Topological link sensing task collaborative unloading method for V2V and V2I combined system

Abstract

The invention discloses a topological link sensing task collaborative unloading method for a V2V and V2I combined system. A delay minimization problem of sequential subtasks in a V2I and V2V combined system model is expressed as a path optimization problem and modeled as a Markov decision process, an action selected by an optimal unloading scheme under a time window is obtained, a task execution process is further obtained, the task execution process is divided into the time window, and the time window of the task execution process is obtained. Calculating the task calculation amount in each time window under the current pulse parameter, and then carrying out Gaussian fitting on the task calculation amount to obtain a Gaussian curve; and fitting Gaussian curve parameters, and predicting the time window task calculation amount under other pulse parameters. By combining the TLCO and the STW, the method achieves the efficient prediction of the calculation workload of the V2V and V2I systems on different time slices, can provide an optimization basis for an unloading decision, and reduces the calculation delay and communication overhead.

Description

Technical Field

[0001] The present invention relates to the application of reinforcement learning in task offloading in the field of communications, and specifically to a topology link perception task collaborative offloading method for a V2V and V2I joint system. Background Art

[0002] With the continuous advancement of intelligent transportation systems, a series of innovative technologies have emerged in the field of Internet of Vehicles (IoV), among which vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communication technologies are particularly prominent. These technologies enable real-time information exchange through direct communication between vehicles and vehicles, and between vehicles and road infrastructure, providing strong support for applications such as intelligent traffic management, safe driving, and environmental perception.

[0003] In complex traffic environments, such as urban congestion areas or emergency rescue scenarios, vehicles need to respond quickly and work together. To meet these challenges, researchers have proposed a cluster-based computing strategy that organizes vehicles into clusters, with cluster head vehicles centrally processing computing tasks, thereby improving overall computing efficiency and quality of service (QoS). In addition, V2V and V2I communication technologies play an important role in collision warning systems at unsignalized intersections, effectively reducing the risk of traffic accidents through accurate vehicle location and speed information.

[0004] However, the high-speed mobility of vehicles leads to the instability of communication links, which poses a challenge to the stability and reliability of vehicle networks. To address this problem, researchers have proposed a variety of strategies, such as dynamic resource allocation based on link duration and distance, and computational task offloading that considers link reliability, computational resources, and relative speed. These strategies aim to optimize the transmission efficiency in V2V and V2I communications and ensure that key information can be delivered in a timely and accurate manner.

[0005] Privacy protection is also an issue that cannot be ignored in the development of the Internet of Vehicles. When vehicles are conducting V2V communication, sensitive information may be exposed, which raises concerns about personal privacy. In order to balance communication efficiency and privacy protection, researchers are exploring how to build a secure and reliable communication link topology while ensuring communication quality. These structures need to take into account multiple factors, such as link stability, data transmission security, and vehicle mobility.

[0006] In practical applications, it is not feasible to simply rely on statistical information to allocate computing resources because the demand changes dynamically. For this reason, reinforcement learning has become an effective tool for optimizing the resource allocation of Internet of Vehicles. Through reinforcement learning, vehicles can adjust their communication and computing strategies based on real-time environmental feedback to achieve more flexible and efficient resource management. Despite this, existing research often ignores the importance of the topological structure between vehicles, which is crucial in actual V2V communication. Summary of the invention

[0007] In order to achieve efficient resource allocation and task offloading based on various topological link structures, a topological link-aware task co-offloading method for joint V2V and V2I system, STW-TLCO (Topological Link-Aware Task Co-Offloading Method for Joint V2V and V2I System,) is proposed. In order to find the best offloading strategy under different topological link structures, the topological link-aware partial offloading process in the joint V2V and V2I system is modeled as a Markov decision process (MDP), and a double deep-Q-network (DDQN) is used to solve the optimization problem. Then, an innovative prediction framework (STW-TLCO) based on TLCO and sliding time window module is proposed to predict the computational workload of the joint V2V and V2I system at different time slices. This prediction is achieved by analyzing the pulse parameters, which play a crucial role in determining the arrival time of tasks within the system.

