Deep Neural Network for Free Space Optical Communication Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Free space optical communication (FSOC) systems face challenges such as laser signal fading due to atmospheric turbulence and environmental factors like rain and fog, which affect data rates and require accurate performance prediction for flight path planning and communication reliability.
Innovation Solution
A method and toolchain system utilizing a deep neural network (DNN) for near real-time FSOC performance prediction, trained with data from external sources and simulation in five feature categories (propagation range, rain rate, visibility, cloud height, and cloud thickness), employing MODTRAN software for accurate atmospheric modeling and adjusting weights and biases to minimize prediction error.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional simulation methods are used for FSOC performance prediction, then accuracy can be maintained, but computation time becomes excessively long and cannot meet near real-time requirements
Solution Approach 1:
The system performs preliminary actions by collecting extensive FSOC performance data from external sources and simulations in advance, organizing it into training and testing datasets, and training the deep neural network model beforehand. This pre-computation enables the model to make rapid predictions during actual applications without performing complex simulations in real-time, thus achieving near real-time performance while maintaining accuracy.
Solution Approach 2:
The patent creates a computational copy of the complex FSOC simulation process by training a deep neural network model to replicate the behavior of traditional simulation methods. The DNN model learns the underlying patterns and relationships from simulation data, then uses this learned knowledge to predict FSOC performance rapidly without executing the original complex simulations, thereby achieving both speed and accuracy.
2Measurement precision
If deep neural network training is performed with extensive data from multiple sources, then prediction accuracy improves, but data processing complexity increases
Solution Approach 1:
The patent segments the data processing task into distinct phases: data collection from external sources, data organization into structured formats, division into training and testing datasets, and model training. This segmentation allows each phase to be handled systematically with appropriate tools and methods, reducing the overall complexity of processing extensive multi-source data for high-accuracy predictions.
Data Source
AI summary
Various embodiments provide a method for free space optical communication performance prediction method. The method includes: in a training stage, collecting a large number of data representing FSOC performance from external data sources and through simulation in five feature categories; dividing the collected data into training datasets and testing datasets to train a prediction model based on a deep neural network (DNN); evaluating a prediction error by a loss function and adjusting weights and biases of hidden layers of the DNN to minimize the prediction error; repeating training the prediction model until the prediction error is smaller than or equal to a pre-set threshold; in an application stage, receiving parameters entered by a user for an application scenario; retrieving and preparing real-time data from the external data sources for the application scenario; and generating near real-time FSOC performance prediction results based on the trained prediction model.