Cloud Computing Resource Scheduling Algorithm Based on Unsampled Collaborative Knowledge Graph Network

Haichuan Sun

doi:10.1109/ACCESS.2024.3472212

2024-01-01

DOI：10.1109/ACCESS.2024.3472212

原文链接

摘要

En 中文

A cloud computing resource scheduling algorithm based on sampled collaborative knowledge graph network is designed to address the issues of lag in the process of cloud computing resource scheduling, high overall load rate, and large transient amplitude and phase errors. Based on graph convolutional neural networks, analyze the target load of cloud platforms, construct multi hop data transmission paths one by one, and perform deep level information load balancing; Establish a multiplexing information transmission model, correct the initial weights of graph convolutional neural networks, combine reverse transmission calculation methods, integrate and balance cloud computing resources, and confirm the optimal resource scheduling plan; Integrating class convolution and human-machine interaction attention mechanism, the value of the previous time series neural unit is transferred to the current neural unit, and the classification output sequence of knowledge graph relational data feature fragments is analyzed. The knowledge graph data fragments are processed based on class convolution and human-machine interaction attention mechanism, and different sizes of linear aggregators are used to capture deep level information, completing the design of cloud computing resource scheduling algorithm. The experimental results show that although the load rate is on the rise, the highest is only 89%, and the scheduling rate is relatively high, ranging from 38.9 to 43.1bps; The energy consumption is relatively low, not exceeding 40.106mW. In terms of transient amplitude and phase, the proposed method can control the error within 2.0. Ensure the efficiency and practical application effectiveness of cloud computing resource scheduling algorithms.

Keyword:

Cloud computing

Knowledge graphs

Transient analysis

Resource management

Convolutional neural networks

Scheduling algorithms

Load modeling

Relays

Multiplexing

Graph convolutional networks

Knowledge graph