vehicle network symbiotic communication system and a method for optimizing the secondary link capacity thereof. Background Art

[0002] Intelligent Transportation Systems (ITS) are considered a key component of future smart cities. Their large-scale application of wireless communications and advanced sensor technologies will revolutionize transportation, encompassing safety and passenger comfort. Future ITS will expand into multiple application areas, including safety monitoring, autonomous driving, road traffic management, and infotainment. 6G technology is considered the core wireless technology for future ITS, featuring comprehensive connectivity, secure data sharing, high-efficiency transmission, and high-speed computing. 6G will advance transportation planning by ensuring the high reliability of autonomous driving systems, sharing more detailed road traffic information, promoting the application of augmented reality (AR) and virtual reality (VR), and providing advanced multimedia and gaming experiences. 6G-enabled Internet of Vehicles (IoVs) will deliver high data rates in the terabit per second range, increase packet delivery rates to 99.99999%, and reduce communication latency to below 1 millisecond, significantly improving wireless communication reliability.

[0003] In the 6G era, the massive integration of Internet of Vehicles (IoVs) into wireless communication networks has led to a significant increase in spectrum resources. Simultaneously, the integration of a large number of Internet of Things (IoT) devices will also consume significant spectrum resources. This situation not only leads to competition for spectrum resources but also causes high energy consumption by IoT nodes. Furthermore, building this infrastructure also carries significant economic costs. Symbiotic radio (SR) is a collaborative, ambient backscatter communication system that offers high spectral efficiency, high energy efficiency, and low cost. An SR system consists of two primary communication links: a primary link (the base station-to-vehicle link, directly serving vehicle users) and a secondary link (the IoT link, transmitting IoT information collected by IoT nodes). The secondary link modulates its IoT information onto the received primary link signal and then reflects the modulated information back to the receiver (users) via a passive IoT device (the backscatter device, BD).

[0004] However, since the IoT devices in the backscatter link are passive, in order for the receiver to accurately identify the backscatter link and main link information, the symbol period of the IoT information must be much longer than the symbol period of the main link communication information. At the same time, due to the influence of the double fading effect of the secondary link, the transmission rate of the backscatter link will be reduced. Summary of the Invention

[0005] The present invention provides a vehicle network symbiotic communication system and a method for optimizing the secondary link capacity thereof, so as to solve the problem of reduced backscatter link transmission rate due to the influence of the double fading effect of the secondary link.

[0006] The present invention is achieved through the following technical solutions:

[0007] A first aspect of the present invention provides a vehicle-to-vehicle symbiotic communication system, the system comprising a base station, an intelligent omnidirectional assistive surface, a passive Internet of Things device, and a vehicle user cluster, the vehicle user cluster being located within a service range of the base station;

[0008] The base station sends a main link signal to the passive IoT device, the vehicle user cluster, and the intelligent omnidirectional auxiliary surface;

[0009] The passive IoT device collects energy by capturing radio frequency signals in the environment during non-task time slots, and uses the collected energy to modulate IoT information onto the received primary link signal during task time slots to obtain a secondary link signal, and sends the secondary link signal to the vehicle user cluster. The task time slot is the time slot when the vehicle user cluster communicates.

[0010] The intelligent omnidirectional auxiliary surface enhances the main link signal and the secondary link signal by simultaneously transmitting and reflecting the received main link signal.

[0011] The solution of the present invention adopts a new intelligent omnidirectional auxiliary surface (STAR-RIS) to achieve omnidirectional communication link enhancement in a spatial range by providing additional link gain. At the same time, wireless power supply technology is introduced. Passive IoT devices can collect energy by capturing radio frequency signals in the environment. The collected energy is used to supply circuit energy for their own devices and transmit information, thereby improving the energy efficiency of the system. Through the effective coordination of enhancing the communication link and improving the system energy efficiency, the impact of the double fading effect of the secondary link can be effectively alleviated, ensuring that the secondary link capacity is increased while the system capacity of the main link is not interfered with, thereby improving the communication rate.

[0012] Furthermore, the vehicle user clusters communicate with each other using a TDMA access method, and the users within the vehicle user clusters communicate with each other using a NOMA access method.

[0013] Furthermore, the passive IoT device modulates the IoT information onto the received main link signal by using the received main link signal as a carrier and modulating the received IoT information onto the main link signal using binary phase shift keying.

[0014] In a second aspect, the present application provides a method for optimizing the capacity of a secondary link of the vehicle-to-everything symbiotic communication system, comprising: establishing a capacity maximization model of the secondary link based on a TDMA time allocation coefficient, a NOMA power allocation coefficient, and a transmission and reflection coefficient matrix of the intelligent omnidirectional auxiliary surface, and solving the capacity maximization model to obtain an optimization scheme.

