Modelling bi-static uncertainties in sequential Monte Carlo with the GLMB model

Uney, Murat ORCID: 0000-0001-6561-0406, Narykov, Alexey ORCID: 0000-0003-2064-2900, Ralph, Jason ORCID: 0000-0002-4946-9948 and Maskell, Simon ORCID: 0000-0003-1917-2913
(2021) Modelling bi-static uncertainties in sequential Monte Carlo with the GLMB model. In: 2021 Sensor Signal Processing for Defence Conference (SSPD), 2021-9-14 - 2021-9-15.

[img] Text
SSPD21_BistaticTracking_1.3.pdf - Submitted version

Download (666kB) | Preview


Bi-static sensing, where the transmitter and receiver of sensors are separately located, underlies a wide range of collaborative sensing systems. Bi-static detections generally feature a signal time-of-flight (ToF) and an angle-of-arrival (AoA). The current practice in multi-object tracking uses the bi-static geometry to map these pairs onto a selected coordinate frame and filter the mapped detections with a noisy range-bearing (i.e., a mono-static) sensor model. However, the uncertainties in ToFAoA pairs are not equivalently captured by this model, and the sensing geometry may result in significant degradation of the modelling accuracy. We introduce bi-static likelihood and false alarm models together with Monte Carlo (MC) computational methods to accurately capture the uncertainties involved and use them within Bayesian filtering. We demonstrate the efficacy of our proposed model in simulations with multiple objects using a sequential MC version of the generalised labelled multi-Bernoulli (GLMB) track filter. We compare the filtering performance with the conventional approximation mentioned above.

Item Type: Conference or Workshop Item (Unspecified)
Divisions: Faculty of Science and Engineering > School of Electrical Engineering, Electronics and Computer Science
Depositing User: Symplectic Admin
Date Deposited: 29 Oct 2021 10:21
Last Modified: 18 Jan 2023 21:25
DOI: 10.1109/SSPD51364.2021.9541502
Related URLs: