Research On The Problem Of UAV Data Transmission

In recent years,with the rapid development of mobile communication networks,the demand for data traffic is increasing,which will bring challenges to the current terrestrial communication system.It is difficult to solve the above problems through traditional ground deployment.In addition,when faced with environmental problems caused by rugged terrain or force majeure factors,traditional deployment costs are too high and difficult to deploy.UAV(Unmanned Aerial Vehicle,UAV)is widely used in wireless communication systems due to its strong flexibility,small size,and easy deployment.This paper mainly applies the UAV communication system to the grassland environment.Because the grassland environment is characterized by a vast area,a changeable climate,and few base stations,there will be problems in the information transmission of the communication system.Therefore,a UAV cooperative relay transmission system is designed to solve the problem of data transmission and expand the coverage of communication.The main research contents are as follows.First,this paper collects data based on grazing in grassland pastoral areas,and designs the grazing path,taking the end point of the grazing path as the target point for UAV data collection.First,model the pastoral area to obtain the overgrazing area and the usable area of the pastoral area.According to the stock carrying capacity,the number of sheep and the number of grazing days as the criteria for the division of the rotational grazing area,the three operators in the genetic algorithm are optimally designed.The concept of connectivity ensures that the algorithm reasonably divides the rotational grazing area.On this basis,the information backtracking mechanism and dynamic detection mechanism of the ant colony algorithm are improved to avoid the algorithm from falling into an infinite loop,thereby improving the convergence speed of the algorithm and quickly finding the shortest path for grazing,providing technical support for UAV data transmission.Second,a deep reinforcement learning algorithm is proposed for UAV data collection.First,during the grazing process of the flock,in order to ensure the observation of the dynamics of the flock and the environmental conditions of the pastoral area,it is necessary to wear a collar sensor on the flock.The drone ensures the effectiveness of data transmission by collecting data from the flock collar sensor.Then,the sensor capacity of the collar is calculated,and the flight path of the data collected by the UAV is determined by the size of the sensor capacity and the dynamics of the flock.In the process of flight,energy consumption is also a factor to be considered.In the grassland environment,wind speed is a unique variable.Therefore,adding wind speed variables in the process of calculating energy consumption improves the authenticity of the simulation environment.Finally,on this basis,it is ensured that the UAV can complete the data acquisition process with limited energy.The results show that the algorithm can quickly search for the amount of data stored in the sensor capacity,and plan a real-time dynamic acquisition path.Third,a reinforcement learning algorithm is proposed for UAV data transmission.When the transmission distance is relatively long,the data cannot be sent back to the ground user in time.Therefore,it is necessary to join the relay UAV for data transmission to ensure the integrity of the communication system.Stability and throughput as a measure of data transfer size.First,output the data of the planned UAV collection path,and then obtain the transmission distance between the UAVs through the throughput formula to determine the throughput,and calculate the distance between the UAVs.Finally,through the experiment The results show that the throughput between UAVs can be expanded by algorithm calculation,and the data transmission of UAVs can be completed.