A game-based vehicle-to-vehicle (V2V) computing offloading system and method

Abstract

The application relates to a vehicle-to-vehicle (V2V) computing offloading system and method based on a game, and belongs to the field of V2V wireless communication. The application is used for computing-intensive and delay-sensitive task offloading between vehicles. A computing-intensive vehicle as a task vehicle proposes a task offloading request, and a vehicle with strong computing capability and sufficient computing resources in the communication range of the task vehicle as a service vehicle. In consideration of cost and profit, the task vehicle offloads the task to the service vehicle for completion and pays service fees, so that the utility of completing the task is maximized; the service vehicle can obtain the service fees by sharing resources in the task offloading process, so that the resources of the service vehicle are effectively utilized and the obtained benefits are optimized. The task offloading interaction process between the two vehicles is converted into a two-stage game, a Stackelberg Game model is used to prove that the game has a Nash equilibrium point, and a Hooke-jeeves algorithm is used to find the Nash equilibrium point.

Description

Technical Field

[0001] This invention relates to a game-theoretic V2V computation offloading system and method for vehicle-to-everything (V2V) networks, belonging to the field of vehicle-to-everything (V2X) wireless communication. Background Technology

[0002] With the rapid development of intelligent vehicles and communication technologies, researchers have proposed the Vehicle Ad-Hoc Network (VANET) to achieve network connectivity and real-time information sharing among vehicles. VANET integrates sensors and Roadside Units (RSUs) for communication with vehicles on the road. Simultaneously, leveraging Mobile Edge Computing (MEC) technology, computational tasks are offloaded to surrounding vehicles or RSUs via vehicle-to-infrastructure (V2I) and vehicle-to-vehicle (V2V) communication methods. This key technology of computation offloading in MEC provides low-latency, low-energy, low-cost, and highly reliable services for computationally intensive emerging businesses and applications in intelligent vehicles, avoiding the latency, energy consumption, and high cost problems caused by data transmission to remote centers. However, considering the high deployment cost and lack of flexibility of edge servers, offloading application tasks to edge servers within the vehicle network does not effectively solve the problem. Summary of the Invention

[0003] The technical problem solved by this invention is to provide a game-theoretic V2V computational offloading system and method for vehicle-to-everything (V2V) networks. It uses Stackelberg game theory to describe the task offloading interaction between vehicles as a two-stage game, and uses the Hooke-jeeves algorithm to find the optimal offloading strategy for both sides of the game, so as to maximize the joint utility of the task vehicle and the service vehicle.

[0004] The technical solution adopted in this invention is as follows: a game-based V2V computation offloading method describes the computation task offloading process between task vehicles and service vehicles. The Stackelberg game theory is used to describe the offloading interaction process between task vehicles and service vehicles as a two-stage game. The Hooke-jeeves algorithm is used to find the optimal computation task offloading strategy for both sides in the game process, so as to maximize the joint utility of task vehicles and service vehicles.

[0005] A game-theoretic V2V computation offloading system for vehicle-to-everything (V2V) communication includes:

[0006] The task vehicle is used to send a computing task offloading request to the roadside unit and receive the initial resource allocation strategy of the service vehicle from the roadside unit. It communicates directly with the service vehicle to play a game. Based on the resource allocation strategy of the service vehicle, it uses the Hooke-jeeves algorithm to adjust the service pricing strategy and finally obtain the optimal service pricing strategy.

[0007] The roadside unit is responsible for responding to unloading requests issued by the task vehicle, broadcasting task unloading request information to service vehicles within the unit's coverage area, and sending the initial resource allocation strategy fed back by the service vehicle to the task vehicle.

[0008] The service vehicle responds to the broadcast from the roadside unit, formulates an initial resource allocation strategy based on its own comprehensive conditions, and feeds the initial resource allocation strategy back to the roadside unit. It then directly engages in a game with the task vehicle, continuously adjusting the resource allocation strategy using the Hooke-Jeeves algorithm, and finally obtains the optimal resource allocation strategy to complete the computation task for the task vehicle.

[0009] Preferably, the mission vehicle is a mobile vehicle node that is computationally intensive and latency-sensitive.

[0010] Preferably, the service vehicle is a mobile vehicle node with powerful computing capabilities and sufficient computing resources.

[0011] Preferably, the roadside unit is a stationary node with storage capacity fixed on both sides of the road, which communicates with all vehicles within the coverage area via V2I to collect basic information of all vehicles within the coverage area.

[0012] Preferably, the roadside unit has a coverage area of ​​ten kilometers.

[0013] Preferably, the comprehensive conditions of the service vehicle itself include the amount of available computing resources, link reliability, relative distance and relative speed with the task vehicle.

[0014] Preferably, the basic information of the vehicle is the vehicle performance index value.

[0015] A game-theoretic V2V computation offloading method for vehicle-to-everything (V2V) networks, with the following specific steps:

[0016] Step 1: The task vehicle issues an unload request;

[0017] Step 2: The roadside unit responds to the unloading request and broadcasts the unloading request information to all service vehicles within the unit's coverage area;

[0018] Step 3: The service vehicle receives information from the roadside unit, provides an initial resource allocation strategy based on its own comprehensive conditions, and feeds back the initial resource allocation strategy to the roadside unit.

[0019] Step 4: The roadside unit sends the initial resource allocation strategy reported by the service vehicle to the task vehicle.

[0020] Step 5: The task vehicle uses the Hooke-jeeves algorithm to provide the corresponding service price based on the resource allocation strategy of the service vehicle and sends the service price directly to the corresponding service vehicle; the service vehicle then uses the Hooke-jeeves algorithm to adjust the resource allocation strategy based on the service price sent by the task vehicle, formulates the updated resource allocation strategy, and sends it directly to the task vehicle.

[0021] Step 6: The task vehicle and the service vehicle repeatedly repeat Step 5 and engage in multiple games until both parties decide not to change the service pricing and resource allocation strategies, so that the computation task offloading strategy reaches the Nash equilibrium point and the optimal computation task offloading strategy is obtained.

[0022] Step 7: Computation task unloading complete.

[0023] The beneficial effects of this invention are: it transforms the computational task offloading process into a two-stage dynamic game using the Stackelberg game algorithm. The task vehicle proposes a service pricing strategy based on the available computing resources of different service vehicles, and the service vehicle updates its resource allocation strategy according to the corresponding service pricing. The task vehicle and the service vehicle engage in a game, using the Hooke-Jeeves algorithm to iterate the service pricing and resource allocation strategies during the game until both sides decide not to change their strategies, reaching a Nash equilibrium point. This yields the optimal service pricing and resource allocation strategies, ensuring that the utility of both sides is optimized. Attached Figure Description

[0024] Figure 1 This is a schematic diagram of the system structure of the present invention;

[0025] Figure 2 This is a schematic diagram of the communication method of the present invention;

[0026] Figure 3 This is an overall block diagram of the present invention;

[0027] Figure 4 This is a schematic diagram of the game process in this invention. Detailed Implementation

[0028] The present invention will be further described below with reference to the accompanying drawings and specific embodiments.

[0029] Example 1: As Figure 1-4 As shown, a game-theoretic V2V computation offloading system is used for offloading computationally intensive and latency-sensitive tasks between vehicles, including:

[0030] The task vehicle is used to send a computing task offloading request to the roadside unit and receive the initial resource allocation strategy of the service vehicle from the roadside unit. It communicates directly with the service vehicle to play a game. Based on the resource allocation strategy of the service vehicle, it uses the Hooke-jeeves algorithm to adjust the service pricing strategy and finally obtain the optimal service pricing strategy.

[0031] The roadside unit is responsible for responding to unloading requests issued by the task vehicle, broadcasting task unloading request information to service vehicles within the unit's coverage area, and sending the initial resource allocation strategy fed back by the service vehicle to the task vehicle.

[0032] The service vehicle responds to the broadcast from the roadside unit, formulates an initial resource allocation strategy based on its own comprehensive conditions, and feeds the initial resource allocation strategy back to the roadside unit. It then directly engages in a game with the task vehicle, continuously adjusting the resource allocation strategy using the Hooke-Jeeves algorithm, and finally obtains the optimal resource allocation strategy to complete the computation task for the task vehicle.

[0033] Furthermore, the mission vehicle is a mobile vehicle node that is computationally intensive and latency-sensitive.

[0034] Furthermore, the service vehicle is a mobile vehicle node with powerful computing capabilities and sufficient computing resources.

[0035] Furthermore, the roadside unit is a stationary node with storage capacity, fixed on both sides of the road. It communicates with all vehicles within its coverage area via V2I to collect basic information about all vehicles within the coverage area. The coverage area of ​​the roadside unit is ten kilometers.

[0036] Furthermore, the comprehensive conditions of the service vehicle itself include the amount of available computing resources, link reliability, relative distance and relative speed with the task vehicle.

[0037] Furthermore, the basic information of the vehicle refers to its performance index values.

[0038] A game-theoretic V2V computation offloading method for vehicle-to-everything (V2V) networks, with the following specific steps:

[0039] Step 1: The task vehicle issues an unload request;

[0040] Step 2: The roadside unit responds to the unloading request and broadcasts the unloading request information to all service vehicles within the unit's coverage area;

[0041] Step 3: The service vehicle receives information from the roadside unit, provides an initial resource allocation strategy based on its own comprehensive conditions, and feeds back the initial resource allocation strategy to the roadside unit.

[0042] Step 4: The roadside unit sends the initial resource allocation strategy reported by the service vehicle to the task vehicle.

[0043] Step 5: The task vehicle uses the Hooke-jeeves algorithm to provide the corresponding service price based on the resource allocation strategy of the service vehicle and sends the service price directly to the corresponding service vehicle; the service vehicle then uses the Hooke-jeeves algorithm to adjust the resource allocation strategy based on the service price sent by the task vehicle, formulates the updated resource allocation strategy, and sends it directly to the task vehicle.

[0044] Step 6: The task vehicle and the service vehicle repeatedly repeat Step 5 and engage in multiple games until both parties decide not to change the service pricing and resource allocation strategies, so that the computation task offloading strategy reaches the Nash equilibrium point and the optimal computation task offloading strategy is obtained.

[0045] Step 7: Once the task unloading is complete, the service vehicle will receive the corresponding service fee paid by the task vehicle.

[0046] like Figure 2 As shown, the system uses two communication methods: V2V communication and V2I communication. Roadside units (RSUs) communicate with vehicles via V2I to collect basic vehicle information (vehicle performance metrics). Vehicles interact with each other via V2V communication to offload computational tasks.

[0047] like Figure 4 The process of task unloading between the task vehicle and the service vehicle is transformed into a two-stage dynamic game. In the first stage, the task vehicle proposes a service pricing strategy based on the initial resource allocation strategies of different service vehicles. The service vehicles then update their own resource allocation strategies accordingly. In the second stage, the task vehicle adjusts its service pricing strategy in response to the updated resource allocation strategies proposed by the service vehicles. This two-stage game continues until both sides decide not to change their strategies, resulting in a Nash equilibrium point for the computational task unloading strategy (service pricing and resource allocation strategies), thus achieving the optimal computational task unloading strategy and maximizing their own benefits.

[0048] Stackelberg game model: The interaction between the task vehicle and the service vehicle is described as a two-stage game. It is proved that there is a Nash equilibrium point in the computational task unloading strategy of both the task vehicle and the service vehicle so that the benefits of both parties are optimized.

[0049] Hooke-jeeves algorithm: In the game between task vehicles and service vehicles, the Hooke-jeeves algorithm is used to find the Nash equilibrium point (the optimal resource allocation strategy and the optimal service price strategy).

[0050] The Stackelberg Game is a two-stage dynamic game. The main idea is that both players choose their own strategies based on the possible strategies of their opponents to maximize their own benefit under those strategies, thus reaching a Nash equilibrium. In the game model, the player who makes the first decision is called the leader. After the leader, the remaining players make decisions based on the leader's decisions and are called followers. The leader then adjusts its own decisions based on the followers' decisions, and this process continues until a Nash equilibrium is reached. In this invention, the task vehicle, as the leader, sets service prices based on the resource allocation strategies of different service vehicles. The service vehicles, as followers, update their own resource allocation strategies based on the corresponding service prices. The task vehicle then adjusts its service prices in response to the updated resource allocation strategies of the service vehicles, until both sides decide not to change their strategies to achieve their optimal benefit. This involves one or more two-stage interactions.

