Vehicle-mounted twinborn migration resource transaction method in vehicle-mounted element universe

Abstract

The invention discloses a resource transaction method for vehicle-mounted twinning migration in a vehicle-mounted element universe, and the method comprises the steps: constructing a queuing time delay prediction model based on a yoke theory for the real-time migration process of vehicle-mounted twinning between a source unmanned aerial vehicle and a target unmanned aerial vehicle, deducing the time delay violation probability of the migration process through the analysis of an arrival process and a service process, and carrying out the calculation of the time delay violation probability. And establishing a quantitative relationship between the migration reliability and the allocated bandwidth. Furthermore, the bandwidth transaction behavior between the meta universe service provider and the vehicle-mounted user is considered, a Stackelberg game model is constructed, the service provider is used as a leader to formulate the optimal bandwidth pricing, and the vehicle-mounted user is used as a follower to select the bandwidth purchase amount to minimize the quality index. Finally, game equilibrium is solved through iteration of a reverse induction method, an optimal pricing and resource allocation strategy is obtained, and the real-time performance of twinborn migration and the overall benefit of the system are effectively improved. The method can be widely applied to edge computing resource scheduling and communication optimization scenes in a high-dynamic vehicle-mounted element universe environment.

Description

TECHNICAL FIELD

[0001] The application relates to the technical field of vehicle-mounted meta-universe communication and resource optimization, and particularly relates to a resource transaction method for vehicle-mounted twin migration in a vehicle-mounted meta-universe. BACKGROUND

[0002] With the development of the integration of vehicle-mounted communication, AR / VR and meta-universe technology, a vehicle-mounted meta-universe (Vehicular Metaverse) has become an important application form in an intelligent transportation system. A vehicle-mounted twin (Vehicular Twins) is a digital mapping entity deployed on a UAV and supports services such as AR navigation and virtual interaction. However, due to the high mobility of vehicles and the limited communication coverage of UAVs, the vehicle-mounted twin needs to be migrated between a source UAV and a destination UAV to ensure the continuity and real-time performance of the service.

[0003] The prior art has ignored the interaction between the time delay uncertainty in the data migration process and the resource pricing mechanism, resulting in low resource allocation efficiency, high service interruption probability and difficulty in meeting the demand for low delay and high reliability.

[0004] Therefore, the application provides a resource transaction method for vehicle-mounted twin migration based on the martingale theory and game modeling, which is used to improve the migration efficiency and service continuity of the vehicle-mounted twin in the vehicle-mounted meta-universe. SUMMARY

[0005] The application provides a resource transaction method for vehicle-mounted twin migration based on the martingale theory and Stackelberg game, which realizes the linkage of the time delay prediction in the migration process and the optimization of the resource scheduling strategy. The method considers the real-time performance and economy of the migration through queuing delay modeling and game theory resource pricing, and improves the overall efficiency of the system.

[0006] The technical scheme of the application is as follows:

[0007] A resource transaction method for vehicle-mounted twin migration based on the martingale theory and Stackelberg game, characterized by: for the bandwidth resource game problem between a vehicle-mounted meta-universe user and a meta-universe service provider in a vehicle-mounted twin migration scenario, a Stackelberg game model based on the martingale theory is constructed. By solving the game equilibrium point, the optimal resource pricing and migration strategy are obtained, the linkage of the time delay prediction in the migration process and the optimization of the resource scheduling strategy is realized, and the migration efficiency and service continuity of the vehicle-mounted twin in the vehicle-mounted meta-universe are improved. The specific steps are as follows:

[0008] 1.1) In the vehicle-mounted meta-universe system, there are J vehicle-mounted meta-universe users, and the jth user (j e {1, 2,..., J}) is deployed on a source UAV U soOnboard vehicle twins can access metaverse services, such as AR navigation and virtual interaction. This is due to the high-speed movement of the vehicle and the source drone U... so The communication coverage is limited (coverage distance is L). To ensure service continuity, the vehicle-mounted twin needs to be transferred from the source UAV. so Real-time migration to the destination UAV de UAV source so To the destination U drone de The distance is d (L < d), and this invention mainly studies the vehicle twin migration process within this distance.

