Task offloading method for improved DBSCAN and many-to-many matching algorithms in 6G vehicle-to-everything (V2X) networks

Abstract

This invention discloses an improved DBSCAN algorithm and a many-to-many matching algorithm for 6G vehicle-to-everything (V2X) networks, belonging to the field of V2X technology. It designs an improved density-based spatial application noise clustering algorithm to cluster more dispersed vehicles, employs a many-to-many matching algorithm to pair vehicles with auxiliary MEC servers, and designs a multi-agent-based task offloading mechanism to reduce latency and improve resource utilization efficiency. The improved DBSCAN algorithm proposed in this invention overcomes the sensitivity of parameter selection, does not require particularly precise parameter settings, performs well on datasets with uneven density, overcomes the influence of vehicle density on clustering results, and the proposed many-to-many matching mechanism improves the utilization of remaining vehicle computing resources and greatly alleviates the pressure on edge servers, while ensuring that high-priority tasks are computed first.

Description

Technical Field

[0001] This invention relates to the field of vehicle networking technology, and in particular to a task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 6G vehicle networking. Background Technology

[0002] With the increasing number of vehicles and the development of intelligent connected vehicles, intelligent transportation systems have become feasible thanks to the revolutionary development of 6G networks. To meet the demands of autonomous driving and in-vehicle entertainment, vehicles have high requirements for the real-time performance and reliability of services. Offloading a large number of tasks to edge servers can lead to high processing latency for many tasks. Furthermore, the high-speed movement of vehicles brings rapidly changing topologies to the vehicle-to-everything (V2X) network. Due to the limited coverage of edge servers, vehicles may drive out of the server's coverage area, causing computation results to be transferred between servers, increasing communication latency.

[0003] While some research has been conducted on vehicle edge computing and task offloading, several challenges and shortcomings remain. Reliable communication and transmission latency between vehicles are significant issues, especially in scenarios with high vehicle mobility or low vehicle density. Better communication mechanisms and transmission schemes are needed to ensure reliable data transmission. Resource constraints on edge devices are a limiting factor, potentially impacting the ability to handle complex tasks and large-scale data processing. More efficient resource management and utilization strategies are required to improve the computing and storage capabilities of edge devices. Utilizing parked vehicles as computing resources may be limited by their availability and coverage. Better vehicle selection and distribution strategies are needed to fully leverage the computing power of parked vehicles. Adaptive task offloading algorithms require better convergence speed and an exploration-utilization balance to find the optimal task offloading strategy. Further research is needed to improve the performance and efficiency of these algorithms. Real-time performance and complexity are challenges in vehicle-to-vehicle communication scenarios. More accurate vehicle trajectory prediction and communication protocol design are needed to meet the requirements of inter-vehicle communication. In summary, although some progress has been made, challenges remain in the field of vehicle edge computing and task offloading. Further research and improvements will contribute to enhancing the performance, reliability, and practicality of the system. Summary of the Invention

[0004] The purpose of this invention is to solve the above-mentioned problems by providing a task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 6G vehicle-to-everything (V2X) networks.

[0005] To achieve the above objectives, the technical solution adopted by this invention is as follows: a task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 6G vehicle-to-everything (V2X) networks. This includes designing an improved density-based spatial application noise clustering algorithm to cluster more dispersed vehicles, employing a many-to-many matching algorithm to pair vehicles with auxiliary MEC servers, and designing a many-to-many matching task offloading mechanism to reduce latency and improve resource utilization efficiency. The preset algorithm is as follows:

[0006] Step 1: First, perform vehicle clustering using the improved DBSCAN algorithm;

[0007] Step 2: Form a platoon based on the clustering results;

[0008] Step 3: Next, within the formation, the proposed many-to-many matching algorithm is used to assist the MEC server in matching with the mission vehicles;

[0009] Step 4: Based on the matching results, perform task unloading and calculation. Vehicles with tasks that do not match the auxiliary MEC server will have their tasks directly unloaded to the edge server.

[0010] Step 5: The server returns the task calculation results;

[0011] Step 6: Repeat steps 1-5 after one calculation cycle is completed.

[0012] Furthermore, it includes a reward step to incentivize vehicles to act as auxiliary MEC servers rather than refusing to provide services.

[0013] Furthermore, the improved DBSCAN algorithm in step 1 is as follows:

[0014] enter:

[0015] D: A dataset containing location information of N vehicle objects

[0016] Radius parameter

[0017] minpts: Density threshold of the neighborhood

[0018] Output:

[0019] A set of density-based clusters

[0020] Step 1: Select point p that has not yet been checked, and center on point p, with... Set the points within the radius to U.

[0021] Step 2: If |U| ≥ minpts, then perform the following steps:

[0022] Step 3: Create a new cluster C and add all points in U to the candidate set N.

[0023] Step 4: Repeat the above steps until U = .

[0024] Step 5: Repeat the above steps until all points p have been examined.

[0025] Step 6: For a point q in the noise points M, perform the following steps:

[0026] Step 7: If d(q, C) < max(d(pi, pj)), where pi, pj ∈ C, then perform the following steps:

[0027] Step 8: Add the point q to the cluster C.

[0028] Furthermore, the many-to-many matching algorithm in Step 3 is as follows:

[0029] Input:

[0030] S: Set of auxiliary MEC servers;

[0031] Ψ: Priority order of the auxiliary MEC server si;

[0032] Tsi: List of tasks to be offloaded to si;

[0033] U: Set of user vehicles;

[0034] Φui: Priority order of the user vehicle ui;

[0035] Tui: List of tasks of the user vehicle ui;

[0036] Output:

[0037] Matching result

[0038] Step 1: For each user u in the user set U, perform the following steps;

[0039] Step 2: For each task t in the task list Tui of the user vehicle u, perform the following steps;

[0040] Step 3: Apply the task t to si in the priority order Pui of the user vehicle ui;

[0041] Step 4: If Tsi is not full, then perform the following steps;

[0042] Step 5: Add the task t to Tsi;

[0043] Step 6: Otherwise, if the priority t.level of t is greater than the priority of any task ti in Tsi, then perform the following steps;

[0044] Step 7: Reject task t and start providing service from the next si in Pui;

[0045] Step 8: Otherwise, proceed with the following steps;

[0046] Step 9: Reject the highest priority task tj in Tsi, where tj.level is the highest in Tsi;

[0047] Step 10: Re-execute step 8 to process the rejected task tj;

[0048] Step 11: If no vehicle is rejected, proceed with the following steps;

[0049] Step 12: Complete the matching process;

[0050] Step 13: Return the matching results.

[0051] Compared with the prior art, the present invention has the following beneficial effects:

[0052] 1. The improved DBSCAN algorithm proposed in this invention overcomes the sensitivity of parameter selection and does not require particularly precise parameter settings.

[0053] 2. The improved DBSCAN algorithm proposed in this invention can perform well on datasets with uneven density, overcoming the influence of vehicle density on clustering results.

[0054] 3. The many-to-many matching mechanism proposed in this invention improves the utilization rate of the vehicle's remaining computing resources and greatly alleviates the pressure on the edge server, while ensuring that high-priority tasks are computed first. Attached Figure Description

[0055] Figure 1 This is a schematic diagram of the overall system according to an embodiment of the present invention. Detailed Implementation

[0056] To make the technical means, creative features, objectives and effects of this invention easier to understand, the invention will be further described below in conjunction with specific embodiments.

[0057] In view of the shortcomings of the current research mentioned in the background art, an embodiment of the present invention provides a task offloading method for vehicles and an auxiliary mobile edge computing (MEC) server based on an improved DBSCAN algorithm and a many-to-many matching algorithm. This method selects vehicles with remaining computing resources as auxiliary MEC servers, uses the improved DBSCAN algorithm to cluster the vehicles, and then matches the vehicles with the auxiliary MEC servers through the many-to-many matching algorithm to achieve task offloading and computing.

[0058] I. Implementation of the improved DBSCAN algorithm

[0059] Input preparation: Obtain the set S of auxiliary MEC servers, the set U of user vehicles, and the corresponding priority orders Ψ and Φui.

[0060] Input:

[0061] D: A dataset containing the location information of N vehicle objects

[0062] : Radius parameter

[0063] minpts: Density threshold of the neighborhood

[0064] Output:

[0065] A set of density-based clustering collections

[0066] Step 1: Select an unexamined point p, and set the points within the range with p as the center and as the radius as U.

[0067] Step 2: If |U| ≥ minpts, perform the following steps:

[0068] Step 3: Establish a new cluster C, and add all points in U to the candidate set N.

[0069] Step 4: Repeat the above steps until U = .

[0070] Step 5: Repeat the above steps until all points p have been examined.

[0071] Step 6: For the point q in the noise point M, perform the following steps: [[ID=4,9]]

[0072] Step 7: If d(q, C) < max(d(pi, pj)), where pi, pj ∈ C, perform the following steps:

[0073] Step 8: Add the point q to the cluster C.

[0074] II. Implementation of the many-to-many matching algorithm

[0075] Input preparation: Obtain the auxiliary MEC server set S, the user vehicle set U, and the corresponding priority order Ψ and Φui.

[0076] For each user u, perform the following steps in the user set U:

[0077] Step 1: For each task t in user vehicle u, perform the following steps in the task list Tui:

[0078] Step 2: Match task t to the corresponding auxiliary MEC server si according to the priority order of user vehicle ui (Pui);

[0079] Step 3: If the task list Tsi of server si is not full, add task t to Tsi;

[0080] Step 4: Otherwise, if the priority t.level of task t is greater than the priority of any task ti in Tsi, perform the following steps:

[0081] i. Reject task t and start providing service from the next si in Pui;

[0082] ii. Otherwise, reject the highest priority task tj in Tsi, where tj.level is the highest in Tsi; re-execute step d to process the rejected task tj;

[0083] Step 5: If no vehicle is rejected, the matching process is complete;

[0084] Step 6: Return the matching results.

