Directional antenna neighbor discovery method and device for bidirectional transceiving reinforcement learning, and medium

• Main Authors: SU ZEBIN, WEN JIANFENG, CHEN XINYE, ZHONG LINYUN, GUO JUNYI, WEI ZONGHENG, LIN ZHIYONG, LIU QINGLIN, WEN QINGJI, LIN ZHANHUI, WU HUAKUN, ZOU RUHAO
• Format: Patent
• Language: Chinese; English
• Published: 22.03.2024
• Subjects: ELECTRIC COMMUNICATION TECHNIQUE; ELECTRICITY; WIRELESS COMMUNICATIONS NETWORKS

The invention discloses a directional antenna neighbor discovery method and device for bidirectional transceiving reinforcement learning and a medium, and relates to the technical field of communication, and the method comprises the steps: S1, constructing a wireless ad hoc network model; s2, initializing a node state, a sector selection probability set and a neighbor list of each node in the wireless ad hoc network model; s3, starting learning circulation, randomly selecting the real-time state of the current time slot, and selecting the real-time action of the current time slot according to the sector selection probability set; s4, generating an enhanced signal according to the real-time state, the real-time action and the monitoring result of each sub-time slot in the current time slot; s5, updating a sector selection probability set according to the enhanced signal; s6, updating the neighbor list of each node according to the neighbor nodes found by the current learning cycle; and S7, returning to the ste