A method and system for UAV trajectory optimization in Internet of Things data collection

Abstract

The invention discloses a method and system for optimizing the UAV trajectory in the data collection of the Internet of Things. First, a deep reinforcement learning framework network is constructed, and then state information including pheromone concentration is input into the strategy network to obtain the UAV trajectory. Action, loop this process; and use the pheromone concentration to calculate the instantaneous reward value and obtain the accumulated reward value. Through the status of the accumulated reward value, it is judged whether the optimization training is completed, and finally the optimized trajectory of the UAV is obtained according to the trained policy network. The present invention performs channel modeling on UAVs and IoT nodes in a simulation environment. The channel modeling considers the existence of line-of-sight LoS links and non-line-of-sight NLoS links at the same time, and can reflect the dynamic changes of the actual Internet of Things communication environment. Under the premise, the UAV can quickly complete the data collection of the Internet of Things. Executing the data acquisition task according to the optimized UAV trajectory can better meet the actual flight requirements of the UAV and complete the data acquisition task more accurately.

Description

technical field [0001] The invention relates to the technical field of unmanned aerial vehicle wireless communication, in particular to a method and system for optimizing trajectory of an unmanned aerial vehicle in data collection of the Internet of Things. Background technique [0002] UAVs have flexible mobility, and can approach potential IoT nodes and collect data through trajectory optimization with a low-power connection scheme. Therefore, UAV communication technology is expected to play a key role in the next generation of wireless communication systems, providing wider and deeper coverage and connections for the growing mass of wireless terminals. Compared with the IoT system based on the ground base station, the air base station system based on the UAV has remarkable characteristics, such as improving the line-of-sight channel probability, improving spectrum and energy efficiency, etc. [0003] In the existing UAV trajectory optimization training, the first thing i...

AI Technical Summary

Problems solved by technology

In addition, in the early stage of training, the strategy executed by the UAV is random, and reward acquisition requires a series of complex operations

Therefore, the data collection task of UAV is a sparse reward problem, which will cause the reinforcement learning algorithm to progress slowly in the iterative process, and even difficult to converge

Images