[0085] III. Implementation of Task Unloading and Computation

[0086] Step 1: Form a platoon of vehicles based on the clustering results obtained from the improved DBSCAN algorithm;

[0087] Step 2: Within the formation, an auxiliary MEC server is used to match the mission vehicles based on a many-to-many matching algorithm;

[0088] Step 3: Perform task unloading and calculation based on the matching results: Task vehicles that do not match the auxiliary MEC server will have their tasks directly unloaded to the edge server;

[0089] Step 4: The server returns the task calculation results;

[0090] Step 5: Repeat steps 1-4 after one calculation cycle is completed.

[0091] The task offloading method for vehicles and auxiliary edge computing servers proposed in this invention, based on an improved DBSCAN algorithm and a many-to-many matching algorithm, enables task offloading and computation, and achieves more efficient task allocation and computation through clustering and grouping. Furthermore, a reward mechanism is proposed to incentivize vehicles with surplus computing resources to act as auxiliary MEC servers, rather than refusing to provide services.

[0092] It will be apparent to those skilled in the art that the present invention is not limited to the details of the exemplary embodiments described above, and that the invention can be implemented in other specific forms without departing from its spirit or essential characteristics. Therefore, the embodiments should be considered in all respects as exemplary and non-limiting, and the scope of the invention is defined by the appended claims rather than the foregoing description. Thus, all variations falling within the meaning and scope of equivalents of the claims are intended to be included within the present invention. No reference numerals in the claims should be construed as limiting the scope of the claims.

[0093] Furthermore, it should be understood that although this specification describes embodiments, not every embodiment contains only one independent technical solution. This narrative style is merely for clarity. Those skilled in the art should consider the specification as a whole, and the technical solutions in each embodiment can also be appropriately combined to form other embodiments that can be understood by those skilled in the art.

Claims

A task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 1.6G vehicle-to-everything (V2X) networks, characterized in that, This includes designing an improved density-based spatial application noise clustering algorithm to cluster more dispersed vehicles, employing a many-to-many matching algorithm to pair vehicles with auxiliary MEC servers, and designing a many-to-many matching task offloading mechanism to reduce latency and improve resource utilization efficiency. The preset algorithm is as follows: Step 1: First, perform vehicle clustering using the improved DBSCAN algorithm. The improved DBSCAN algorithm is as follows: enter: D: A dataset containing location information of N vehicle objects radius parameter minpts: Density threshold of the neighborhood Output: A set of density-based clusters S1: Select point p that has not yet been checked, and center on point p, ... Set U as the radius for points within that range; S2: If Then perform the following steps: S3: Create a new cluster C and add all points from U to the candidate set N; S4: Repeat the above steps until U = ; S5: Repeat the above steps until all points p have been checked; S6: For point q in the noise point M, perform the following steps: S7: If d(q, C) < max(d(pi, pj)), where pi, pj C, then perform the following steps: S8: Add point q to cluster C; Step 2: Form a platoon based on the clustering results; Step 3: Next, within the formation, the proposed many-to-many matching algorithm is used to assist the MEC server in matching with the mission vehicles; Step 4: Based on the matching results, perform task unloading and calculation. Vehicles with tasks that do not match the auxiliary MEC server will have their tasks directly unloaded to the edge server. Step 5: The server returns the task calculation results; Step 6: Repeat steps 1-5 after one calculation cycle is completed.

2. The task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 6G vehicle-to-everything (V2X) as described in claim 1, characterized in that, It also includes a reward step to incentivize vehicles with surplus computing resources to act as auxiliary MEC servers instead of refusing to provide services.

3. The task offloading method for the improved DBSCAN algorithm and many-to-many matching algorithm in 6G vehicle-to-everything (V2X) as described in claim 1, characterized in that, The many-to-many matching algorithm in step 3 is as follows: enter: S: Auxiliary MEC server set; Priority order of auxiliary MEC server si; Tsi: A list of tasks to be uninstalled to si; U: User vehicle set; UI: Priority order of user vehicle UI; Tui: The user's vehicle UI task list; Output: Matching results Step 1: For each user u, perform the following steps in the user set U; Step 2: For each task t in user vehicle u, perform the following steps in task list Tui; Step 3: Apply task t to si in the priority order Pui of the user vehicle ui; Step 4: If Tsi is not full, proceed with the following steps; Step 5: Add task t to Tsi; Step 6: Otherwise, if the priority t.level of t is greater than the priority of any task ti in Tsi, then perform the following steps; Step 7: Reject task t and start providing service from the next si in Pui; Step 8: Otherwise, proceed with the following steps; Step 9: Reject the highest priority task tj in Tsi, where tj.level is the highest in Tsi; Step 10: Re-execute step 8 to process the rejected task tj; Step 11: If no vehicle is rejected, proceed with the following steps; Step 12: Complete the matching process; Step 13: Return the matching results.

Images