[0015] Further, the capacity of the secondary link is represented as:

[0016]

[0017] wherein, Cm(k) represents the capacity of the secondary link at the user k in the vehicle user cluster m, t m K represents the total number of users in the vehicle user cluster m, and is the signal-to-noise ratio of the secondary link at the user k in the vehicle user cluster m, and is represented as:

[0018]

[0019] wherein, K represents the total number of users in the vehicle user cluster m, and P B is the base station transmission power, is the passive Internet of Things device channel between the passive Internet of Things device BD is the channel between the base station and the vehicle user cluster m, h BR is the channel between the base station and the intelligent omnidirectional auxiliary surface, is the channel between the intelligent omnidirectional auxiliary surface and the vehicle user cluster m, and m represents the reflection coefficient of the passive Internet of Things device BD m v i = diag{φ1,φ2,…,φ N} represents the transmission and reflection coefficient matrix of the intelligent omnidirectional auxiliary surface, represents the thermal noise power generated by the active enhancement of the passive Internet of Things device BD m σ 2 represents the power of the zero-mean additive Gaussian white noise.

[0020] Further, the capacity maximization model of the secondary link is solved under the conditions of satisfying the minimum communication rate constraint of the primary link, the maximum power allocation constraint of NOMA, the TDMA time allocation coefficient constraint, and the unit modulus constraint of the intelligent omnidirectional auxiliary surface.

[0021] Further, the capacity maximization model of the secondary link is:

[0022]

[0023] wherein, Represents the coefficient moment of the intelligent omnidirectional auxiliary surface, i∈{t,r} is used to represent the transmission coefficient matrix and the reflection coefficient matrix respectively, ω=[ω m,1 ,ω m,2 ,…,ω m,k ,…,ω m,K ] T is the NOMA power allocation coefficient vector of K users in the vehicle user cluster m, T=[t1,t2,…,t m ,…,t M ] T is the time allocation coefficient vector of M vehicle user clusters.

[0024] The main link minimum communication rate constraint is:

[0025]

[0026] Among them, R m,k Indicates the main link communication rate, R m,k =E[t m log2(1+γ m,k )],t m represents the time allocation coefficient of vehicle user cluster m, γ m,k is the signal-to-noise ratio of the main link at user k in vehicle user cluster m, It is the minimum communication rate of the main link communication;

[0027]

[0028]

[0029] Where K is the total number of users in vehicle user cluster m, P B is the base station transmit power, ω m,k is the power allocation coefficient of user k in vehicle user cluster m, It is a passive IoT device The channel between user k in vehicle user cluster m, is the channel between the base station and the vehicle user cluster m, h BR It is the channel between the base station and the intelligent omnidirectional auxiliary surface. is the channel between the intelligent omnidirectional auxiliary surface and the vehicle user cluster m, α m Indicates passive IoT device BD m The reflection coefficient, v i =diag{φ1,φ2,…,φ N} represents the transmission and reflection coefficient matrix of the auxiliary surface of the smart omnidirectional surface, Indicates passive IoT device BD m Thermal noise power generated by active enhancement, σ2 represents the power of zero-mean additive white Gaussian noise;

[0030] The NOMA maximum power allocation constraint is:

[0031] The TDMA time allocation coefficient constraints are:

[0032] The unit modulus constraint of the intelligent omnidirectional auxiliary surface is:

[0033] Furthermore, the DDPG algorithm is used to solve the capacity maximization model, including the following steps:

[0034] S1, initialize the parameters of the Actor network and Critic network, set the experience replay buffer, define the learning rate parameters, reward decay coefficient, number of training rounds and number of training steps per round of the Actor network and Critic network;

[0035] S2, in each training round, an action value is generated by the Actor network based on the current state value;

[0036] S3, the action value obtained from S2 interacts with the environment to obtain a reward value and generate the next state value;

[0037] S4, storing the experience in the experience replay buffer, wherein the experience includes the current state value, the current action value, the current reward value and the next state value;

[0038] S5, randomly sample a batch of experiences from the experience replay buffer to train the Actor network and Critic network;

[0039] S6, calculate the value function Q value of the current state and update the parameters of the value function using the Critic network;

[0040] S7, using the Actor network and the current state value as input, calculates the value function Q value of the generated action value, and uses the value function Q value of the generated action value to update the Actor network parameters;

[0041] S8, repeat S2-S7 until the predetermined number of training steps is reached or the stopping condition is met;

[0042] S9, use the trained Actor network to make decisions and obtain the final strategy.

