UAS Path Planning Using SINR Heatmaps for BVLOS Connectivity
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Unmanned aerial systems (UAS) face challenges in optimizing paths for safe and efficient operation beyond visual line of sight (BVLOS), particularly due to signal interference, limited cellular network coverage, bandwidth constraints, communication dropouts, battery life, and varying atmospheric conditions, which hinder reliable communication and data transmission.
Innovation Solution
A system and method using a machine learning model trained on various wireless characteristics to optimize UAS paths, generating signal-to-interference-plus-noise ratio (SINR) heatmaps and state space information, and employing reinforcement learning to determine optimal movement actions for maintaining connectivity, minimizing energy consumption, and maximizing communication quality.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional path planning objectives (minimization of flight time or distance) are used, then flight efficiency is improved, but communication reliability deteriorates in areas with limited cellular coverage
Solution Approach 1:
The patent transforms the path planning problem from minimizing flight time/distance to maximizing communication quality metrics (SINR, bandwidth, connectivity). The machine learning model learns to optimize paths based on wireless channel characteristics rather than traditional geometric parameters, fundamentally changing the optimization parameters to resolve the contradiction between flight efficiency and communication reliability
Solution Approach 2:
The patent introduces SINR heatmaps and wireless channel state information as intermediary elements that mediate between the drone's physical path and communication quality. These heatmaps serve as a bridge that allows the path planner to indirectly optimize communication reliability by planning paths through regions with favorable wireless characteristics rather than directly controlling communication parameters
2Adaptability or versatility
If drones operate beyond cellular network coverage areas to expand operational range, then operational flexibility is improved, but communication connectivity deteriorates
Solution Approach 1:
The patent performs preliminary generation of SINR heatmaps and wireless channel state predictions before the drone actually flies. This advance preparation allows the system to identify and plan paths through regions with predicted good communication characteristics, ensuring connectivity is maintained even when operating beyond traditional coverage areas. The preliminary action of creating communication maps enables expanded operational range without sacrificing reliability
3Measurement precision
If drones transmit high-quality video feeds in real-time to operators, then data quality is improved, but bandwidth consumption increases leading to network congestion
Solution Approach 1:
The patent changes the optimization parameter from transmitting all data at maximum quality to selectively transmitting data based on learned channel conditions. The machine learning model determines appropriate transmission parameters (quality, rate, timing) based on predicted bandwidth availability, transforming the approach from constant high-quality transmission to adaptive transmission that maintains data quality when possible while conserving bandwidth when necessary
Solution Approach 2:
The patent applies partial action by transmitting only the necessary portion of video data at high quality rather than continuously transmitting full-resolution feeds. The system selectively transmits critical data when channel conditions permit and reduces transmission quality or skips non-critical frames when bandwidth is constrained, thereby maintaining operational data quality while reducing overall bandwidth consumption
4Reliability
If drones actively search for cellular networks to maintain connectivity, then communication reliability is improved, but energy consumption increases reducing flight time
Solution Approach 1:
The patent performs preliminary mapping of wireless channel conditions and identifies regions with reliable cellular coverage before the drone departs. This advance knowledge allows the drone to fly directly through predicted good-signal areas without needing to actively search for networks during flight, thereby maintaining communication reliability while minimizing the energy-consuming network search operations
Solution Approach 2:
The patent enables the drone to self-navigate to regions with favorable wireless characteristics using the pre-computed SINR heatmaps and learned channel state information. Rather than continuously searching for networks, the drone autonomously follows paths that the machine learning model has identified as having reliable connectivity, reducing active searching and associated energy consumption while maintaining communication reliability
Data Source
AI summary
An electronic device and a method for path optimization for unmanned aerial system (UAS) is provided. The electronic device includes a memory to store a machine learning model to be trained. The electronic device retrieves a first plurality of parameters related to one or more wireless environments to be utilized by the UAS and generates signal-to-interference-plus-noise ratio (SINR) heatmap information. The electronic device further generates state space information based on the generated SINR heatmap information and trains the machine learning model based on the generated state space information. The trained machine learning model indicates one or more movement based actions to be taken by the UAS at a plurality of states defined in the generated state space information. The electronic device further controls the UAS to travel a predefined path in at least one of the one or more wireless environments based on the trained machine learning model.