[0009] 1.2) Due to the high real-time requirements of vehicle twin migration, the arrival and processing of data are characterized by suddenness and time-varying nature, leading to dynamic and random changes in queuing delays, which become a key factor affecting the migration success rate. This invention employs martingale theory, introducing a delay violation probability index, i.e., the probability of failing to complete the migration within a given time window. The lower this probability, the higher the reliability of the migration process, thus accurately predicting queuing delays. Based on this, a Stackelberg game model is constructed to achieve optimized linkage between delay prediction and resource scheduling strategies. Below, we use martingale theory to solve for the delay violation probability of vehicle twin migration.

[0010] 1.3) Due to the dynamic and random characteristics of data during the vehicle-mounted twin migration process (including changes in data volume, time-varying UAV location and data processing capabilities, and channel state fluctuations), traditional models struggle to accurately characterize its statistical properties. This invention will use the source UAV... so The data arrival process at k is modeled as a Markov switching process using the Markov Monte Carlo method. This process is then simplified to have two states: 0 and 1. In state 0, no data arrives; in state 1, the amount of data at time k is a. j (k). The data state transition probability during the vehicle twin migration process for the j-th user is defined as: the probability of transitioning from state 0 to state 1 is... The probability of transitioning from state 1 to state 0 is Its state transition probability matrix can be expressed as:

[0011]

[0012] The cumulative data received during the vehicle twin migration process of the j-th user within time [m,n] is:

[0013]

[0014] Furthermore, the martingale model for the data arrival process is defined as follows:

[0015]

[0016] Where θj The decay exponent, Let n be the amount of data arriving at time n. for The right eigenvector, A j (n) represents the amount of data that arrives within the time interval [0, n]. in denoted as spectral radius.

[0017] 1.4)U so The received data is processed and then sent to the target drone. de Service capacity is dynamically changing due to various factors such as channel status and bandwidth allocation. Similar to the data arrival process, the data service process is modeled as a Markov switching process using the Markov Monte Carlo method. In state 0, the link is not established; in state 1, the amount of service data at time k is s. j (k). The probability of transitioning from state 0 to state 1 is Otherwise, the probability is Its state transition probability matrix can be expressed as:

[0018]

[0019] The cumulative amount of service data completed by this user within time [m,n] is:

[0020]

[0021] U de The received power is:

[0022]

[0023] in For the transmission power, γ j The channel gain is represented by l, and the channel fading parameter is l. This represents the probability of a successful transmission.

[0024] Source U drone so and target UAV de The transmission rate between them is expressed as:

[0025]

[0026] in Where N is the received power of the path unit, N0 is the average noise power, and b j The bandwidth resources purchased by the j-th vehicle-mounted metaverse user from the service provider, who is the manager of the drone and responsible for providing bandwidth resources during the vehicle-mounted twin migration process.

[0027] The data service rate of the j-th user at time k can be further expressed as:

[0028]

[0029] It can be seen that the service rate s j (k) and bandwidth b j Proportional.

[0030] Therefore, the service process martingale model is constructed as follows:

[0031]

[0032] Where θ j The decay exponent, The amount of data served at time n. for The right eigenvector, S j (n) represents the amount of data served within time n. in Indicates the spectral radius.

[0033] To ensure the model is non-trivial, the arrival and service rates must satisfy the system's stability conditions, namely:

[0034]

[0035] 1.5) Based on the data arrival and service processes of vehicle twin migration, the data departure process is defined as:

[0036] D j (n)≥inf{A j (m)+S j (k,n)}

[0037] The time delay process can be further represented as follows:

[0038]

[0039] Therefore, the probability of delay violation is defined as:

[0040]

[0041] 1.6) Combining the arrival and service martingale processes, martingale theory can be used to derive the probability of delay violation satisfying:

[0042]

[0043] in for The right eigenvector, This represents the amount of data arriving at time 0.