[0043] Furthermore, the state values ​​include: the channel gain from the base station directly reaching the vehicle user cluster, the channel gain from the base station reaching the vehicle user cluster through the intelligent omnidirectional auxiliary surface, the channel gain from the base station reaching the vehicle user cluster through the passive IoT device, and the action (T, ω, v) adopted by the algorithm. t ;

[0044] The action values ​​include: TDMA time allocation coefficient, transmission and reflection coefficient matrix of intelligent omnidirectional auxiliary surface and NOMA power allocation coefficient;

[0045] The reward value includes: In this training, if the primary link rate at user k meets the minimum rate requirement, then the secondary link rate at user k is used as the instantaneous reward and the instantaneous reward is added to the total reward. If the minimum rate constraint is not met, the instantaneous reward is given a penalty term of -0.2.

[0046] Compared with the prior art, the present invention has the following advantages and beneficial effects:

[0047] 1. The present invention adopts a smart omnidirectional auxiliary surface (STAR-RIS) combined with wireless power supply technology. Through the effective cooperation of STAR-RIS to enhance the communication link and wireless power supply technology to improve the system energy efficiency, it can effectively alleviate the impact of the double fading effect of the secondary link, ensuring that the system capacity of the primary link is not disturbed, while increasing the secondary link capacity, thereby improving the communication rate.

[0048] 2. Vehicle user clusters communicate using TDMA access, and users within a vehicle user cluster communicate using NOMA access, reducing main link information interference between different clusters and within clusters.

[0049] 3. By jointly optimizing the STAR-RIS coefficient matrix, NOMA power allocation coefficient, and TDMA time allocation coefficient, the secondary link capacity is maximized while meeting the minimum communication rate requirements of the primary and secondary links, significantly improving the backscatter link communication rate of all vehicle user clusters, and giving full play to the advantages of the symbiotic communication system of the present invention.

[0050] 4. By using deep reinforcement learning to solve the problem and verifying it through simulation, the present invention can ensure normal communication of the user's main link without generating additional spectrum and energy overhead or increasing additional costs. It can significantly improve the Internet of Things link transmission performance of the symbiotic communication system and has strong application value and development potential. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions of the exemplary embodiments of the present invention, the following briefly introduces the drawings required for use in the examples. It should be understood that the following drawings only illustrate certain embodiments of the present invention and should not be considered as limiting the scope. A person of ordinary skill in the art can also derive other relevant drawings based on these drawings without inventive effort. In the drawings:

[0052] Figure 1 This is an architecture diagram of a symbiotic communication system for an Internet of Vehicles according to an embodiment of the present invention;

[0053] Figure 2 1 is a schematic diagram of a frame structure of a signal sent by a base station according to an embodiment of the present invention;

[0054] Figure 3 This is a comparison chart of the secondary link and communication rate of the solution of the present invention and other solutions at different base station transmission powers;

[0055] Figure 4 This is a comparison chart of secondary links and rates between the solution of the present invention and other solutions at different numbers of intelligent auxiliary plane units. DETAILED DESCRIPTION

[0056] In order to make the objectives, technical solutions and advantages of the present invention more clearly understood, the present invention is further described in detail below in conjunction with examples and drawings. The exemplary embodiments of the present invention and their descriptions are only used to explain the present invention and are not intended to limit the present invention.

[0057] It should be noted that the terms "including" and "having" and any variations thereof in the specification and claims of the present invention and the above-mentioned drawings are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to other steps or units inherent in the device.

[0058] The terms used in the various embodiments of the application are only used to describe the purpose of specific embodiments and are not intended to limit the various embodiments of the application. As used herein, the singular form is intended to also include the plural form, unless the context clearly indicates otherwise. Unless otherwise limited, all terms used here (including technical terms and scientific terms) have the same meaning as the meaning generally understood by those of ordinary skill in the art of the application. The terms (such as the terms defined in the dictionary generally used) will be interpreted as having the same meaning as the contextual meaning in the relevant technical field and will not be interpreted as having idealized meaning or too formal meaning, unless clearly defined in the various embodiments of the application.

[0059] An embodiment of the present invention provides a vehicle network symbiotic communication system, which is suitable for 6G vehicle network communication, is conducive to alleviating the impact of the double fading effect of the secondary link and improving the communication rate of the vehicle network symbiotic system.

[0060] Example 1

[0061] like Figure 1 As shown, Figure 1 This is an architecture diagram of the vehicle network symbiotic communication system of the present invention, which includes a base station BS, an intelligent omnidirectional auxiliary surface STAR-RIS, a passive Internet of Things device BD and a vehicle user cluster.

[0062] Figure 1 This is an example of a communication network composed of a base station. Within the coverage area of ​​the base station, there is a STAR-RIS with N intelligent units, one or more passive Internet of Things devices BD, and one or more vehicle user clusters. The above constitute a minimum unit of the vehicle network symbiotic communication system. The vehicle network symbiotic communication system of the present invention can also include a plurality of such units.