[0008] In a first aspect, the present invention provides a method for collaborative offloading of topology link awareness tasks for a V2V and V2I joint system, which is implemented by the following steps:

[0009] In a second aspect, the present invention provides a topology link awareness task collaborative offloading system for implementing the method, comprising:

[0010] Data acquisition module, responsible for acquiring Internet of Vehicles data;

[0011] The data processing module is responsible for obtaining the actions for selecting the optimal unloading solution in each time window under the V2I and V2V joint system model;

[0012] The time window task computing amount prediction module is responsible for obtaining the task execution process according to the action selected by the optimal unloading plan under the time window, dividing the task execution process into time windows, and calculating the task computing amount in each time window; performing Gaussian fitting on the time window task computing amount under the current pulse parameters, and predicting the time window task computing amount under other pulse parameters.

[0013] In a third aspect, the present invention provides an electronic device, comprising a processor and a memory, wherein the memory stores machine executable instructions that can be executed by the processor, and the processor executes the machine executable instructions to implement the method.

[0014] In a fourth aspect, the present invention provides a machine-readable storage medium storing machine-executable instructions. When the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the method.

[0015] The present invention has the beneficial effects:

[0016] The present invention adopts the TLCO algorithm to provide accurate scheduling execution time for vehicle tasks by considering task arrival time and computing power, ensuring that tasks are completed efficiently within a limited time window. The STW framework dynamically adjusts the time window for task execution through the sliding time window technology, adapts to the uncertainty of task arrival and changes in system load, thereby optimizing resource allocation, and uses Gaussian curves to fit the computational amount of each time window. The present invention can more accurately predict the workload of the system. Combining TLCO and STW, the present invention realizes efficient prediction of the computational workload of V2V and V2I systems in different time slices, can provide an optimization basis for offloading decisions, and reduce computational delays and communication overhead. Therefore, the present invention develops a fine-grained partial offloading model for sequential subtasks in dynamic V2I and V2V systems with topological links to optimize collaborative offloading strategies, thereby achieving efficient utilization of computing resources. By accurately predicting and optimizing offloading strategies, the waste of computing resources caused by improper task scheduling is reduced, and the overall energy efficiency ratio of the system is improved. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] Figure 1 It is a schematic diagram of the method proposed in the present invention.

[0018] Figure 2 This is the overall architecture of the V2V and V2I joint system proposed in the present invention.

[0019] Figure 3 These are all possible structural modes of topological links when the number of vehicles M = 4.

[0020] Figure 4 is the average delay of different numbers of subtasks, where the number of subtasks K = [5, 6, 7].

[0021] Figure 5 is the average delay for different numbers of vehicles, where the number of vehicles M = [5, 6, 7].

[0022] Figure 6is the average delay of the assisting vehicle at different speeds, where speed v0 = [16, 18, 20] m / s.

[0023] Figure 7 It compares the effects of learning rate, seed, and speed on the reward of the training process.

[0024] Figure 8 It is the amount of computation in different time windows when δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1].

[0025] Fig. 9 It is the cumulative amount of calculation in different time windows when δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1].

[0026] Fig.10 It is the difference between the predicted value and the actual value of the computational amount of the time window task under different δ, where δ in Figures (a)-(c) is 1 / 6, 1 / 8, and 1 / 10 respectively. DETAILED DESCRIPTION

[0027] The present invention is further described below in conjunction with the accompanying drawings and specific embodiments.

[0028] Vehicle-to-Vehicle (V2V) and vehicle-to-Infrastructure (V2I) joint offloading is an effective way to utilize the remaining computing resources of neighboring devices, thereby expanding the coverage of the computing resource supply of the Internet of Vehicles. However, many studies have overlooked the importance of topological communication caused by the rapid movement of vehicles, privacy, and communication intentions. In order to achieve efficient task offloading in the face of various topological link structures, this paper proposes a novel topological link-aware task collaborative offloading method designed for partial offloading in V2V and V2I joint systems. The sequential subtask offloading process is modeled as a Markov decision process, and the deep double Q-network (DDQN) algorithm is used to optimize the total delay of the proposed system. Based on the innovative prediction framework of TLCO and sliding time window module (STW-TLCO), the pulse parameters are used to accurately predict the computational load of different time windows. Compared with the state-of-the-art methods, the proposed TLCO algorithm is more effective and superior. Furthermore, the STW-TLCO algorithm also demonstrated high accuracy with an R-squared value of over 99%, confirming its predictive ability.

