Research On Unmanned Swarms Information Processing Task Scheduling For Cloud-to-End Fusion

With the rapid development of wireless communication,mechanical manufacturing,artificial intelligence and other related technologies,unmanned technologies are becoming gradually mature.Unmanned systems are widely used in many fields such as material delivery,disaster relief,and target reconnaissance.However,due to the increasing demand for unmanned applications,single unmanned platform is facing the challenge of endurance,load capacity,communication capability,and computing capability.The unmanned swarm with multiple unmanned platforms is gradually becoming the research hotspots for unmanned systems.Effective collaborative interaction is the key of the unmanned swarm.Flexible scheduling of resources such as communication and computing is the basis of collaborative interaction,which directly determines the collaboration ability and overall performance of the unmanned swarm.Based on this,this paper aims to optimize the task scheduling efficiency of information processing for cloud-to-end fusion in the unmanned swarm.Specifically,the following four aspects are studied:(1)A cloud-to-end fusion task scheduling framework for unmanned swarm is designed.Considering the organizational relationship between unmanned swarm and unmanned platform,the application characteristics of unmanned swarm,and the technical bottleneck of communication and computing in unmanned swarm,this paper designs an unmanned swarm cloud-to-end fusion task scheduling framework,which is the foundation for the cloud-to-end collaborative information processing.Combining the task scheduling framwork,this paper takes the battlefield situation reconnaissance and judgment as the background to concretize the key issues of unmanned swarm collaborative task scheduling.Besides,the scope of subsequent researches is determined.(2)A reliable data transmission method with energy optimization for unmanned swarm under unstable communication channel state is proposed.Aiming at the collaborative data transmission in unmanned swarm under unstable channel states,considering the constraints of unstable communication quality,variable communication topology,and unbalanced communication energy consumption,this paper models the unmanned swarm collaborative data transmission as the least energy consumption problem.The energy consumption of the unmanned swarm is reduced while ensuring the data transmission reliability.Futhermore,this paper proposes a novel method to reduce the time complexity of obtaining the optimal data transmission strategy.The experimental results show that the proposed dynamic transmission method has good timeliness and robustness under unstable channel states.(3)A reliable real-time task scheduling method with resource optimization for tactical cloud under unstable operating conditions is presented.Aiming at the real-time task scheduling of unmanned swarm in tactical cloud,considering the poor operating environment,the volatility of information processing capability and the failure of computing host,this paper proposes a dynamic selection strategy of task replicant and task resubmission.Considering the high volatility of information processing capability,a novel method for estimating the uncertain information processing time is presented.On this basis,this paper extends the fault-tolerant task scheduling method to improve the resource utilization while system reliability and task timeliness are guranteed.The experimental results show that the reliable real-time task scheduling method in tactical cloud has strong adaptability and reliability under unstable operating conditions.(4)A reliable scheduling method for unmanned swarm tasks with energy optimization under unstable channel states and working conditions is proposed.Aiming at the real-time scheduling by end-to-end collaboration,considering the constraints of unstable communication quality,variable communication topology,and the volatility of information processing capability,this paper proposes a dynamic selection strategy of local computing and collaborative computing.Through this method,the energy consumption of the unmanned swarm is greatly reduced while ensuring the data transmission reliability.In addition,this paper proposes a novel method to improve the time efficiency without affecting the performance of the method.The experimental results show that the end-to-end collaboration scheduling method has better energy efficiency and reliability under unstable channel states and working conditions.