[0063] The vehicle user cluster is to divide the users in the service area of ​​the base station BS according to the service area of ​​each passive IoT device BD. For example, Figure 1 The base station BS server in the BD is divided into M vehicle user clusters, each of which includes K users. m represents the mth vehicle user cluster in the M vehicle user clusters, 1≤m≤M, and k represents the kth user in cluster m. m represents the passive IoT devices covering cluster m.

[0064] The base station BS sends the main link signal to the passive IoT device BD, the vehicle user cluster and the intelligent omnidirectional auxiliary surface.

[0065] The base station (BS) uses time division multiple access (TDMA) via V2I to serve a cluster of M vehicles within its service area. TDMA is a communication technology for sharing a transmission medium (typically radio) or network. It allows multiple users to use the same frequency in different time slices (time slots). Each user uses their own time slice, allowing multiple users to share the same transmission medium (e.g., radio frequency).

[0066] The links between the base station BS and each passive IoT device BD, the link between the base station BS and each vehicle user cluster, and the link between the base station BS and the intelligent omnidirectional auxiliary surface are the main links. The base station BS sends signals to the passive IoT device BD, vehicle user cluster, and intelligent omnidirectional auxiliary surface through the main links. These signals become the main link signals.

[0067] The passive IoT device collects energy by capturing the RF signal in the environment during the non-task time slot. During the task time slot when the vehicle user cluster is communicating, the collected energy is used to modulate the IoT information onto the received primary link signal to obtain the secondary link signal, and then send the secondary link signal to the vehicle user cluster. Figure 2 The figure shows the frame structure diagram of the base station sending the signal. In M time slots, the vehicle user cluster m is in time slot t m Information is transmitted within t m Passive IoT device BD in time slot m The energy collected in other time slots is used for active enhanced information transmission.

[0068] Furthermore, in the task time slot, the passive IoT device BD uses the received main link signal as a carrier and modulates the received IoT information into the main link signal using binary phase shift keying (BPSK).

[0069] To accurately decode IoT information at the receiver, the IoT information symbol period must be significantly longer than the primary link symbol period. To accurately distinguish between primary and secondary link signals, the receiver (user) uses Successive Interference Cancellation (SIC) to accurately decode both signals. SIC is also used to decode the superimposed signals of NOMA user pairs sent by the base station.

[0070] The intelligent omnidirectional auxiliary surface enhances the main link signal and the secondary link signal by transmitting and reflecting the received main link signal at the same time.

[0071] Furthermore, to reduce interference between different clusters and within clusters on the main link, a TDMA+NOMA access method is considered. Vehicle user clusters communicate using TDMA access, while users within a vehicle user cluster communicate using non-orthogonal multiple access (NOMA). NOMA transmits multiple information streams at varying power levels on overlapping channels in the time, frequency, and code domains, providing wireless services to multiple users simultaneously on the same wireless resources. This access method minimizes inter-cluster interference, and continuous interference cancellation technology is used within the cluster to decode the NOMA superimposed signals.

[0072] The main link signal at the receiver includes the signal sent directly by the base station BS to the user cluster, the auxiliary signal passing through the auxiliary surface of the smart omnidirectional surface, and the main link information decoded from the IoT information transmitted to the user cluster from the IoT backscatter device BD.

[0073] The present invention adopts a smart omnidirectional auxiliary surface (STAR-RIS) combined with wireless power supply technology. Through the effective coordination of STAR-RIS enhanced communication links and wireless power supply technology to improve system energy efficiency, it can effectively alleviate the impact of the double fading effect of the secondary link, ensuring that the secondary link capacity is increased while the system capacity of the primary link is not interfered with, thereby improving the communication rate.

[0074] At the same time, in order to reduce the main link information interference between different clusters and within a cluster, a TDMA+NOMA access method is considered, which can eliminate inter-cluster interference to the greatest extent.

[0075] Example 2

[0076] In order to ensure the normal communication of the user main link in the user cluster, and without generating additional spectrum and energy overhead and without increasing additional costs, the sum rate of the backscatter links of all vehicle user clusters is significantly improved, and the advantages of the symbiotic communication system are fully utilized. This embodiment optimizes the secondary link capacity of the vehicle network symbiotic communication system in Example 1. Based on the TDMA time allocation coefficient, the NOMA power allocation coefficient, and the transmission and reflection coefficient matrix of the intelligent omnidirectional auxiliary surface, a capacity maximization model for the secondary link is established, and the optimization solution is obtained by solving the capacity maximization model. Specifically, the process of establishing the secondary link capacity maximization model is as follows.