[0044] To meet the real-time requirements of vehicle twin migration, the queuing delay for the vehicle twin migration of the j-th user is specified as k. j The upper bound of the latency violation probability of the vehicle twin migration of the j-th user can be expressed as:

[0045]

[0046] in Expressing expectations,

[0047] Further based on bandwidth b j With service rate s j (k) and The linear relationship can be established as follows:

[0048]

[0049] Where η j With φ j These are the positive coefficients obtained from linear regression fitting or simulation methods.

[0050] Therefore, the upper bound for the latency violation probability of the j-th user's in-vehicle twin migration is:

[0051]

[0052] 1.7) During the vehicle twin migration process, vehicle users need to purchase bandwidth resources from the Metaverse service provider to reduce the probability of migration latency violations. Since the service provider controls all bandwidth resources, forming a monopolistic market structure, it has pricing power. Users need to decide the purchase quantity based on its pricing. When bandwidth pricing is low, users tend to purchase more bandwidth resources to improve the reliability of vehicle twin migration; conversely, when the price is too high, users' willingness to purchase will be suppressed, leading to an increased probability of latency violations and a decrease in migration reliability.

[0053] The QoS performance of users in the vehicular metaverse depends not only on migration reliability (i.e., latency violation probability) but also on the bandwidth cost paid. The QoS metric for the j-th user is defined as follows:

[0054]

[0055] Where β j Here, p is the balance factor, p is the price per unit bandwidth, and b is the price per unit bandwidth. j Purchase bandwidth for the j-th user. The lower the QoS metric, the higher the user's benefit.

[0056] The unit bandwidth transmission cost of the metaverse service provider is C, and the provider's utility function is:

[0057]

[0058] The larger the utility function, the higher the bandwidth revenue for the provider.

[0059] Furthermore, a Stackelberg game model based on martingale theory is constructed to solve for the Stackelberg game equilibrium point, thereby obtaining the optimal resource pricing and vehicle twin migration strategy to improve migration efficiency and service continuity. The specific steps are as follows:

[0060] 2.1) To optimize bandwidth pricing strategies and resource trading efficiency, this invention introduces a Stackelberg game model, modeled as follows:

[0061] Leader: The metaverse service provider, who selects the optimal pricing p;

[0062] Follower: In-vehicle users who select bandwidth purchase amount b based on pricing. j .

[0063] 2.2) Solving for the follower strategy:

[0064] The j-th in-vehicle user determines the optimal bandwidth by minimizing QoS metrics:

[0065]

[0066] stb min ≤b j ≤b max

[0067] C≤p≤p max

[0068] Where b min The minimum amount of bandwidth purchased for users of the in-vehicle metaverse, b max The maximum bandwidth purchased for users of the in-vehicle metaverse, p max Pricing is set for the highest unit of bandwidth.

[0069] For b j By finding the first and second derivatives, we can obtain:

[0070]

[0071] The optimal bandwidth can be obtained by solving the first-order optimality condition:

[0072]

[0073] 2.3) Leader Strategy Solution:

[0074] The service provider maximizes total revenue based on the user response function:

[0075]

[0076] C≤p≤p max

[0077] Among them B max The maximum amount of bandwidth sold to metaverse service providers.

[0078] The best response of the follower Substitute the utility function of the service provider We can obtain:

[0079]

[0080] The leader's problem is transformed into:

[0081]

[0082] Taking the first and second derivatives with respect to p, we get:

[0083]

[0084] Based on the first-order optimality condition, the optimal pricing p can be obtained. * .

[0085] 2.4) Based on the above model, the iterative algorithm for solving the Stackelberg equilibrium is as follows:

[0086] Step 1: Initialize unit bandwidth pricing p (0) Define the maximum number of iterations. Set the convergence threshold ε and the iteration counter k = 0;

[0087] Step 2: Iterative execution: a) Fix the current unit bandwidth pricing p (k) Calculate the optimal bandwidth for all users. b) Fixed bandwidth selection for all users Substitute the bandwidth response and update the service provider's optimal unit bandwidth pricing. c) If |p (k+1) -p (k) |<∈| or k=K max If the value is -1, stop the iteration; otherwise, return to step a.