[0029] This embodiment provides a topology link awareness task collaborative offloading method for a V2V and V2I joint system. Figure 1 The following steps are involved:

[0030] Step 1: Obtain Internet of Vehicles data, which includes the data transmission rate R of the vehicle and the intelligent roadside equipment RSU, the task computing capacity f, and the connection mode Mode of the topological link (such as Figure 3 ), as well as the vehicle's mission data volume D, mission computation workload ω, coordinates (x, y), speed v, and mission initial time T start ;

[0031] Step 2: Establish a V2I and V2V joint system model, and formulate the delay minimization problem of sequential subtasks in the V2I and V2V joint system model as a path optimization problem;

[0032] The V2V and V2I joint system model is as follows Figure 2 At least a number of ordinary vehicles V1, V2, ..., V M , an auxiliary vehicle V0 and an intelligent roadside equipment RSU; ordinary vehicles selectively communicate with other ordinary vehicles to form a V2V link, auxiliary vehicles can communicate with any vehicle to form a V2A link, and intelligent roadside equipment RSU can communicate with any vehicle to form a V2I link; wherein, the auxiliary vehicle is responsible for generating tasks, and has the ability to execute its own designated tasks and unload tasks from other networked ordinary vehicles; the ordinary vehicle is responsible for generating tasks, and has the ability to execute its own designated tasks and unload tasks from other networked ordinary vehicles; the intelligent roadside equipment has the ability to execute tasks from the auxiliary vehicle itself and unload tasks from ordinary vehicles;

[0033] The specific task offloading relationship is: m The task is divided into K subtasks, where m∈{0,1,2,…,M} and the subtasks related to the same vehicle are arranged in order, and the subsequent subtasks are generated according to the previous subtasks; the initial task arrival time assigned to each ordinary vehicle is random, Indicates vehicle V m The arrival time of a task is equal to the arrival time of its initial task;

[0034] For each vehicle V m , whose subtask set is represented as Task m ={task m,1 ,task m,2 ,…,task m,K Specifically, vehicle V m The first subtask task m,1 It can be executed within the vehicle itself or offloaded to intelligent roadside equipment, auxiliary vehicle V0 or other ordinary communication vehicles; the last subtask task m,K After successful completion, the results need to be transmitted back to the vehicle V m ;

[0035] The delay minimization problem of sequential subtasks consists of three parts: i) execution time, which depends on the amount of computation and the computing power of the device; ii) transmission time, which is calculated based on the data size and transmission rate (if there is a link); iii) waiting time, which is determined by the offloading strategy of the previous subtask and the limitation of available computing resources;

[0036] Assume that Dev0 corresponds to auxiliary vehicle V0, Dev M+1 Corresponding to intelligent roadside equipment, Dev1, Dev2, ..., Dev M Corresponding to ordinary vehicles V1, V2, ..., V M , so it is recorded as device Dev l (where l∈{0,1,…,M+1}; device Dev l All have the ability to perform the subtasks assigned to them;

[0037] 1) Execution time:

[0038] Subtask m,k The execution time is expressed as:

[0039]

[0040] in For ordinary vehicles V m The kth subtask task m,k ,Device l Execution time; exe represents the execution process; ω m,k Represents a subtask task m,k The computational workload of f, m∈{0,1,2,…,M},k∈{1,2,…,K}; l Indicates the device Dev l The computing power of , in cycles per second;

[0041] Used to determine the task task m,K Is it on the device Dev l The calculation formula is as follows:

[0042]

[0043] 2) Transmission time:

[0044] Intelligent roadside equipment RSU and auxiliary vehicles have the ability to communicate with all ordinary vehicles; however, due to privacy, trust authorization and communication intent, the links between ordinary vehicles are restricted; these restrictions are manifested as Mask m =[mask m,1 ,maskm,2 ,…,mask m,M ].