[0077] Since the base station BS uses NOMA non-orthogonal multiple access in different time slots to transmit superimposed information for the cluster m served in the current time slot, without loss of generality, the signal-to-noise ratio between the base station and the K users in the vehicle user cluster m satisfies:

[0078]

[0079] In the vehicle user cluster m, the channel strong user The decoding is performed by using the serial interference cancellation technique SIC, that is, before decoding the information, the first K-1 user information is eliminated. SIC technology will not be used, but other user information will be decoded as interference.

[0080] At the base station BS, superposition coding technology is used to allocate different transmission powers to each signal. According to the above signal-to-noise ratio arrangement order, the power allocation coefficient arrangement can be obtained as follows:

[0081] ω m,1 ≥ω m,2 ≥…≥ω m,k ≥…≥ω m,K

[0082] From the above analysis, we can conclude that the superimposed signal sent by the base station BS to the K users in the cluster m is:

[0083]

[0084] Based on the above information, the signal received by user k in cluster m is:

[0085]

[0086] Among them, x m is the superimposed information symbol sent by the base station BS to the K users in the cluster m, and c represents the BD received by the cluster. m The information symbol sent by the device to cluster m, n represents the power σ 2 Zero-mean additive Gaussian white noise, v i =diag{φ1,φ2,…,φ N} represents the matrix coefficient of the auxiliary surface of the intelligent omnidirectional surface, z b The power generated by the active enhancement of the BD circuit is Thermal noise, α m Indicates BD m The reflection coefficient, is the channel between BS and user k in the mth cluster, h BR ∈C 1×N is the channel between BS and STAR-RIS, is the channel between STAR-RIS and user k in the mth cluster, for The channel between user k in the mth cluster, BS and BD m The channel between For STAR-RIS and BD m The channel between.

[0087] Define the STAR-RIS beamforming vector:

[0088]

[0089] With 0≤λ t ≤1 and 0≤λ r ≤1 represents the time allocation coefficient of STAR-RIS working in transmission mode and reflection mode respectively. In the system, the lengths of these two variables are consistent with the service time slot lengths on the transmission side and reflection side respectively.

[0090] In our design, wireless power supply technology is introduced to alleviate the impact of BD double fading effect, making BD mEnergy is collected in non-task slots, and in task slots, the collected energy is used to enhance the reflected incident signal. Assuming the current time slot length is t m Then the total time slot length of the energy collection phase is 1-t m Therefore, the energy collection formula of BDm is expressed as: η is the energy collection efficiency, P B The table is the base station transmit power.

[0091] The maximum transmission power of the reflection transmission phase is obtained as:

[0092]

[0093] The reflection coefficient is:

[0094]

[0095] According to the above information, the signal received at user k in vehicle user cluster m is:

[0096]

[0097] In the above received signal, the first segment is the received primary link signal, the second segment is the secondary link signal transmitted by the BD device, and the third segment is the thermal noise generated at the BD, which satisfies The last term is the received Gaussian white noise at the user.

[0098] In the mutualistic symbiotic wireless communication system, in order to ensure that the primary link information and the secondary link information can be accurately decoded, the transmission symbol period of the secondary link information is much smaller than that of the primary link information, so we can consider the secondary transmission link signal as an additional multipath component of the primary link information for decoding.

[0099] At the same time, when decoding information at user k in vehicle user cluster m, the SIC serial interference cancellation technology is used to cancel the interference of the information with weaker channel gain, and the information of the user with stronger channel gain is considered as interference for decoding. According to the above analysis, the signal-to-noise ratio at user k in vehicle user cluster m after SIC decoding is:

[0100]

[0101] where

[0102] Since the BD device uses BPSK modulation to modulate the collected ambient IoT information onto the ambient RF signal, the backscatter link information symbol c∈{0,1} is unknown at the user. Therefore, according to the Shannon formula theorem, the communication capacity of the system at user k in cluster m is approximately expressed as:

[0103] R m,k =E[t m log2(1+γ m,k )]

[0104] It is necessary to take the expected value of the random variable.

