一种通感一体化网络多任务学习资源分配方法及系统
By constructing an ISAC network architecture that supports federated learning, beamforming and computational offloading are transformed into a multi-objective optimization problem. The MGDA-UB algorithm with multi-gradient descent is used to solve the joint optimization problem of resource allocation in the ISAC network, achieving efficient allocation of sensing, communication and computing resources, alleviating communication pressure and protecting data privacy.
Patent Information
- Authority / Receiving Office
- CN · China
- Patent Type
- Patents(China)
- Current Assignee / Owner
- UNIV OF SCI & TECH BEIJING
- Filing Date
- 2023-05-29
- Publication Date
- 2026-07-17
AI Technical Summary
In the Industrial Internet, there is a lack of joint optimization methods for sensing, communication and computing resources in ISAC networks, especially in the research of edge intelligence-assisted ISAC networks, which lacks deep coupling.
An ISAC network architecture supporting federated learning is constructed. By offloading beamforming and computation, the problem is transformed into a multi-objective optimization problem. The upper bound of the multi-gradient descent method MGDA-UB algorithm is adopted to transform it into a multi-task learning model, thereby achieving joint optimization of resource allocation.
It achieves joint optimization of sensing, communication and computing resources, solves data privacy protection issues, reduces computing costs, and alleviates communication pressure.
Smart Images
Figure CN117155436B_ABST