[0045]

[0046] where m'∈{1,2,…,M}, m≠m'; task task m,k Not always from the vehicle V m Transmission, unless k = 1; introduction of CD m,k Indicates the current task task m,k equipment, namely:

[0047] When k = 1, the vehicle V m Execute subtasks task by itself m,k ;

[0048] When k>1, execute subtask task m,k The device is recorded as CD m,k ; Therefore, the subtask task m,k Transmission time The calculation formula is:

[0049]

[0050] Where cur represents CD m,k ,D m,k Represents a subtask task m,k The data size, tr represents the transmission process; It indicates a subtask task m,k Whether to transfer to the target device Dev l The indication value of mask cur,l Indicates from CD m,k To Dev l An indication value of whether a link exists;

[0051] Current Device CD m,k To deviceDev l The transmission rate R cur,l The calculation formula is:

[0052]

[0053] Where h is the channel fading coefficient, P is the transmission power, θ is the path loss index, and ω0 is the white Gaussian noise power; B cur,l and d cur,l Represents the current device CD m,k With the deviceDev l Bandwidth and distance between

[0054] 3) Waiting time:

[0055] When a vehicle or intelligent roadside equipment is m,k When the device is occupied, other subtasks that are offloaded to the occupied device need to wait for the subtask task m,k Finish;

[0056] When the subtask m,k Select Device l When executing, wait time The calculation is as follows:

[0057]

[0058]

[0059] in, Device l The occupation time, K m is the vehicle V m The number of completed subtasks, It is a task m,k Arrival DeviceDev l The arrival time, ΔT m,k,l Is on the device Dev l Tasks executed on m,k Total delay of Indicates vehicle V m The time when subtasks are generated, K m is the vehicle V m The number of completed subtasks;

[0060] The delay minimization problem for a dynamic V2I and V2V joint system is formulated as a path optimization problem, taking into account the possibility of offloading tasks according to their sequential subtask dependencies; the problem is formulated as follows:

[0061]

[0062] Where T sum Represents the total latency of executing all subtasks on all devices;

[0063] Step 3: Based on step 2, the path optimization problem is modeled as a Markov decision process;

[0064] 3-1 Define the state space and the state of each time step; specifically:

[0065] Time step t p The state is defined as:

[0066]

[0067]

[0068] Among them, is the result of the last subtask that needs to be transmitted back to vehicle V m where MO ∈ Mode M is the mode of the topological link; is m the current position of V at the p-th time step, and is m the speed of V;;

[0069] 3-2 Define the action space. The action selection at each time step is given by the reinforcement learning method; the reinforcement learning method uses the deep double Q-network DDQN;

[0070] The action space is A = [a0, a1, …, a M+1 , where a(t p ) = a0 means that at time t p subtask is executed on the auxiliary vehicle, a(t p ) = [a1, …, a M means that at time t p subtask is executed on a certain ordinary vehicle, a(t p ) = a M+1 means that at time t p subtask is executed on the intelligent roadside unit RSU;

[0071] 3-3 Define the reward. Different rewards and punishments are defined according to different situations. The core is to minimize the total delay;

[0072]

[0073] Among them C1, C2, C3 are three different constants used to adjust the punishment and reward according to the preferences of the optimization problem;

[0074] The definition of condition 1 is as follows: If a(t p ) is not in , and the task number k of time t p = K, then the task p result is preferentially transmitted back to the vehicle Condition 2 is defined as follows: If a(t ) is not in p , and k p < K, then the task p result is preferentially transmitted to the connected device in ; ;

[0075] 3-4 The deep dual Q network DDQN learns and selects the action that obtains the highest reward according to the state at each moment to minimize the total delay of the system; the deep dual Q network DDQN includes a main network and a target network;

[0076] The target network is based on the current action a p , output Q T ;

[0077] The Q function of the main network is updated as:

[0078]

[0079] where r p It is through t p The reward measured by the time delay in h p+1 and a p+1 are the hidden representation and action at the next moment, Q M (t p+1 ) is the main network at t p+1 Q value prediction at the moment, Q T is the output of the target network; is time t p The Q function of the main network; Represents the updated Q function of the main network, that is, the Q function is obtained after updating

[0080] By being able to obtain the maximum of Output a p+1 That is, at time t p+1 The action selected by the optimal unloading solution;

[0081] Step 4: According to the action selected by the optimal unloading solution under the time window in step 3, obtain the task execution process, and then divide the task execution process into time windows. Each window is of the same size and fixed, and the task calculation amount in each time window is calculated; Gaussian fitting is performed on the task calculation amount of the time window under the current pulse parameters to obtain a Gaussian curve; then the Gaussian curve parameters are fitted to predict the Gaussian curve parameters under other pulse parameters, and the task calculation amount of the time window under other pulse parameters is predicted by reconstructing the Gaussian curve.