[0105] Since the backscatter link information is modulated onto the main link signal, and the length of the backscatter link information symbol is K times the length of the main link information symbol, when decoding the backscatter link information, the main link information symbol is regarded as a spread spectrum code of length K. And when the main link information is decoded, considering perfect SIC continuous interference cancellation, the secondary link information can be completely decoded and processed. In this process, we first subtract the main link information from the original signal, and then decode the IoT information transmitted by BD according to the maximum ratio combining MCR method. Therefore, the signal-to-noise ratio of the secondary link at user k in the vehicle user cluster m can be obtained as follows:

[0106]

[0107] When K is large enough, the decoding interference of the main link information on the backscatter link information is 0. According to Shannon's theorem, the secondary link capacity at user k in vehicle user cluster m can be obtained as:

[0108]

[0109] According to the model established above, in order to maximize the secondary link rate of all clusters while satisfying the minimum primary link rate constraint, the TDMA time allocation coefficient, NOMA power allocation coefficient, and STAR-RIS beamforming matrix are jointly optimized to reduce the impact of double fading and increase the secondary link system capacity. This ensures that the system capacity of the primary and secondary links is not interfered with, thus establishing a secondary link capacity maximization model:

[0110]

[0111] in, Represents the coefficient moment of the intelligent omnidirectional auxiliary surface, i∈{t,r} is used to represent the transmission coefficient matrix and the reflection coefficient matrix respectively, ω=[ω m,1 ,ω m,2 ,…,ω m,k ,…,ω m,K ] Tis the NOMA power allocation coefficient vector of K users in the vehicle user cluster m, T=[t1,t2,…,t m ,…,t M ] T is the time allocation coefficient vector of M vehicle user clusters.

[0112] Under the conditions of satisfying the minimum communication rate constraint of the primary link, the maximum power allocation constraint of NOMA, the time allocation coefficient constraint of TDMA, and the unit modulus value constraint of the intelligent omnidirectional auxiliary surface, the above secondary link capacity maximization model is solved. The optimization problem is as follows:

[0113]

[0114]

[0115]

[0116]

[0117]

[0118] The first constraint is the minimum communication rate requirement for the main link; the second constraint is used to ensure the maximum power allocation constraint of NOMA; the third constraint is the time allocation coefficient constraint; and the last constraint is the STAR-RIS unit modulus constraint.

[0119] Because the variables to be optimized include three coupled variables and the objective function involves the summation of logarithmic terms, Problem P1 is a difficult non-convex optimization problem. In this solution, rather than addressing the problem directly mathematically, we consider using Deep Reinforcement Learning (DRL) to obtain a feasible STAR-RIS coefficient matrix, NOMA power allocation coefficients, and time allocation coefficients between different clusters.

[0120] We considered using the DDPG algorithm as a solution. It consists of two networks: an actor network and a critic network. The actor network takes the state as input and outputs a continuous action, which is then fed into the critic network along with the state. The actor network is used to continuously approximate the optimal action, eliminating the non-convex optimization problem of finding the action that maximizes the Q value given the next state.

[0121] At the beginning of the algorithm, the experience replay buffer M, the parameters of the Critic network and the Actor network, and the action T,ω,v need to be initialized. In this paper, we simply use the identity matrix to initialize T,ω,v.

[0122] The algorithm runs over N rounds, with T training steps per round. In each round, the algorithm terminates when convergence or the maximum number of steps allowed is reached. The optimal action combination T, ω, v yields the highest immediate reward. The goal of this algorithm is to use DRL to obtain the optimal action combination T, ω, v, rather than training a neural network for online processing.

[0123] Problem P1 described in this article is solved using the DDPG algorithm. The specific process is as follows:

[0124] S1, initialize the parameters of the Actor network and Critic network, set the experience replay buffer, define the learning rate parameters, reward decay coefficient, number of training rounds and number of training steps per round of the Actor network and Critic network;

[0125] S2, in each training round, an action value is generated by the Actor network based on the current state value;

[0126] S3, the action value obtained from S2 interacts with the environment to obtain a reward value and generate the next state value;

[0127] S4, store the experience in the experience replay buffer, the experience includes the current state value, current action value, current reward value and next state value;

[0128] S5, randomly sample a batch of experiences from the experience replay buffer to train the Actor network and Critic network;

[0129] S6, calculate the value function Q value of the current state and update the parameters of the value function using the Critic network;

[0130] S7, using the Actor network and the current state value as input, calculates the value function Q value of the generated action value and uses this value to update the Actor network parameters;

[0131] S8, repeat S2-S7 until the predetermined number of training steps is reached or the stopping condition is met;

[0132] S9, use the trained Actor network to make decisions and obtain the final strategy.

[0133] The specific state s, action (T, ω, v), and immediate reward are described in detail as follows:

[0134] The state values ​​include: the channel gain from the base station directly reaching the vehicle user cluster, the channel gain from the base station reaching the vehicle user cluster through the intelligent omnidirectional auxiliary surface, the channel gain from the base station reaching the vehicle user cluster through the passive IoT device, and the action (T, ω, v) adopted by the algorithm.

