Probabilistic Tracking with Database Systems

Erscheinungsjahr 2014
ISBN: 978-3-89863-258-4
Verlag: GCA

Buch, Englisch, Band ISSN 2196-1352, 228 Seiten, PB, Format (B × H): 148 mm x 210 mm, Gewicht: 355 g

Reihe: Fraunhofer Series Advances in Sensor Data and Information Fusion

Probabilistic tracking is a standard technique for estimating the state of moving objects from sensor data. With increase in the numbers of sensors, however, sensor data has become overwhelming. This has led to the usage of database systems for sensor data processing. However, tracking tasks so far have been realized in external components. In this thesis, a solution for integrating tracking methods into generic relational data¬base systems (RDBMS) is provided. For four classes of representative sensor data pro¬cessing algorithms, relational and deductive rules are derived for imple¬menting proba¬bilistic tracking algorithms into RDBMS. These rules are declarative, descriptive formu¬lations for sensor data processing. For neces¬sary additional functionalities concerning probabilistic functions and matrix implementations, several implementation solutions are discussed and evaluated. A new phase concept on data streams enhances the implemen¬tation of anomaly detection. An analysis of the asymptotical runtime be¬havior of the deductive implementation shows its superlinear runtime which is un¬wanted for real-time applications. As a solution, incremental update propagation methods are applied which restores the linear runtime behavior.
The methods are tested on a prototypical system for the detection of criti¬cal situations in airspace monitoring. The test proofs the suitability for real-world scenarios, in this case allowing for processing flight data for all Ger¬man airspace in real time on stand-ard computer hardware.

Autoren/Hrsg.

Schüller, Gereon

Inhaltsverzeichnis

Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
1.1 Related Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10
1.2 Outline of this Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2 Design of a Tracking Information System . . . . . . . . . 21
2.1 The Coalition Shared Database – an Analysis . . . . . . 23
2.2 Tracking Inside a Database System – a new
Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
2.3 Components of a Tracking Information System . . . . . 30
3 Probabilistic Tracking . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
3.1 Excursus: Radar Data and Uncertainty . . . . . . . . . . . 39
3.2 Mathematical Foundations . . . . . . . . . . . . . . . . . . . . . . . 42
3.3 The Kalman Filter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
3.4 The PMHT Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . 51
3.5 The Interacting Multiple Model Filter . . . . . . . . . . . . . 60
3.6 Anomaly Criteria for Enhanced Situation Awareness 66
4 Database Technology . . . . . . . . . . . . . . . . . . . . . . . . . .. . . 71
4.1 The Relational Data Model . . . . . . . . . . . . . . . . . . . . . . 72
4.2 The Query Language SQL . . . . . . . . . . . . . . . . . . . . . . . 76
4.3 Deductive Databases . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.4 Probabilistic Databases . . . . . . . . . . . . . . . . . . . . . . . . . 89
4.5 Temporal Databases . . . . . . . . . . . . . . . . . . . . . . . . . . .. 91
4.6 Moving Object Databases . . . . . . . . . . . . . . . . . . . . . . . 94
4.7 Data Stream Management . . . . . . . . . . . . . . . . . . . . . . . 97
5 Implementing Tracking Algorithms in SQL . . . . . . . 103
5.1 Implementing the Kalman Filter . . . . . . . . . . . . . . . . . 109
5.2 PMHT in a Relational Database System . . . . . . . . . . . 112
5.3 Matrix Representation in Relational Databases . . . . . 117
5.4 Filtering Multiple Models. . . . . . . . . . . . . . . . . . . . .. . . 124
5.5 Expressing Anomaly Criteria in SQL . . . . . . . . . . . . . . 128
6 Accelerating Track Processing . . . . . . . . . . . . . . . . . . . 133
6.1 Incremental Update Propagation . . . . . . . . . . . .. . . . . 134
6.2 Incremental Calculation of PMHT . . . . . . . . . . . . . . . . 141
6.3 Incremental Calculation of the Kalman Filter . . . . . . 147
6.4 Evaluation of Update Propagation Methods . . . . . . . . 150
6.5 Indexing Covariance Matrices . . . . . . . . . . . . . . . . . . 154
6.6 Evaluation of the Bregman Ball Index . . . . . . . . . . . . . 160
7 Phase Management for Situation Assessment and
Awareness . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 163
7.1 Events and Phases in SQL . . . . . . . . . . . . . . . . . . .. . . 165
7.2 Phases as an Additional Layer in a Tracking Scenario 166
7.3 A Proposal for Extending SQL with Phases . . . . . . . . 171
7.4 Performance Issues . . . . . . . . . . . . . . . . . . . . . . . . . . . 176
8 Case Study: AIMS—An Airspace Monitoring
System . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 177
8.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . 178
8.2 System Architecture . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
8.3 View-Based Flight Analysis . . . . . . . . . . . . . . . . . . . . . 183
8.4 Maintaining Flight Phase Transitions . . . . . . . . . . . . . 189
8.5 In-Memory Caching for Fast Stream History Access . 191
8.6 Performance of AIMS . . . . . . . . . . . . . . . . . . . . . . . . . . 198
9 Conclusion and Future Work . . . . . . . . . . . . . . . . . . . . . 203
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Own Publications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 217
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223