Research On Collaborative Awareness Business Scenario Modeling Of Unmanned Cluster Network

Unmanned swarm sensing and detection is an emerging field of UAV research,and has a wide range of application prospects in field reconnaissance,target search and rescue and other scenarios.Establishing a service model is the prerequisite and key to build and test unmanned swarm networks.A reasonable service model can provide effective guidance for the structure and protocol design of unmanned swarm networks,and the generated service traffic can be used to test the performance of unmanned swarm networks.However,there are few academic discussions on the construction scheme of service models in unmanned trunked networks today,so this thesis presents a detailed study of this issue.In this thesis,we construct a new unmanned swarm cooperative sensing service model,design a sensing service transmission path selection strategy to determine the transmission path of sensing services in the swarm network,and verify the effectiveness of the above model and strategy through simulation experiments.Specifically,the main research of this thesis includes:The research of unmanned swarm network is still at the beginning stage,and this thesis investigates how to build a model of unmanned swarm cooperative sensing service.By analyzing the perception logic of UAV sensors when swarms perform perception tasks,three typical types of perception services in unmanned swarm networks are summarized:periodical,event-triggered and video-based.By analyzing the sensing logic of the sensing service,the mathematical distribution of packet interval of the sensing service source generating service messages is summarized,and the traffic model of the sensing service source is constructed.On the basis of the service source traffic model,this thesis further analyzes and constructs the service traffic model when the sensed services are aggregated in three aggregation modes:cascade,parallel and cascaded-parallel on the relay nodes in the unmanned swarm network,respectively,and designs a service traffic generation software suitable for the service source traffic model.Through simulation experiments,the rationality of the service source model and the practicality of the traffic generation software are verified.Since the determination of the service path is anchored to the application scenario of the unmanned swarm,in the collaborative sensing scenario,the mission objective of the unmanned swarm is to collect the heterogeneous sensing data of the sensing UAVs quickly and completely and complete the situational awareness of the swarm deployment area after comprehensive processing.Therefore,the selection of service transmission paths in this scenario should consider minimizing the end-to-end delay of the service,but selecting the transmission paths only by minimizing the end-to-end delay of the service will lead to network link congestion and thus service data loss.To solve the above problems,this thesis first proposes an end-to-end delay calculation method based on network calculus to derive the end-to-end delay of sensing service relaying in unmanned swarm networks using network calculus theory,and further designs a service path selection strategy based on reinforcement learning using this calculation method.The strategy gives the optimal transmission paths for different sensing services in the unmanned swarm network by considering the end-to-end delay constraint of sensing services and the resource constraint within the unmanned swarm.Simulation results show that the perceptual service path selection strategy designed in this thesis can significantly reduce the delay of service transmission and the consumption of swarm resources by service transmission.The service model established in this thesis has some reference value for designing unmanned swarm networks under cooperative sensing scenarios.