[0135] That is, the current state value includes: the state s at time t (t) , the channel gains from the base station BS directly reaching M*K users at time t, the channel gains from the BS reaching M*K users through STAR-RIS, the channel gains from the BS reaching M*K users through M BDs, and the action (T, ω, v) taken by the algorithm at time t t ;

[0136] The action values ​​include: TDMA time allocation coefficient T, transmission and reflection coefficient matrix v of the intelligent omnidirectional auxiliary surface, and NOMA power allocation coefficient ω;

[0137] The reward value includes: In this training, if the primary link rate at user k meets the minimum rate requirement, then the secondary link rate at user k is used as the instantaneous reward and the instantaneous reward is added to the total reward. If the minimum rate constraint is not met, the instantaneous reward is given a penalty term of -0.2.

[0138] Example 3

[0139] This embodiment verifies the superiority of the system performance of the present invention through simulation experiments. Figure 1 As shown in the figure, the system consists of a single-antenna base station (BS), N intelligent STAR-RIS units, M*K vehicle users, and M BDs. The M*K users are divided into M vehicle user clusters based on the BD's service location. TDMA access is used between clusters, while NOMA is used to transmit information between vehicle users within a cluster. The intelligent omnidirectional auxiliary surface (STAR-RIS) reconstructs the incident signal, decomposing it into reflected and transmitted signals, intelligently reconstructing the wireless communication channel environment.

[0140] The system simulation parameters are set as follows: the number of intelligent omnidirectional auxiliary surface units is between 10 and 80, the base station BS transmission power p = 10 ~ 45dBm, the additive white Gaussian noise power is, the multiple of the main link information symbol period and the IoT link information symbol period K = 128; the distance from the base station BS to different users in the cluster is 10 ~ 40m, and the distance from STAR-RIS to different users in the cluster is 7.5 ~ 30m; the other distance variables are all fixed values, the distance from BS to BD, the distance from STAR-RIS to BD, and the distance from BD to user are 30m, 20m, and 20m respectively, the path loss average channel gain at the reference distance of 1m is 20dB, and the path loss exponent is -2.2. In M time slots, vehicle user cluster m in time slot t m Information transmission is carried out within BD m The energy collected in other time slots is used for active enhanced information transmission.

[0141] Figure 3 The following chart compares the secondary link and communication rate of the proposed solution (STAR-RIS combined with wireless power) with other solutions at different BS transmit powers. The comparison schemes include wireless power only, STAR-RIS only, STAR-RIS in place of traditional RIS and wireless power, and a traditional solution without any additional support. It can be seen that the introduction of STAR-RIS and wireless power significantly improves the total secondary link rate at different BD transmit powers.

[0142] Figure 4 The following chart compares the sub-link and rate of the proposed solution (STRA-RIS combined with wireless power supply technology) and other solutions at different numbers of intelligent auxiliary plane units. The comparison schemes include only wireless power supply, only STAR-RIS, replacing STAR-RIS with traditional RIS and wireless power supply, and traditional RIS without any auxiliary enhancement. It can be seen that as the number of intelligent units increases, the solutions using the intelligent auxiliary plane all have a certain improvement in sub-link and rate. At the same time, when both have auxiliary planes and wireless power supply technology for assistance, STAR-RIS and RIS have more significant improvements in sub-link and rate when configured with the same number of units.

[0143] The specific implementation methods described above further illustrate the objectives, technical solutions and beneficial effects of the present invention in detail. It should be understood that the above description is only a specific implementation method of the present invention and is not intended to limit the scope of protection of the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the scope of protection of the present invention.

Claims

1. A method for optimizing the secondary link capacity of a symbiotic communication system of an Internet of Vehicles, characterized in that: The optimization method includes: establishing a capacity maximization model for the secondary link based on the TDMA time allocation coefficient, the NOMA power allocation coefficient, and the transmission and reflection coefficient matrix of the intelligent omnidirectional auxiliary surface, and solving the capacity maximization model under the conditions of satisfying the minimum communication rate constraint of the primary link, the NOMA maximum power allocation constraint, the TDMA time allocation coefficient constraint, and the unit modulus value constraint of the intelligent omnidirectional auxiliary surface to obtain an optimization solution; The capacity of the secondary link is expressed as: in, represents the capacity of the secondary link at user k in vehicle user cluster m, t m represents the time allocation coefficient of vehicle user cluster m, K is the total number of users in vehicle user cluster m, is the signal-to-noise ratio of the secondary link at user k in vehicle user cluster m, expressed as: Among them, P B is the base station transmit power, It is a passive IoT device The channel between user k in vehicle user cluster m, is the channel between the base station and the vehicle user cluster m, h BR It is the channel between the base station and the intelligent omnidirectional auxiliary surface. is the channel between the intelligent omnidirectional auxiliary surface and the vehicle user cluster m, α m Indicates passive IoT device BD m The reflection coefficient, v i =diag{φ1,φ2,…,φ N } represents the transmission and reflection coefficient matrix of the auxiliary surface of the smart omnidirectional surface, Indicates passive IoT device BD m Thermal noise power generated by active enhancement, σ 2 represents the power of zero-mean additive white Gaussian noise; The capacity maximization model of the secondary link is: in, Represents the coefficient matrix of the intelligent omnidirectional auxiliary surface, i∈{t,r} is used to represent the transmission coefficient matrix and the reflection coefficient matrix respectively, is the NOMA power allocation coefficient vector of K users in the vehicle user cluster m, T=[t1,t2,…,t m ,…,t M ] T is the time allocation coefficient vector of M vehicle user clusters; The main link minimum communication rate constraint is: Among them, R m,k Indicates the main link communication rate, R m,k =E[t m log2(1+γ m,k )],t m represents the time allocation coefficient of vehicle user cluster m, γ m,k is the signal-to-noise ratio of the main link at user k in vehicle user cluster m, It is the minimum communication rate of the main link communication; Among them, ω m,k is the power allocation coefficient of user k in vehicle user cluster m, and c represents the BD received by the cluster. m Information symbols sent by the device to cluster m; The NOMA maximum power allocation constraint is: The TDMA time allocation coefficient constraints are: The unit modulus constraint of the intelligent omnidirectional auxiliary surface is:

