Using Apache Spark To Speed Analysis Of Ads B Aircraft Tracking Data Techniques
Abstract
The U.S. Navy is exploring the feasibility of using a big-data platform and machine-learning algorithms to analyze combat-identification data. Combat identification involves a large number of remote sensors that report back data for aggregation and analysis. In this thesis, we used a sample of ADS-B aircraft-tracking data to test big-data methods for machine-learning methods developed previously. We showed large speed improvements in the analysis setup over the previous single-processor methods, and a lesser speed improvement for machine-learning based anomaly analysis.
Document Details
- Document Type
- Technical Report
- Publication Date
- Jun 01, 2018
- Accession Number
- AD1060123
Entities
People
- Jim Z. Zhou
Organizations
- Naval Postgraduate School