[0082] 4-1 The actions obtained in step 3 are used to execute all tasks, divide the task execution process into time windows, and calculate the task calculation amount in each time window; specifically:

[0083] The task computation amount of each time window is the sum of the task computation amounts of multiple subtasks in the time window. The task computation amount of each subtask is:

[0084]

[0085] in Represents task m,k exist Execute on, The computing power, t wd Indicates the start time of the time window, t u Indicates the size of the time window; represents the vehicle V under pulse parameter δ m task m,k Arrival time, lie in within the scope; The time slots in the impulse workflow are randomly arrived at by tasks from multiple vehicles within a duration multiple of δ; a k ∈[a0,a M+1 ] represents the action of selecting the optimal unloading solution obtained in step 3; C is a constant; Indicates a k Execute subtask task m,k Total delay of

[0086] Therefore, the task computation amount of the time window is:

[0087]

[0088] 4-2 According to the task calculation amount of each time window under the pulse parameter δ, the following Gaussian curve is used for fitting:

[0089]

[0090] Where y = Com(t wd ,δ), x represents the time window, b, c, d are Gaussian curve parameters.

[0091] Then use the power function to fit b, c, and d to predict the corresponding Gaussian curve parameters under other pulse parameters δ'. According to the predicted Gaussian curve parameters under other pulse parameters, a Gaussian curve is constructed to obtain the task calculation amount of each time window under the pulse parameter δ'.

[0092] Figure 3 This is an example of all possible structural modes of topological links when the number of vehicles M = 4. There are 10 connection modes between the 4 devices, forming different Masks.

[0093] Figure 4 is the average delay of different numbers of subtasks, where the number of subtasks K = [5, 6, 7]. Figure 4As shown, the performance of TLCO-DDQN (i.e., the method without step 4 of the present invention) is significantly better than other methods. When M=5 and K=5, TLCO-DDQN is 5.19% higher than TLCO-DQN (i.e., DDQN in TLCO-DDQN is replaced by deep Q network DQN), 3.23% higher than TLCO-DuelingDQN, 22.57% higher than Greedy, 43.34% higher than Greedy-Noseg (i.e., Greedy algorithm that merges subtasks into a single task), 58.72% higher than All Edge, 136.12% higher than All Local, and 186.82% higher than Random. As K increases, the average execution time increases due to the increase in the total computational workload of the subtasks. At the same time, the total computing power of the device remains unchanged, resulting in a narrower optimization space. Therefore, the advantages of the DRL-based algorithm gradually weaken. Among the non-DRL based algorithms, Greedy achieves the best offloading decision, while the three DRL based algorithms, TLCO-DDQN, TLCO-DQN and TLCO-DuelingDQN, consistently outperform the non-DRL based algorithms with different K values.

[0094] Figure 5 is the average delay for different numbers of vehicles, where the number of vehicles M = [5,6,7]. Figure 5 As shown in the figure, the results of TLCO-DDQN are significantly better than other methods. For example, when the experiment is set to M=6 and K=5, TLCO-DDQN is significantly improved by 7.57% over TLCO-DQN, 1.52% over TLCO-DuelingDQN, 11.44% over Greedy, 30.35% over Greedy-Noseg, 66.47% over All Edge, 96.91% over All Local, and 162.31% over Random. As the value of M increases, the average execution time also increases. Compared with the increase in K value, the results based on the DRL algorithm show a greater efficiency improvement. This is because as M increases, the available equipment of the vehicle also increases, thereby expanding the optimization space of the average time. Among the non-DRL-based algorithms, the greedy algorithm achieves the best offloading decision, while the three DRL-based algorithms TLCO-DDQN, TLCODQN, and TLCO-DuelingDQN consistently outperform the non-DRL-based algorithms at different values ​​of M. As the number of available devices increases, the optimization space also expands, and the superiority of DRL-based methods becomes apparent, indicating that they can effectively handle various offloading situations in V2I and V2V systems.

