Real-time compute offload recovery
By building a virtual computing resource pool in the edge computing scenario of the Internet of Vehicles, and using fuzzy logic and deep reinforcement learning algorithms for load balancing and task resource allocation, the problems of limited computing resources and low task resource allocation efficiency are solved, and task processing delay is minimized and efficiency is improved.
CN119697188APending Publication Date: 2025-03-25CHONGQING UNIV OF POSTS & TELECOMM
0 Cites 0 Cited by
Patent Information
- Application Number
- CN202411839360.0
- Authority / Receiving Office
- CN · China
- Patent Type
- Applications(China)
- Current Assignee / Owner
- Filing Date
- 2024-12-13
- Publication Date
- 2025-03-25
AI Technical Summary
Technical Problem
In the Internet of Vehicles edge computing scenario, the computing resources of a single edge server are limited, which leads to too long task execution time, and how to efficiently allocate resources to multiple tasks is also a problem.
Method used
By predicting the vehicle's running trajectory, the edge servers along the trajectory are integrated into a virtual computing resource pool, and the edge servers in the virtual computing resource pool are load balancing and task resource allocation using interval two-type fuzzy logic and deep deterministic strategy gradient reinforcement learning algorithms.
Benefits of technology
It minimizes task processing delay, improves task execution efficiency, and solves the problem of load balancing of edge servers.
✦ Generated by Eureka AI based on patent content.
Abstract
The invention provides a calculation unloading method considering load balancing, and in an edge scene of the Internet of Vehicles, in order to solve the problem of task real-time constraint caused by insufficient local calculation resources in calculation unloading, a virtual calculation resource pool is proposed to integrate calculation resources of an edge server. The method comprises the following steps: firstly, predicting a running track of a vehicle by using a long-short-term memory neural network, and abstractly integrating edge servers around a future track of the vehicle into a virtual computing resource pool; and secondly, evaluating the performance of an edge server in the virtual computing resource pool by utilizing interval type-2 fuzzy logic, selecting the edge server according to an evaluation result, and then determining resource allocation and scheduling of the edge server to tasks by utilizing a depth deterministic strategy gradient reinforcement learning algorithm. Finally, a simulation result and analysis show that the method can effectively improve the service quality of the user and reduce the total time delay of task processing.
Need to check novelty before this filing date? Find Prior Art