UAV Positioning via Reinforcement Learning for Mobile Communication
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current mobile communication systems face challenges in adaptively supporting user equipment (UE) communication using unmanned aerial vehicles (UAVs), particularly in dynamically changing situations such as disaster monitoring or emergency scenarios, where traditional deployment methods are inflexible and may not optimize quality of service (QoS) for multiple users.
Innovation Solution
A method employing a reinforced learning network to determine the optimal position of a UAV based on real-time position information and communication state feedback from UEs, allowing the UAV to move and adjust its position dynamically to enhance communication performance and QoS, using a base station that transmits control information to the UAV for movement adjustments.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If traditional fixed deployment methods are used for UAVs, then device complexity is reduced, but adaptability to different communication situations deteriorates
Solution Approach 1:
The patent implements dynamic deployment by enabling the UAV to move between different positions (first position and second position) based on real-time communication requirements. The base station determines positioning information dynamically and transmits control information to adjust the UAV's position, transforming a static system into a dynamic one that adapts to changing communication scenarios.
Solution Approach 2:
The patent employs feedback mechanisms where the base station receives communication state information from the UAV and users, processes this information through a reinforcement learning network, and generates control information to adjust the UAV's positioning. This closed-loop feedback system enables continuous optimization of UAV deployment based on actual communication performance.
2Reliability
If reinforcement learning networks are used to optimize UAV positioning, then communication performance is improved, but device complexity increases
Solution Approach 1:
The patent introduces a reinforcement learning network as an intermediary component that mediates between the base station's control functions and the UAV's positioning adjustments. This specialized module processes communication state information and generates optimized control decisions, separating the complex optimization logic from the basic control functions and managing system complexity through functional decomposition.
3Reliability
If the UAV moves dynamically to optimize QoS, then service quality is improved, but loss of time for position adjustment increases
Solution Approach 1:
The patent applies preliminary action by having the base station determine positioning information and generate control information in advance based on predicted communication requirements. The reinforcement learning network processes information proactively to prepare optimal positioning decisions before communication performance degrades, reducing the time penalty associated with dynamic adjustments.
4Loss of information
If real-time control information is transmitted to the UAV, then communication state monitoring is improved, but use of energy increases
Solution Approach 1:
The patent implements partial action by transmitting control information selectively based on actual communication needs and state changes. Rather than continuous full-information transmission, the system monitors communication states and transmits positioning control information only when necessary to maintain or improve QoS, reducing energy consumption while preserving essential monitoring capabilities.
Data Source
AI summary
A method of a base station in a mobile communication system is provided, which includes receiving position information of at least one UE; determining an initial position of a UAV based on the position information; transmitting, to the UAV, control information related to the initial position and association information between the at least one UE and the UAV; receiving, from the UAV, first feature information related to a communication state between the at least one UE and the UAV; and transmitting control information related to a movement position of the UAV based on an output of a reinforced learning network to which the first feature information is input.