[0095] Figure 6is the average delay of the assisting vehicle at different speeds, where speed v0 = [16, 18, 20] m / s. In order to study the effect of vehicle speed on the experimental results, we set the vehicle speed to 16, 18 and 20 m / s for comparison. Figure 6 As shown, the change in the auxiliary vehicle speed has little effect on the experimental results. This observation shows that the proposed method can robustly determine the optimal unloading strategy regardless of the change in vehicle speed.

[0096] Figure 7 This is to compare the effects of learning rate, seed and speed on the reward of the training process. Considering that different random seeds will lead to randomness in the experimental results, we set the seeds of TLCO-DDQN to 0, 500 and 1000 respectively to reduce the impact of randomness on the experimental results. In addition, we also compared the effects of different learning rates and auxiliary vehicle speeds on the experimental results. The learning rates were set to 0.005, 0.007 and 0.01, and the speeds were set to 16, 18 and 20 m / s. Figure 7 As shown, the different colors of the range represent the average rewards of different parameter sets of TLCO-DDQN. The smaller the range, the smaller the impact of different parameters. Different speeds and seeds have almost no effect on the convergence speed and results, while different learning rates only speed up the convergence speed. The convergence results are consistent with the cases of different speeds and seeds.

[0097] Figure 8 is the amount of computation in different time windows when δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1]. We set the pulse parameter δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1], which corresponds to the degree of task congestion. As δ increases, the arrival time of tasks will be delayed, resulting in a decrease in task density. The size of the time window is set to 0.1 seconds, and the interval of the time window is set to 0.02 seconds. Figure 8 As shown in the figure, as δ decreases, the peak of the computation arrives earlier and reaches a higher level, resulting in early completion of the task. For example, when δ = 1, the computation amount is 0 near the 100th sliding time window; when δ = 1 / 7, the computation amount is 0 near the 80th sliding time window, and the peak when δ = 1 / 7 is higher and earlier. When the tasks are sparse, the peak of the computation amount will appear after a period of time. When the tasks are dense, the peak of the computation amount appears earlier.

[0098] Fig. 9 is the cumulative amount of computation in different time windows when δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1]. We set the pulse parameter δ∈[1 / 9,1 / 7,1 / 5,1 / 3,1]. As δ increases, the task completion time will be advanced, that is, the cumulative amount of computation will reach the peak earlier.

[0099] Fig.10is the difference between the predicted and actual values ​​of the computational effort for different δ. We predict the parameters of the Gaussian curve corresponding to the computational effort within the sliding window for different values ​​of δ∈[1 / 8, 1 / 6, 1 / 4, 1 / 2]. These predicted values ​​are compared with the actual values ​​calculated by TLCO. The fitting results for different values ​​of δ are shown in Fig.10 .

Claims

1. A collaborative offloading method for topology link awareness tasks in a V2V and V2I joint system, characterized in that The following steps are involved: Step 1: Obtaining Internet of Vehicles data, which includes the data transmission rate, task computing capability and connection mode of topological links of vehicles and intelligent roadside equipment RSU, as well as the task data volume, task computing workload, coordinate position, speed and task initial time of the vehicle; Step 2: Establish a V2I and V2V joint system model, and formulate the delay minimization problem of sequential subtasks in the V2I and V2V joint system model as a path optimization problem; Step 3: Model the path optimization problem as a Markov decision process to obtain the action of selecting the optimal unloading solution under the time window; Step 4: According to the action selected by the optimal unloading solution under the time window in step 3, obtain the task execution process, and then divide the task execution process into time windows. Each window is of the same size and fixed, and the task calculation amount in each time window is calculated; Gaussian fitting is performed on the task calculation amount of the time window under the current pulse parameters to obtain a Gaussian curve; then the Gaussian curve parameters are fitted to predict the Gaussian curve parameters under other pulse parameters, and the task calculation amount of the time window under other pulse parameters is predicted by reconstructing the Gaussian curve.

2. The method according to claim 1, characterized in that The V2V and V2I joint system model consists of at least a plurality of ordinary vehicles V1, V2, ..., V M , an auxiliary vehicle V0 and an intelligent roadside device RSU; ordinary vehicles selectively communicate with other ordinary vehicles to form a V2V link, auxiliary vehicles can communicate with any vehicle to form a V2A link, and intelligent roadside devices RSU can communicate with any vehicle to form a V2I link; wherein, the auxiliary vehicle is responsible for generating tasks, and has the ability to execute its own designated tasks and unload tasks from other networked ordinary vehicles; the ordinary vehicle is responsible for generating tasks, and has the ability to execute its own designated tasks and unload tasks from other networked ordinary vehicles; the intelligent roadside device has the ability to execute tasks from the auxiliary vehicle itself and unload tasks from ordinary vehicles.