[0051] This invention addresses the offloading of computationally intensive and latency-sensitive tasks between vehicles. A computationally intensive vehicle, acting as the task vehicle, initiates the task offloading request, while a vehicle within its communication range with powerful computing capabilities and sufficient resources acts as the service vehicle. Considering cost and profit, the task vehicle offloads the task to the service vehicle for completion and pays a service fee, maximizing the utility of task completion. The service vehicle, by sharing resources during the task offloading process, earns a service fee, effectively utilizing its own resources and optimizing its revenue. The task offloading interaction between the two types of vehicles is transformed into a two-stage game. The existence of a Nash equilibrium is proven using the Stackelberg Game model, and the Hooke-Jeeves algorithm is used to find the Nash equilibrium (the optimal resource allocation strategy and service pricing strategy).

[0052] The specific embodiments of the present invention have been described in detail above with reference to the accompanying drawings. However, the present invention is not limited to the above embodiments. Within the scope of knowledge possessed by those skilled in the art, various changes can be made without departing from the spirit of the present invention.

Claims

1. A game-theoretic V2V computational offloading system for vehicle-to-everything (V2V) networks, characterized in that: include The task vehicle is used to send a computation task offloading request to the roadside unit and receive the initial resource allocation strategy of the service vehicle from the roadside unit. It communicates directly with the service vehicle to play a game. Based on the resource allocation strategy of the service vehicle, it uses the Hooke-jeeves algorithm to adjust the service pricing strategy and finally obtain the optimal service pricing strategy. The roadside unit is responsible for responding to unloading requests issued by the task vehicle, broadcasting task unloading request information to service vehicles within the unit's coverage area, and sending the initial resource allocation strategy fed back by the service vehicles to the task vehicle. The service vehicle responds to the broadcast from the roadside unit, formulates an initial resource allocation strategy based on its own comprehensive conditions, and feeds the initial resource allocation strategy back to the roadside unit. It then directly engages in a game with the task vehicle, continuously adjusting the resource allocation strategy using the Hooke-Jeeves algorithm, and finally obtains the optimal resource allocation strategy to complete the computation task for the task vehicle.

2. The game-theoretic V2V computational offloading system for vehicle networking according to claim 1, characterized in that: The task vehicle is a mobile vehicle node that is computationally intensive and latency-sensitive.

3. The game-theoretic V2V computational offloading system for vehicle networking according to claim 1, characterized in that: The service vehicle is a mobile vehicle node with powerful computing capabilities and sufficient computing resources.

4. The game-theoretic V2V computational offloading system for vehicle networking according to claim 1, characterized in that: The roadside unit is a stationary node with storage capacity that is fixed on both sides of the road. It communicates with all vehicles within its coverage area via V2I and collects basic information about all vehicles within the coverage area.

5. The game-theoretic V2V computational offloading system for vehicle networking according to claim 1, characterized in that: The roadside unit has a coverage area of ​​ten kilometers.

6. The game-theoretic V2V computational offloading system for vehicle networking according to claim 1, characterized in that: The comprehensive conditions of the service vehicle itself include the amount of available computing resources, link reliability, relative distance and relative speed with the task vehicle.

7. A game-theoretic V2V computational offloading system for vehicle networking according to claim 4, characterized in that: The basic information about the vehicle refers to its performance index values.

8. A game-theoretic V2V computational offloading method for vehicle-to-everything (V2V) networks, characterized in that: The specific steps are as follows: Step 1: The task vehicle issues an unload request; Step 2: The roadside unit responds to the unloading request and broadcasts the unloading request information to all service vehicles within the unit's coverage area; Step 3: The service vehicle receives information from the roadside unit, provides an initial resource allocation strategy based on its own comprehensive conditions, and feeds back the initial resource allocation strategy to the roadside unit. Step 4: The roadside unit sends the initial resource allocation strategy reported by the service vehicle to the task vehicle. Step 5: The task vehicle uses the Hooke-jeeves algorithm to provide the corresponding service price based on the resource allocation strategy of the service vehicle and sends the service price directly to the corresponding service vehicle; the service vehicle then uses the Hooke-jeeves algorithm to adjust the resource allocation strategy based on the service price sent by the task vehicle, formulates the updated resource allocation strategy, and sends it directly to the task vehicle. Step 6: The task vehicle and the service vehicle repeatedly repeat Step 5 and engage in multiple games until both parties decide not to change the service pricing and resource allocation strategies, so that the computation task offloading strategy reaches the Nash equilibrium point and the optimal computation task offloading strategy is obtained. Step 7: Computation task unloading complete.

Images