[0088] Step 3: Output the Stackelberg equilibrium solution

[0089] The design concept of this invention is as follows:

[0090] This invention addresses the real-time migration process of a vehicle-mounted twin between a source and a destination UAV. It constructs a queuing delay prediction model based on martingale theory, deriving the latency violation probability of the migration process by analyzing the arrival and service processes, and establishing a quantitative relationship between migration reliability and allocated bandwidth. Furthermore, considering bandwidth trading behavior between the metaverse service provider and the vehicle user, a Stackelberg game model is constructed. The service provider, as the leader, sets the optimal bandwidth pricing, while the vehicle user, as the follower, selects the amount of bandwidth to purchase to minimize its quality indicators. Finally, the game equilibrium is iteratively solved using backward induction to obtain the optimal pricing and resource allocation strategy, effectively improving the real-time performance and overall system efficiency of the twin migration. This model can be widely applied to edge computing resource scheduling and communication optimization scenarios in highly dynamic vehicle-mounted metaverse environments.

[0091] The beneficial effects of this invention are as follows:

[0092] This invention introduces martingale theory to accurately predict queuing delays and introduces a delay violation probability index to construct a Stackelberg game model based on martingale theory. This effectively solves the problem of low resource allocation efficiency caused by delay uncertainty in existing technologies. By solving the game equilibrium point, the optimal resource pricing and migration strategy are obtained, realizing the optimal linkage between delay prediction and resource trading strategy in the migration process. This meets the requirements of low latency and high reliability, and improves the migration efficiency and service continuity of vehicle twins in the vehicle metaverse. Attached Figure Description

[0093] Figure 1 This is a model diagram of bandwidth resource trading in a vehicle twin migration scenario according to an embodiment of the present invention;

[0094] Figure 2 This is a flowchart of a bandwidth resource trading method for vehicle twin migration according to an embodiment of the present invention;

[0095] Figure 3 This is a flowchart of an algorithm for solving game equilibrium according to an embodiment of the present invention;

[0096] Figure 4 This is a comparison chart of the utility of the method with other solutions under different unit bandwidth transmission costs in this embodiment of the invention;

[0097] Figure 5 This chart compares the average QoS of the user under different unit bandwidth transmission costs in this invention with other solutions. Detailed Implementation

[0098] The technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings in the embodiments of the present invention. This embodiment provides a resource trading method for vehicle twin migration in a vehicle metaverse, the model diagram and flowchart of which are shown below. Figure 1 and Figure 2 As shown, the specific steps include the following:

[0099] A resource trading method for in-vehicle twin migration based on martingale theory and Stackelberg game theory is proposed. Addressing the bandwidth resource game problem between in-vehicle twin users and metaverse service providers in in-vehicle twin migration scenarios, a Stackelberg game model based on martingale theory is constructed. By solving for the game equilibrium point, the optimal resource pricing and migration strategy are obtained, achieving optimal linkage between latency prediction and resource scheduling strategies during the migration process, thereby improving the migration efficiency and service continuity of in-vehicle twins in the in-vehicle metaverse. The specific steps are as follows:

[0100] 1.1) The system has a total of J vehicle-mounted metaverse users. The j-th user (j∈{1,2,...,J}) continuously updates the deployment on the source UAV. so Onboard vehicle twins can access metaverse services, such as AR navigation and virtual interaction. This is due to the high-speed movement of the vehicle and the source drone U... so The communication coverage is limited (coverage distance is L). To ensure service continuity, the vehicle-mounted twin needs to be transferred from the source UAV. so Real-time migration to the destination UAV de UAV source so To the destination U drone de The distance is d (L < d), and this invention mainly studies the vehicle twin migration process within this distance.

[0101] 1.2) Due to the high real-time requirements of vehicle twin migration, the arrival and processing of data are characterized by suddenness and time-varying nature, leading to dynamic and random changes in queuing delays, which become a key factor affecting the migration success rate. This invention employs martingale theory, introducing a delay violation probability index, i.e., the probability of failing to complete the migration within a given time window. The lower this probability, the higher the reliability of the migration process, thus accurately predicting queuing delays. Based on this, a Stackelberg game model is constructed to achieve optimized linkage between delay prediction and resource scheduling strategies. Below, we use martingale theory to solve for the delay violation probability of vehicle twin migration.

