Air-based network mixed service flow intelligent scheduling method based on deep reinforcement learning

• Main Authors: JI XIN, WANG JING, PENG YUHUAI, WU SHUQING
• Format: Patent
• Language: Chinese; English
• Published: 27.02.2024
• Subjects: ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; WIRELESS COMMUNICATIONS NETWORKS

The invention provides an air-based network mixed service flow intelligent scheduling method based on deep reinforcement learning. The method comprises the following steps: designing a management and control separated network architecture; the method comprises the following steps of: dividing a data transmission process into time slot fragments with the same size in a time dimension, and describing data in the air-based network; designing a queue scheduling module for the communication base station; a multi-stage queue scheduling mechanism is designed in combination with the transmission thought of time slot fragmentation; based on a deep reinforcement learning method, designing a reliable data scheduling strategy for the time-sensitive priority queue TSQ; designing emergency degree and cache time state elements; designing a reward function, and in a data transmission process, selecting a proper data transmission sequence according to the state of a node where a data packet is currently located; based on a re