UAV-IoT AoI Path Planning With DRL and Device Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing space-air-ground integrated UAV-assisted IoT data collection systems face high computation complexity in path planning, neglect local optimization, and lack techniques combining deep reinforcement learning with matching theory, leading to suboptimal data freshness and coordination between UAVs and IoT devices.
Innovation Solution
A method that constructs a UAV-assisted space-air-ground integrated IoT system using deep reinforcement learning and matching theory to minimize Age of Information (AoI), involving a Markov decision process, neural networks, and the Soft Actor-Critic algorithm to optimize UAV flight paths and device matching, while introducing virtual agents for point-to-multipoint data collection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional path planning methods are used for UAVs in space-air-ground integrated systems, then the system can collect data from IoT devices, but the computation complexity becomes excessively high and the system falls into local optimization
Solution Approach 1:
The patent segments the complex path planning problem into multiple sub-problems by dividing the service area into grid cells and processing UAV paths in sequential time slots. This segmentation reduces the overall computation complexity while maintaining data freshness through coordinated multi-UAV operation.
Solution Approach 2:
The patent employs deep reinforcement learning to create dynamic path planning that adapts to changing system states in real-time. The UAVs continuously update their paths based on current positions, IoT device data generation rates, and system AoI states, avoiding static optimization that leads to local optima.
2Productivity
If existing path optimization techniques are applied to UAVs, then flight paths can be optimized, but the matching between UAVs and IoT devices is neglected leading to suboptimal system performance
Solution Approach 1:
The patent merges path optimization and UAV-IoT device matching into a unified deep reinforcement learning framework. The joint optimization simultaneously determines optimal UAV trajectories and device assignments, improving data collection efficiency while managing coordination complexity through integrated decision-making.
Solution Approach 2:
The patent creates a universal matching mechanism where UAVs can dynamically serve multiple IoT devices based on real-time conditions. The same reinforcement learning agent handles both path planning and device matching functions, making the system adaptable to different service scenarios without requiring separate optimization mechanisms.
3Area of stationary object
If the distribution of IoT devices is extended on the ground, then coverage is improved, but the coordination between multiple UAVs becomes more complex and directly influences overall system AoI
Solution Approach 1:
The patent segments the extended service area into discrete grid cells, with each cell potentially served by different UAVs in different time slots. This spatial segmentation manages coordination complexity by localizing UAV-device interactions while maintaining broad coverage through systematic grid traversal.
Solution Approach 2:
The patent adds the time dimension to UAV coordination by implementing sequential time-slot-based service. Instead of attempting to coordinate all UAVs simultaneously across the extended area, the system processes UAV movements and device assignments in temporal sequence, reducing coordination complexity while maintaining coverage.
4Reliability
If deep reinforcement learning is used for UAV path planning, then optimization performance can be improved, but the high-dimensional state space creates computational challenges
Solution Approach 1:
The patent extracts and focuses on the most critical state variables for UAV path planning, such as UAV positions, target device locations, and AoI values. By selecting only the essential high-impact features from the full state space, the system maintains optimization performance while reducing the effective dimensionality that requires computational processing.
Data Source
AI summary
A space-air-ground integrated UAV-assisted IoT data collection method based on AoI comprises: constructing a UAV-assisted space-air-ground integrated IoT system, constructing a UAV channel model and an AoI model, establishing an AoI-based UAV-assisted space-air-ground integrated IoT data collection model, transforming a problem into a Markov problem, introducing a neural network to solve a high-dimensional state problem, introducing a deep reinforcement learning algorithm to train UAVs to find optimal collection points, and introducing a matching theory to match the UAVs and IoT devices. To meet the requirement for the timeliness of information collection, the invention finds the optimal configuration of flight parameters of UAVs and deduces the restrictive relation between performance indicators such as AoI, system capacity and energy utilization rate, thus effectively improving the timeliness of information collection, reducing the management and control complexity of the system, and improving the application level of AI in the IoT field.