[0102] 1.3) Due to the dynamic and random characteristics of data during the vehicle-mounted twin migration process (including changes in data volume, time-varying UAV location and data processing capabilities, and channel state fluctuations), traditional models struggle to accurately characterize its statistical properties. This invention will use the source UAV... soThe data arrival process at k is modeled as a Markov switching process using the Markov Monte Carlo method. This process is then simplified to have two states: 0 and 1. In state 0, no data arrives; in state 1, the amount of data at time k is a. j (k). The data state transition probability during the vehicle twin migration process for the j-th user is defined as: the probability of transitioning from state 0 to state 1 is... The probability of transitioning from state 1 to state 0 is Its state transition probability matrix can be expressed as:

[0103]

[0104] The cumulative data received during the vehicle twin migration process of the j-th user within time [m,n] is:

[0105]

[0106] Furthermore, the martingale model for the data arrival process is defined as follows:

[0107]

[0108] Where θ j The decay exponent, Let n be the amount of data arriving at time n. for The right eigenvector, A j (n) represents the amount of data that arrives within the time interval [0, n]. in denoted as spectral radius.

[0109] 1.4)U so The received data is processed and then sent to the target drone. de Service capacity is dynamically changing due to various factors such as channel status and bandwidth allocation. Similar to the data arrival process, the data service process is modeled as a Markov switching process using the Markov Monte Carlo method. In state 0, the link is not established; in state 1, the amount of service data at time k is s. j (k). The probability of transitioning from state 0 to state 1 is Otherwise, the probability is Its state transition probability matrix can be expressed as:

[0110]

[0111] The cumulative amount of service data completed by this user within time [m,n] is:

[0112]

[0113] U deThe received power is:

[0114]

[0115] in For the transmission power, γ j The channel gain is represented by l, and the channel fading parameter is l. This represents the probability of a successful transmission.

[0116] Source U drone so and target UAV de The transmission rate between them is expressed as:

[0117]

[0118] in Where N is the received power of the path unit, N0 is the average noise power, and b j The bandwidth resources purchased by the j-th vehicle-mounted metaverse user from the service provider, who is the manager of the drone and responsible for providing bandwidth resources during the vehicle-mounted twin migration process.

[0119] The data service rate of the j-th user at time k can be further expressed as:

[0120]

[0121] It can be seen that the service rate s j (k) and bandwidth b j Proportional.

[0122] Therefore, the service process martingale model is constructed as follows:

[0123]

[0124] Where θ j The decay exponent, The amount of data served at time n. for The right eigenvector, S j (n) represents the amount of data served within time n. in Indicates the spectral radius.

[0125] To ensure the model is non-trivial, the arrival and service rates must satisfy the system's stability conditions, namely:

[0126]

[0127] 1.5) Based on the data arrival and service processes of vehicle twin migration, the data departure process is defined as:

[0128] Dj (n)≥inf{A j (m)+S j (k,n)}

[0129] The time delay process can be further represented as follows:

[0130]

[0131] Therefore, the probability of delay violation is defined as:

[0132]

[0133] 1.6) Combining the arrival and service martingale processes, martingale theory can be used to derive the probability of delay violation satisfying:

[0134]

[0135] in for The right eigenvector, This represents the amount of data arriving at time 0.

[0136] To meet the real-time requirements of vehicle twin migration, the queuing delay for the vehicle twin migration of the j-th user is specified as k. j The upper bound of the latency violation probability of the vehicle twin migration of the j-th user can be expressed as:

[0137]

[0138] in Expressing expectations,

[0139] Further based on bandwidth b j With service rate s j (k) and The linear relationship can be established as follows:

[0140]

[0141] Where η j With φ j These are the positive coefficients obtained from linear regression fitting or simulation methods.

[0142] Therefore, the upper bound for the latency violation probability of the j-th user's in-vehicle twin migration is:

[0143]

[0144] 1.7) During the vehicle twin migration process, vehicle users need to purchase bandwidth resources from the Metaverse service provider to reduce the probability of migration latency violations. Since the service provider controls all bandwidth resources, forming a monopolistic market structure, it has pricing power. Users need to decide the purchase quantity based on its pricing. When bandwidth pricing is low, users tend to purchase more bandwidth resources to improve the reliability of vehicle twin migration; conversely, when the price is too high, users' willingness to purchase will be suppressed, leading to an increased probability of latency violations and a decrease in migration reliability.