2. The secondary link capacity optimization method of the vehicle network symbiotic communication system according to claim 1, characterized in that: The DDPG algorithm is used to solve the capacity maximization model, including the following steps: S1, initialize the parameters of the Actor network and Critic network, set the experience replay buffer, define the learning rate parameters, reward decay coefficient, number of training rounds and number of training steps per round of the Actor network and Critic network; S2, in each training round, an action value is generated by the Actor network based on the current state value; S3, the action value obtained from S2 interacts with the environment to obtain a reward value and generate the next state value; S4, storing the experience in the experience replay buffer, wherein the experience includes the current state value, the current action value, the current reward value and the next state value; S5, randomly sample a batch of experiences from the experience replay buffer to train the Actor network and Critic network; S6, calculate the value function Q value of the current state and update the parameters of the value function using the Critic network; S7, using the Actor network and the current state value as input, calculates the value function Q value of the generated action value, and uses the value function Q value of the generated action value to update the Actor network parameters; S8, repeat S2-S7 until the predetermined number of training steps is reached or the stopping condition is met; S9, use the trained Actor network to make decisions and obtain the final strategy.

3. The secondary link capacity optimization method of the vehicle network symbiotic communication system according to claim 2, characterized in that: The state values ​​include: the channel gain from the base station directly reaching the vehicle user cluster, the channel gain from the base station reaching the vehicle user cluster through the intelligent omnidirectional auxiliary surface, the channel gain from the base station reaching the vehicle user cluster through the passive IoT device, and the action (T, ω, v) adopted by the algorithm. t ; The action values ​​include: TDMA time allocation coefficient, transmission and reflection coefficient matrix of the intelligent omnidirectional auxiliary surface and NOMA power allocation coefficient; The reward value includes: in this training, if the primary link rate at user k meets the minimum rate requirement, then the secondary link rate at user k is used as the instantaneous reward, and the instantaneous reward is added to the total reward; if the minimum rate constraint is not met, the instantaneous reward is given a penalty item of -0.

2.

4. A symbiotic communication system for an Internet of Vehicles, characterized in that: The system is optimized by applying the secondary link capacity optimization method of the vehicle network symbiotic communication system according to any one of claims 1 to 3; the system includes a base station, an intelligent omnidirectional auxiliary surface, a passive Internet of Things device and a vehicle user cluster, and the vehicle user cluster is located within the service range of the base station; The base station sends a main link signal to the passive IoT device, the vehicle user cluster, and the intelligent omnidirectional auxiliary surface; The passive IoT device collects energy by capturing radio frequency signals in the environment during non-task time slots, and uses the collected energy to modulate IoT information into a received primary link signal during a task time slot to obtain a secondary link signal, and sends the secondary link signal to the vehicle user cluster. The task time slot is a time slot when the vehicle user cluster communicates. The intelligent omnidirectional auxiliary surface enhances the main link signal and the secondary link signal by simultaneously transmitting and reflecting the received main link signal.

5. The vehicle network symbiotic communication system according to claim 4, characterized in that: The vehicle user clusters communicate with each other using a TDMA access mode, and the users within the vehicle user clusters communicate with each other using a NOMA access mode.

6. The vehicle network symbiotic communication system according to claim 4, characterized in that: The passive Internet of Things device modulates the Internet of Things information onto the received main link signal in the following manner: using the received main link signal as a carrier and adopting binary phase shift keying to modulate the received Internet of Things information onto the main link signal.

Images