A Machine Learning Framework for Performance Prediction of an Air Surveillance System

Juha Jylhä, Marja Ruotsalainen, Ville Väisänen, Kai Virtanen, Mikko Harju, Minna Väilä

Research output: Chapter in Book/Report/Conference proceedingConference contributionScientificpeer-review


The optimal use of a surveillance radar system requires proper understanding about the system behavior in different configurations, modes, and operating conditions. This paper proposes a machine learning framework for producing and validating the performance model of the surveillance radar system. The framework consists of an optimization method for the parameterization of a radar model and a machine learning method for the modeling of a tracker. Optimization and machine learning is based on the satellite navigation data of cooperative aircraft and corresponding track data from the surveillance system. The aim is to learn the system performance in a wide range of operating conditions using the extensive measurement history and then to predict the present performance with high accuracy at specified locations in the airspace. The feasibility of the proposed framework is assessed using real data.
Original languageEnglish
Title of host publicationEuropean Microwave Week 2017
Subtitle of host publication"A Prime Year for a Prime Event", EuMW 2017 - Conference Proceedings; 14th European Microwave Conference, EURAD 2017
ISBN (Print)978-2-87487-049-1
Publication statusPublished - 2018
MoE publication typeA4 Article in a conference publication
EventEuropean Radar Conference - Nuremberg, Germany
Duration: 11 Oct 201713 Oct 2017
Conference number: 14


ConferenceEuropean Radar Conference
Abbreviated titleEuRAD


Dive into the research topics of 'A Machine Learning Framework for Performance Prediction of an Air Surveillance System'. Together they form a unique fingerprint.

Cite this