[0145] The QoS performance of users in the vehicular metaverse depends not only on migration reliability (i.e., latency violation probability) but also on the bandwidth cost paid. The QoS metric for the j-th user is defined as follows:

[0146]

[0147] Where β j Here, p is the balance factor, p is the price per unit bandwidth, and b is the price per unit bandwidth. j Purchase bandwidth for the j-th user. The lower the QoS metric, the higher the user's benefit.

[0148] The unit bandwidth transmission cost of the metaverse service provider is C, and the provider's utility function is:

[0149]

[0150] The larger the utility function, the higher the bandwidth revenue for the provider.

[0151] A Stackelberg game model based on martingale theory was constructed to solve for the Stackelberg game equilibrium point, thereby obtaining the optimal resource pricing and vehicle twin migration strategy to improve migration efficiency and service continuity. The specific steps are as follows:

[0152] 2.1) To optimize bandwidth pricing strategies and resource trading efficiency, this invention introduces a Stackelberg game model, modeled as follows:

[0153] Leader: The metaverse service provider, who selects the optimal pricing p;

[0154] Follower: In-vehicle users who select bandwidth purchase amount b based on pricing. j .

[0155] 2.2) Solving for the follower strategy:

[0156] The j-th in-vehicle user determines the optimal bandwidth by minimizing QoS metrics:

[0157]

[0158] stb min ≤b j ≤b max

[0159] C≤p≤p max

[0160] Where b min The minimum amount of bandwidth purchased for users of the in-vehicle metaverse, b max The maximum bandwidth purchased for users of the in-vehicle metaverse, p max Pricing is set for the highest unit of bandwidth.

[0161] For b j By finding the first and second derivatives, we can obtain:

[0162]

[0163] The optimal bandwidth can be obtained by solving the first-order optimality condition:

[0164]

[0165] 2.3) Leader Strategy Solution:

[0166] The service provider maximizes total revenue based on the user response function:

[0167]

[0168] C≤p≤p max

[0169] Among them B max The maximum amount of bandwidth sold to metaverse service providers.

[0170] The best response of the follower Substitute the utility function of the service provider We can obtain:

[0171]

[0172] The leader's problem is transformed into:

[0173]

[0174] Taking the first and second derivatives with respect to p, we get:

[0175]

[0176] Based on the first-order optimality condition, the optimal pricing p can be obtained. * .

[0177] 2.4) As Figure 3As shown, based on the above model, the iterative algorithm flow for solving the Stackelberg equilibrium is as follows:

[0178] Step 1: Initialize unit bandwidth pricing p (0) Define the maximum number of iterations. Set the convergence threshold ε and the iteration counter k = 0;

[0179] Step 2: Iterative execution: a) Fix the current unit bandwidth pricing p (k) Calculate the optimal bandwidth for all users. b) Fixed bandwidth selection for all users Substitute the bandwidth response and update the service provider's optimal unit bandwidth pricing. c) If |p (k+1) -p (k) |<∈| or k=K max If the value is -1, stop the iteration; otherwise, return to step a.

[0180] Step 3: Output the Stackelberg equilibrium solution

[0181] 2.5) The results of the service provider utility and average user QOS achieved using the resource trading method of this embodiment are as follows: Figure 4 and Figure 5 As shown. In this invention, the number of users J is set to 2, the unit bandwidth transmission cost is set to 2, 4, 6, 8, 10, the path loss exponent l is 2, and the channel gain γ is... j It follows an exponential distribution with parameter 1, and its transmit power is... It is 30dBm. The maximum bandwidth sold by the service provider is B. max It is 100MHz.

