Research On Reliable Transmission Technology Of Emergency-oriented Mobile Ad Hoc Network

In emergency scenarios,the real-time perception of on-site situation information and the timely and effective communication between rescuers are very important to improve the efficiency of emergency response and reduce the impact of disasters.However,the communication infrastructure in post-disaster areas is often severely damaged,and it is difficult to recover in a short time,which greatly affects the development of rescue activities.Mobile Ad Hoc network(MANET)is a promising choice to recover wireless communication without preset infrastructure.It can be quickly set up in disaster areas at any time without geographical restrictions,and has good scalability and invulnerability.MANET is considered to be an effective solution to the communication vacuum problem in disaster areas.However,the characteristics of MANET,such as dynamic topology,multi-hop transmission,and time-varying link quality make traditional network protocols less adaptable to MANET.With the development of a new generation of communication systems,various access terminals and rich data types have also brought new challenges to the reliable transmission of emergency-oriented MANET.Therefore,this thesis focuses on how to design an adaptive and reliable network protocol for emergency-oriented MANET,and the source of this thesis is also supported by the National Natural Science Foundation of China(Project No.62171060).The main contributions are summarized as follows:(1)A dynamic priority packet scheduling scheme is proposed for emergency-oriented MANET,since the existing scheduling scheme rarely considers the impact of time-varying network state on the data packet transmission process.The proposed scheme can realize the rational allocation of transmission resources,by adopting a dynamic priority method at the sending end,that is,adjusting the scheduling sequence of data packets according to the real-time network state.To incorporate the node mobility,as well as the topology instability and the time-varying channel quality into the packet’s priority assignment,this thesis firstly models the dynamic characteristics of nodes and the various types of transmission channels in emergency communication networks.The transmission delay of the data packet is then estimated,meanwhile the link maintenance probability between mobile nodes is theoretically analyzed.When designing the priority update function,the estimated future packet transmission situation above are taken into account.Finally,a complete dynamic priority packet scheduling scheme is given,and simulation results show that the proposed scheme can effectively improve the successful transmission rate of data packets and reduce the average end-to-end delay.(2)A multi-objective routing algorithm based on pre-trained deep reinforcement learning is proposed,since,in the existing routing algorithms,the routing criteria are relatively single,and the routing indicators are not suitable for emergency networks,which make it difficult to optimize multiple network performance indicators simultaneously in a dynamic network environment.Firstly,according to the time-varying factors of the MANET,routing metrics based on node energy consumption,load and transmission quality are designed respectively.Then the routing decision problem in dynamic environment is modeled as a Markov decision process,and the routing optimization objective including the above three metrics is given.Finally,to solve the problem of slow convergence in the early stage of traditional deep reinforcement learning algorithms,a routing algorithm based on pre-trained deep reinforcement learning is proposed.The simulation results verify the adaptability of the proposed algorithm to the dynamic network environment.Compared with the existing routing algorithms,the proposed routing algorithm has better performance in network lifetime,end-to-end delay and packet delivery rate.