Research On Beidou Short-message Optimization For Meteorological And Oceanographic Data Transmission

Efficient transmission of meteorological and oceanographic data is the key to realizing real-time data acquisition.Beidou short-message has become an important means of transmitting meteorological and oceanographic data in remote areas,sea,and other areas without signal coverage.In order to overcome the problems of low short-message communication capacity and scarce satellite processing resources.The "Beidou-3" short message satellite communication system and research on Beidou short-message optimization for meteorological and oceanographic data transmission.The purpose is to improve the utilization of satellite processing resources and the reliability of short-message transmission.The main work is as follows:(1)Aiming at the satellite access problem of short-message terminals in the multi-satellite coverage scenario,a short-message terminal access strategy based on Q-learning(QSMAS)is proposed.QSMAS comprehensively considers the parameters of the satellite and the user.QSMAS combines the Q-learning algorithm to adapt to the satellite access selection of different service requests and different types of short-message terminals in a dynamic environment.QSMAS improves the access success rate of short message terminals and communication quality.(2)Aiming at the problems of the low success rate of Beidou short-message communication transmission and unreliable communication link,an interactive short-message data transmission control protocol based on automatic repeat request message cache(ARQ-MC)was proposed.ARQ-MC implements retransmission or buffering of failed packets,which improves the reliability of Beidou short-message meteorological and oceanographic data transmission.(3)Aiming at the scarcity of satellite short-message processing resources,a short-message satellite resource allocation algorithm based on deep reinforcement learning(DRL-SRA)is proposed.DRL-SRA reduces the input data dimension based on the region division strategy and feature extraction network and uses the deep deterministic policy gradient(DDPG)framework to parameterize the continuous space state,reasonably allocates and schedule the short-message satellite processing resources in the multi-satellite coverage area.DRL-SRA improves the resource utilization rate of the satellite system.