Data collection method of roadmap coloring UAV energy-saving endurance based on metric learning

Abstract

The invention discloses a method for collecting energy-saving endurance data of a roadmap coloring unmanned aerial vehicle based on metric learning. It mainly solves the throughput and energy consumption problems in UAV-assisted sensor networks. The method includes: building a model of a UAV-assisted sensor network, and proposing an isomorphic network model. Nodes in the network all have moving lines and have the characteristics of group mobility and group behavior. The probabilistic landmarks are then applied to construct a global map. The metric learning method is used to optimize the trajectory of the UAV, and the metric matrix is ​​trained offline. In the offline training phase, the LMNN algorithm is used to construct the metric matrix. In order to update the node information, the broadcast Hello packet mechanism is adopted, and the Hello packet is broadcast hop by hop from the base station. In order to ensure a high transmission success rate and throughput, the graph coloring method is used to reduce the conflict of multi-hop transmission. Since the transmission of data packets may overlap with the broadcast information collection node, there is a problem of multi-hop channel allocation in this paper, so the network can Represented by a graph, each point in the graph represents a node in the network, and the edge connecting the two points represents the outgoing communication between the two points. Therefore, the channel assignment problem of a multi-hop network can be transformed into a graph coloring problem. In the UAV-assisted network that supports fast movement, the UAV flies faster, and the ground node can be a mobile robot with mobility, so the channel access method of prioritizing the contact time is adopted.

Description

technical field [0001] The invention belongs to the field of unmanned aerial vehicle-assisted sensor networks, in particular to a method for collecting energy-saving endurance data of unmanned aerial vehicles based on metric learning for coloring roadmap images. Background technique [0002] With the rapid development of electronic, sensor and communication counting, UAVs are widely used in various military and civilian fields. Applying drones to wireless sensor networks can effectively collect sensor data, prolong network life, and reduce network energy consumption. In Unmanned Aided Sensor Networks (UWSNs), the channel access algorithm is very important, which affects not only the system performance but also the energy efficiency of the sensor nodes. Therefore, designing an efficient channel access algorithm for UWSNs is a problem worthy of study. [0003] In the UAV-assisted sensor network, because UAVs usually have other tasks, UAVs fly according to a fixed flight traj...

AI Technical Summary

Problems solved by technology

[0006] In order to solve the problem of low throughput performance of fast-moving UAV-assisted network and insufficient energy consumption of UAV flight, the example of the present invention adopts the metric learning method for path planning to optimize the throughput. The method obtains the metric matrix through offline training and uses the metric online The matrix predicts the optimal information of throughput and energy consumption performance. In order to achieve the above-mentioned purpose, the example of the present invention provides a roadmap map coloring UAV energy-saving endurance data collection method for metric learning

Images