ΑΙhub.org
 

A neural network method for satellite anomaly detection

by
05 October 2023



share this:
satellite dish

Rural and remote communities in Canada often rely on satellites to access the internet, but those connections are fraught with many glitches and service interruptions because the technology can be unreliable. The inequity in internet access between these communities and those who live in cities is an ongoing problem with myriad consequences for Canada’s economic productivity.

A team of researchers from the University of Waterloo and the National Research Council (NRC) are tackling this long-standing issue using machine learning. The team’s method, the Multivariate Variance-based Genetic Ensemble Learning Method, merges several existing AI-driven models to detect anomalies in satellites and satellite networks before they can cause major problems.

“For remote areas in Canada and around the world, satellites are often their best option for maintaining internet access,” said Peng Hu, an adjunct professor of computer science and statistics and actuarial science at Waterloo and the corresponding author of the study. “The problem is that the operation of those satellites can be expensive and time-consuming, and issues with them can lead to populations being cut off from the rest of the world.”

The project was conducted at the NRC-Waterloo Collaboration Centre together with Yeying Zhu, associate professor of statistics and actuarial science, in a research project supported by the NRC’s High-throughput and Secure Networks Challenge program.

The researchers tested their method using three datasets: Soil Moisture Active Passive – NASA satellite monitoring soil moisture across Earth, Mars Science Laboratory rover – satellite data from the Mars rover, and Server Machine Dataset – data from a large internet provider.

The researchers chose these datasets both because of their public availability and because they’re representative of a large array of satellite uses.

In a series of tests, their new model outperformed existing models in terms of accuracy, precision, and recall.

“Satellite network systems are going to be more and more important in the future,” Hu said. “This research will help us to design more reliable, resilient, and secure satellite systems.”

The research, Multivariate Variance-based Genetic Ensemble Learning for Satellite Anomaly Detection, appears in the journal IEEE Transactions on Vehicular Technology.

You can read the research in full in the arXiv version.




University of Waterloo




            AIhub is supported by:


Related posts :



#AAAI2024 workshops round-up 4: eXplainable AI approaches for deep reinforcement learning, and responsible language models

We hear from the organisers of two workshops at AAAI2024 and find out the key takeaways from their events.
12 April 2024, by

Deep learning-powered system maps corals in 3D

A system developed at EPFL can produce 3D maps of coral reefs from camera footage in just a few minutes.
11 April 2024, by

Is compute the binding constraint on AI research? Interview with Rebecca Gelles and Ronnie Kinoshita

We hear from authors of work presented at AAAI 2024 studying access to compute and the impact this has on AI research and researchers.
10 April 2024, by

Forthcoming machine learning and AI seminars: April 2024 edition

A list of free-to-attend AI-related seminars that are scheduled to take place between 9 April and 31 May 2024.
09 April 2024, by

Modeling extremely large images with xT

Introducing a new framework to model large images on contemporary GPUs while aggregating global context with local details.
08 April 2024, by

Going top shelf with AI to better track hockey data

Waterloo researchers get an assist from AI in identifying hockey players with greater accuracy and speed.
05 April 2024, by




AIhub is supported by:






©2024 - Association for the Understanding of Artificial Intelligence


 












©2021 - ROBOTS Association