CN-117218832-B - Multi-source multi-target fusion mobile edge computing method and system

CN117218832BCN 117218832 BCN117218832 BCN 117218832BCN-117218832-B

Abstract

The invention relates to a multi-source multi-target fusion mobile edge computing method and system, which comprises the steps of obtaining residual energy of each MEC site, setting an opening and closing energy threshold, determining MEC sites participating in computation according to the magnitude relation between the opening and closing energy threshold and the residual energy of each MEC site, grouping all MEC sites participating in computation, after all MEC sites participating in computation receive computation tasks transmitted by surrounding vehicles, distributing tasks for each MEC site according to the energy state of each MEC site and the transmission time among each MEC site, respectively executing the distributed tasks by each MEC site, and returning to the vehicles after the execution is completed. Compared with the prior art, the method has the advantages of low time delay and suitability for various scenes.

Inventors

XIA BIN
YIN CHENGLIANG
GAO FEI
WANG YU
CHEN ZHIYONG
WU YUEPENG
QIN WENGANG

Assignees

上海智能网联汽车技术中心有限公司
上海交通大学

Dates

Publication Date: 20260508
Application Date: 20230831

Claims (7)

1. The multi-source multi-target fusion mobile edge computing method is characterized by comprising the following steps of: Acquiring the residual energy of each MEC site, setting an opening and closing energy threshold, and determining the MEC sites participating in calculation according to the magnitude relation between the opening and closing energy threshold and the residual energy of each MEC site; Grouping all MEC stations participating in calculation, and after all MEC stations participating in calculation receive calculation tasks transmitted by surrounding vehicles, distributing tasks to each MEC station for each group of MEC stations according to the energy state of each MEC station and the transmission time among each MEC station; each MEC station respectively executes the assigned tasks, and the tasks are returned to the vehicle after the execution is completed; The strategy for grouping all MEC sites participating in calculation is specifically as follows: according to the residual electric quantity condition of MEC sites participating in calculation, MEC sites with low residual electric quantity and MEC sites with more residual electric quantity are distributed to the same group; The specific process of determining MEC sites participating in calculation is as follows: Closing MEC sites with residual energy lower than the opening and closing energy threshold, and taking MEC sites with residual energy higher than the opening and closing energy threshold as MEC sites participating in calculation; And setting a plurality of group distributors for distributing calculation tasks to MEC sites of the group, wherein each group of the group distributors is one MEC site selected by each group after the group completion, and is used for receiving and distributing tasks to each MEC site in the group according to the energy state of each MEC site in the group and the transmission time among each MEC site.
2. The method for computing a multi-source multi-target fusion mobile edge according to claim 1, wherein the strategy for grouping all MEC sites participating in the computation further comprises: Grouping is carried out according to the position information of MEC sites participating in calculation through a clustering method.
3. A mobile edge computing system applying the multi-source multi-target fusion mobile edge computing method according to any one of claims 1-2, comprising: the MEC stations are used for receiving the calculation tasks transmitted by the surrounding vehicles and executing the calculation tasks; an MEC management component to participate in determining MEC sites and MEC groupings to participate in the computation; and the plurality of group distributors are used for distributing calculation tasks to MEC sites of the group.
4. The mobile edge computing system of claim 3 wherein the MEC management component receives the remaining energy cases of all MEC sites, sets an on-off energy threshold, determines MEC sites participating in the computation based on a magnitude relationship between the on-off energy threshold and the remaining energy of each MEC site, and groups the MEC sites participating in the computation.
5. The mobile edge computing system of claim 3 wherein the MEC management component sets an opening and closing energy threshold and sends the opening and closing energy threshold to each MEC site, each MEC site determining whether to act as a MEC site participating in the computation based on the magnitude between its remaining energy and the opening and closing energy threshold and performing ad hoc grouping.
6. The mobile edge computing system of any of claims 4 or 5, wherein MEC sites with residual energy below the opening and closing energy threshold are closed and MEC sites with residual energy above the opening and closing energy threshold are considered MEC sites involved in the computation.
7. A mobile edge computing system according to claim 3, wherein the MEC site communicates with the vehicle via an RSU and/or base station.