[0182] This invention employs backward induction to iteratively solve the Stackelberg game equilibrium, obtaining the optimal resource pricing and migration strategy. From Figure 4 and Figure 5 As can be seen, compared with random and greedy strategies, the method proposed in this invention performs better in improving the utility of service providers and can significantly reduce the average QoS index value of users. This indicates that while ensuring migration reliability, it achieves high efficiency and fairness in resource allocation, and improves the migration efficiency and service continuity of the vehicle twin in the vehicle metaverse.

[0183] The above embodiments are merely preferred embodiments of the present invention and are not intended to limit the technical solutions of the present invention. Any technical solution that can be implemented based on the above embodiments without creative effort should be considered to fall within the scope of protection of the patent of the present invention.

Claims

1. A resource transaction method for vehicle-mounted twin migration in a vehicle-mounted meta-universe, characterized in that, In the context of vehicle-mounted twin migration scenarios, the bandwidth resource game between vehicle-mounted meta-universe users and meta-universe service providers is addressed. The queuing delay is accurately predicted using the martingale theory method, and the delay violation probability index is introduced to construct the revenue indicators of both meta-universe service providers and users. A Stackelberg game model based on the martingale theory is constructed, and the optimal resource pricing and vehicle-mounted twin migration strategy are obtained by solving the game equilibrium point. This realizes the optimal linkage of delay prediction and resource scheduling strategy in the migration process, and improves the migration efficiency and service continuity of vehicle-mounted twins in the vehicle-mounted meta-universe.

2. The resource transaction method for vehicle-mounted twin migration in a vehicle-mounted meta-universe according to claim 1, characterized in that, The queuing delay is accurately predicted using the martingale theory method, and the delay violation probability index is introduced to construct the revenue indicators of both meta-universe service providers and users. The specific steps are as follows: 1.1) There are J vehicle meta-universe users in the vehicle meta-universe system in total, and the jth user (j ∈ {1, 2,..., J}) accesses the meta-universe service by constantly updating the vehicle twin deployed on the source unmanned aerial vehicle U so ; the vehicle twin is migrated from the source unmanned aerial vehicle U so to the target unmanned aerial vehicle U de in real time, and the distance from the source unmanned aerial vehicle U so to the target unmanned aerial vehicle U de is d (L < d), and L is the communication coverage distance of the source unmanned aerial vehicle U so ; 1.2) The martingale theory method is used to introduce the delay violation probability index, i.e. the probability of not completing migration within a given time window. The lower the delay violation probability, the higher the reliability of the migration process, and thus the queuing delay is accurately predicted. On this basis, a Stackelberg game model is constructed to realize the optimal linkage of delay prediction and resource scheduling strategy; 1.3) The data arrival process at the source unmanned aerial vehicle U so The data arrival process at the source unmanned aerial vehicle U j (k); then the data state transition probability in the jth user's vehicle-mounted twin migration process is defined as: the probability of transition from state 0 to state 1 is The probability of transition from state 1 to state 0 is The state transition probability matrix is represented as: The cumulative arrival data of the jth user's vehicle-mounted twin migration process within time [m, n] is: The martingale model of the data arrival process is defined as follows: where θ j is an exponential decay, is the amount of data arriving at time n, is the right eigenvector of A j (n) denotes the amount of data arriving in time [0, n], where is the spectral radius; 1.4) Source drone U so The received data is processed and sent to the destination drone U de The data service process is modeled as a Markov switching process using the Markov Monte Carlo method, with state 0 when the link is not established, and state 1 when the service data volume at time k is s j (k); the probability of transitioning from state 0 to state 1 is Otherwise, the probability is The state transition probability matrix is represented as: The cumulative service data volume completed by the user within time [m, n] is: U de The received power is: wherein P is the transmit power, γ j denotes the channel gain, l is a channel fading parameter, is the probability of successful transmission. The transmission rate between the source drone U so and the destination drone U de is represented as: wherein, is the received power of the path unit, N0 is the average noise power, b j is the bandwidth resource purchased by the jth on-board user of the metaverse from the service provider, and the metaverse service provider is the manager of the UAV, responsible for providing bandwidth resources in the on-board twin migration process; The data service rate of the jth user at time k is further expressed as: It is known that the service rate s j (k) is proportional to the bandwidth b j . The service process martingale model is constructed as follows: where θ j is an exponential decay, is the amount of data served at time n, is the right eigenvector of S j (n) denotes the amount of data served at time n, where denotes the spectral radius; To make the model non-trivial, the arrival and service rates must satisfy the stability condition of the system, i.e. 1.5) According to the data arrival process and service process of vehicle-mounted twin migration, the data departure process is defined as: D j (n) ≥ inf {A j (m) + S j (k, n)} The expression of the delay process is obtained as follows: Thus, the delay violation probability is defined as: 1.6) Combining the arrival and service martingale processes, the martingale theory is used to derive that the delay violation probability satisfies: wherein is the right eigenvector of is the amount of data arrived at time 0; In order to meet the real-time requirement of the vehicle twin migration, the queuing delay of the vehicle twin migration of the jth user is regulated as k j The upper bound of the delay violation probability of the vehicle twin migration of the jth user is obtained as wherein indicates that, According to the bandwidth b j With the service rate s j (k) and The following relationship is established: where η j with φ j is a positive coefficient obtained from a linear regression fit or simulation method; Therefore, the upper bound of the delay violation probability of the jth user's vehicle-mounted twin migration is: 1.7) In the vehicle-mounted twin migration process, the vehicle-mounted user needs to purchase bandwidth resources from the meta-universe service provider to reduce the migration delay violation probability; The QoS performance of the vehicle-mounted meta-universe user depends not only on the migration reliability, i.e. the delay violation probability, but also on the payment of bandwidth fees; the QoS index of the jth user is defined as: where β j is the balance coefficient, p is the unit bandwidth price, b j is the bandwidth purchased by the jth user, and the smaller the QoS index, the higher the user's revenue. The unit bandwidth transmission cost of the meta-universe service provider is C, and the provider's utility function is: The larger the utility function, the higher the bandwidth revenue of the provider.