3. The method according to claim 1, characterized in that The delay minimization problem of sequential subtasks in the V2V and V2I joint system model includes execution time, transmission time, and waiting time; The delay minimization problem of the dynamic V2I and V2V joint system is formulated as a path optimization problem, which is formulated as follows: in They represent the V m The kth subtask task m,k ,Device l Execution time, transmission time, waiting time, T sum Represents the total latency of executing all subtasks on all devices.

4. The method according to claim 1, characterized in that Step 3 specifically is: 3-1 Define the state space and the state of each time step; 3-2 Define the action space, and the action selection at each time step is given by the reinforcement learning method; the reinforcement learning method uses the deep dual Q network DDQN; 3-3 Define rewards. Define different rewards and penalties according to different situations to minimize the total delay. 3-4 The deep dual Q network DDQN learns and selects the action that obtains the highest reward according to the state at each moment to minimize the total delay of the system; the deep dual Q network DDQN includes a main network and a target network; The target network is based on the current action a p , output Q T ; The Q function of the main network is updated as: where r p It is through t p The reward measured by the time delay in h p+1 and a p+1 are the hidden representation and action at the next moment, Q M (t p+1 ) is the main network at t p+1 Q value prediction at the moment, Q T is the output of the target network; is time t p The Q function of the main network; represents the updated Q function of the main network; By being able to obtain the maximum of Output a p+1 That is, at time t p+1 The action selected by the optimal unloading solution.

5. The method according to claim 1, characterized in that Step 4 is as follows: 4-1 The actions obtained in step 3 are used to execute all tasks, divide the task execution process into time windows, and calculate the task calculation amount in each time window; 4-2 According to the task calculation amount of each time window under the current pulse parameter δ, the Gaussian curve is fitted; Then, the power function is used to fit the Gaussian curve parameters to predict the corresponding Gaussian curve parameters under other pulse parameters δ'. According to the predicted corresponding Gaussian curve parameters under other pulse parameters δ', a Gaussian curve is constructed to obtain the task calculation amount of each time window under the pulse parameter δ'.

6. The method according to claim 5, characterized in that Step 4-1 is as follows: The task computation amount of each time window is the sum of the task computation amounts of multiple subtasks in the time window. The task computation amount of each subtask is: in Represents task m,k exist Execute on, The computing power of wd Indicates the start time of the time window, t u Indicates the size of the time window; represents the vehicle V under pulse parameter δ m task m,k The arrival time, a k ∈[a0,a M+1 ] represents the action of selecting the optimal unloading solution obtained in step 3, C is a constant; Indicates a k Execute subtask task m,k Total delay of Therefore, the task computation amount in the time window 7. The method according to claim 5, characterized in that The Gaussian curve under the current pulse parameter δ in step 4-2 is specifically: Where y = Com(t wd ,δ), x represents the time window, b, c, d are the Gaussian curve parameters.

8. A topology link awareness task collaborative offloading system implementing the method described in any one of claims 1 to 7, characterized in that include: Data acquisition module, responsible for acquiring Internet of Vehicles data; The data processing module is responsible for obtaining the actions for selecting the optimal unloading solution in each time window under the V2I and V2V joint system model; The time window task computing amount prediction module is responsible for obtaining the task execution process according to the action selected by the optimal unloading plan under the time window, dividing the task execution process into time windows, and calculating the task computing amount in each time window; performing Gaussian fitting on the time window task computing amount under the current pulse parameters, and predicting the time window task computing amount under other pulse parameters.

9. An electronic device, characterized in that: The invention comprises a processor and a memory, wherein the memory stores machine executable instructions that can be executed by the processor, and the processor executes the machine executable instructions to implement the method according to any one of claims 1 to 7.

10. A machine-readable storage medium, characterized in that: The machine-readable storage medium stores machine-executable instructions. When the machine-executable instructions are called and executed by a processor, the machine-executable instructions prompt the processor to implement the method described in any one of claims 1 to 7.

Images