Description

Multi-source multi-target fusion mobile edge computing method and system Technical Field The invention relates to the technical field of wireless communication of the Internet of vehicles, in particular to a multi-source multi-target fusion mobile edge computing method and system. Background Under the actual road scene, mass data are generated in the running process of the vehicle, and are generally interfered by various real environments, so that the multi-source data with errors and redundancy are preprocessed by utilizing methods such as compressed sensing and the like, and the mass multi-source information with high robustness is fused and processed in real time, and the processing methods bring new requirements for the computing capacity of hardware equipment, and the higher computing capacity of the hardware equipment needs enough electric energy to ensure. On the roads in remote areas and other roads without cables, the energy of the road side MEC (Multi-ACCESS EDGE Computing) equipment is often provided by wind power or solar energy, the energy storage difference of each station is caused by the energy supply mode, and in extreme cases, a certain station can not work normally due to the failure of power generation equipment, so that the Computing tasks are distributed in a coordinated mode according to the energy storage condition of each station, and the dependence on a single station is worth researching. In the prior art, many efforts are made to solve the above technical problems, for example, in "a method for collecting combined energy and unloading task of a mobile edge computing server" disclosed in patent CN108880893a, by modeling the state of charge of the MEC server and the energy consumption required for executing the task of the user, determining an optimal unloading policy, performing centralized task allocation, and thus, minimizing the combined execution overhead of time delay and energy consumption is achieved. For another example, patent CN109413615A discloses "energy delay tradeoff for MEC-based energy aware offloading in internet of vehicles" that considers MEC server energy and task processing delay, and specific task allocation is performed by setting user preferences (low latency bias or low energy consumption). Although the above prior art all make a contribution to coordinating the allocation of computing tasks, the following drawbacks exist: (1) The prior art only considers the condition that the MEC server has sufficient electric quantity, and the discussion of the scene that the remote areas need to be powered by wind power and solar energy is lacking, so that the method has certain limitation; (2) The centralized task allocation adopted in the prior art can cause the problem of too high time delay under the cooperative scene of the expressway and the road. Disclosure of Invention The invention aims to overcome the defects of the prior art and provide a multi-source multi-target fusion mobile edge computing method and a system. The aim of the invention can be achieved by the following technical scheme: A multi-source multi-target fusion mobile edge computing method comprises the following steps: Acquiring the residual energy of each MEC site, setting an opening and closing energy threshold, and determining the MEC sites participating in calculation according to the magnitude relation between the opening and closing energy threshold and the residual energy of each MEC site; Grouping all MEC stations participating in calculation, and after all MEC stations participating in calculation receive calculation tasks transmitted by surrounding vehicles, distributing tasks to each MEC station for each group of MEC stations according to the energy state of each MEC station and the transmission time among each MEC station; each MEC site respectively executes the assigned tasks, and the tasks are transmitted back to the vehicle after the execution is completed. Further, the policy of grouping all MEC sites participating in the calculation is specifically: Grouping is carried out according to the position information of MEC sites participating in calculation through a clustering method. Further, the policy of grouping all MEC sites participating in the calculation is specifically: And according to the residual capacity conditions of MEC sites participating in calculation, the MEC sites with low residual capacity and the MEC sites with more residual capacity are distributed to the same group. Further, the specific process of determining MEC sites involved in the calculation is: and closing MEC sites with the residual energy lower than the opening and closing energy threshold, and taking MEC sites with the residual energy higher than the opening and closing energy threshold as MEC sites participating in calculation. The invention also provides a mobile edge computing system applying the multi-source multi-target fusion mobile edge computing method, which comprises the following steps: the