3. The resource transaction method for vehicle-mounted twin migration in a vehicle-mounted meta-universe according to claim 1, characterized in that: A Stackelberg game model based on the martingale theory is constructed, and the optimal resource pricing and vehicle-mounted twin migration strategy are obtained by solving the Stackelberg game equilibrium point. This realizes the optimal linkage of delay prediction and resource scheduling strategy in the migration process, and improves the migration efficiency and service continuity of vehicle-mounted twins in the vehicle-mounted meta-universe. The specific steps are as follows: 2.1) To optimize the bandwidth pricing strategy and resource trading efficiency, a Stackelberg game model is introduced, which is modeled as follows: Leader: meta-universe service provider, choose the optimal pricing p; Follower: in-vehicle user, buys bandwidth amount b according to pricing j ; The follower strategy is solved as: The jth on-board user determines the optimal bandwidth by minimizing the QoS indicator: s.t.b min ≤b j ≤b max C < p < p max where b min is the minimum bandwidth amount purchased by the in-vehicle metaverse user, b max is the maximum bandwidth amount purchased by the in-vehicle metaverse user, p max is the highest unit bandwidth pricing; For b j The first and second derivatives are: The optimal bandwidth is obtained by solving the first-order optimality condition: Leader strategy solution: The service provider maximizes the total revenue according to the user response function: C < p < p max where B max the maximum amount of bandwidth sold by the metaverse service provider; Optimal response of follower Substitute the utility function of the service provider Result: The leader problem is transformed into: The first and second derivatives of p are obtained: According to the first order optimality condition, the optimal pricing p is obtained * ; 2.2) Based on the above Stackelberg game model, the iterative algorithm flow of Stackelberg equilibrium solution is as follows: 2.2.1) : initialization of the unit-bandwidth pricing p (0) , define the maximum number of iterations set the convergence threshold ε, iteration counter k = 0; 2.2.2) : Iterative execution: a) fix the current unit bandwidth price p (k) , compute the optimal bandwidth for all users b) fix the bandwidth selection for all users , update the service provider's optimal unit bandwidth price by substituting the bandwidth response c) if |p (k+1) -p (k) | <∈ or k = K max -1, stop iteration; otherwise, go back to step a; 2.2.3): Output Stackelberg equilibrium solution

Images