Parallel Satellite Access Window Calculation via GPU Segmentation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current scheduling systems for imaging satellites face challenges in reducing computation time for determining satellite access windows and efficiently processing large-scale data, which affects customer satisfaction and operational efficiency.
Innovation Solution
The use of graphical processing units (GPUs) for parallel computing to determine satellite access windows and an implementation framework for executing native programs on high-performance distributed computing systems, allowing for simultaneous execution of instructions across large volumes of data and independent operation from underlying data storage and processing platforms.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of time
If traditional sequential computing methods are used to determine satellite access windows, then computation accuracy is maintained, but computation time becomes excessively long
Solution Approach 1:
The patent segments the computation of satellite access windows by dividing the constellation into multiple satellite groups and processing each group independently through separate compute nodes. This segmentation allows parallel computation of access windows for different satellite groups simultaneously, dramatically reducing total computation time while maintaining accuracy for each segment.
Solution Approach 2:
The patent transitions from sequential single-threaded computation to parallel multi-dimensional computation by utilizing multiple compute nodes and GPU accelerators. This dimensional change in computation architecture enables simultaneous processing of multiple satellite access window calculations, transforming the time complexity from linear to logarithmic scale.
2Productivity
If native programs are executed directly on high-performance distributed computing systems, then processing efficiency improves, but platform compatibility and portability are lost
Solution Approach 1:
The patent introduces an intermediary layer (the implementation framework) that sits between the native programs and the distributed computing system. This framework translates generic program requests into platform-specific execution commands, allowing native programs to run efficiently on HPC systems without direct coupling, thus maintaining both efficiency and adaptability.
Solution Approach 2:
The implementation framework provides universal functionality by supporting multiple programming languages and native program types through a common interface. This multi-functional design allows the same framework to execute C++, Java, Python, and other native programs on distributed systems, achieving both high performance and broad compatibility.
3Ease of operation
If imaging satellite control systems perform all tasks sequentially, then system simplicity is maintained, but operational responsiveness to customer requests deteriorates
Solution Approach 1:
The patent implements dynamic task distribution where the control system automatically assigns different computational tasks to available compute nodes based on current system load and resource availability. This dynamic allocation enables the system to maintain simplicity in control logic while achieving fast response times through adaptive parallel processing of imagery requests.
Data Source
AI summary
Systems and methods are provided for calculating satellite access windows for a constellation of imaging satellites. In some implementations, systems and methods are provided for managing execution of native programs on high performance computing systems.