TY - GEN
T1 - Indoor Mapping with a Mobile Radar Using an EK-PHD Filter
AU - Talvitie, Jukka
AU - Kaltiokallio, Ossi
AU - Rastorgueva-Foi, Elizaveta
AU - Barneto, Carlos Baquero
AU - Keskin, Musa Furkan
AU - Wymeersch, Henk
AU - Valkama, Mikko
N1 - Funding Information:
Corresponding author: Jukka Talvitie (jukka.talvitie@tuni.fi). This work was partially supported by the Academy of Finland (grants #319994, #323244, #328214, #338224), Business Finland under the project 5G VIIMA, Tampere University Graduate School, and MSCA-IF grant 888913 (OTFS-RADCOM).
Publisher Copyright:
© 2021 IEEE.
jufoid=57448
PY - 2021/9/13
Y1 - 2021/9/13
N2 - Integrated communications, localization and sensing is one of the most addressed technologies considered for future mobile communications systems. In this context, a user equipment (UE)-centric mobile radar has been proposed to introduce improved situational awareness, and consequently potential improvement in network performance. In this paper, we derive an extended Kalman probability hypothesis density (EK-PHD) filter with a novel feature model, for a mobile radar based environment mapping, where range-angle detections are used to track map objects over time for dynamic map construction. In order to evaluate the performance of the proposed filtering approach, we employ a realistic ray-tracing-based simulation setup, which models the full transmission chain from the transmitted IQ-samples to mapping results. Besides this, a simplified measurement model considering solely single-bounce specular reflections is exploited for providing further insight into the filter performance. The obtained results show that the proposed EK-PHD filter is able to provide high-quality mapping results, reaching around 10 cm landmark estimation accuracy in the considered millimeter wave simulation setup.
AB - Integrated communications, localization and sensing is one of the most addressed technologies considered for future mobile communications systems. In this context, a user equipment (UE)-centric mobile radar has been proposed to introduce improved situational awareness, and consequently potential improvement in network performance. In this paper, we derive an extended Kalman probability hypothesis density (EK-PHD) filter with a novel feature model, for a mobile radar based environment mapping, where range-angle detections are used to track map objects over time for dynamic map construction. In order to evaluate the performance of the proposed filtering approach, we employ a realistic ray-tracing-based simulation setup, which models the full transmission chain from the transmitted IQ-samples to mapping results. Besides this, a simplified measurement model considering solely single-bounce specular reflections is exploited for providing further insight into the filter performance. The obtained results show that the proposed EK-PHD filter is able to provide high-quality mapping results, reaching around 10 cm landmark estimation accuracy in the considered millimeter wave simulation setup.
KW - environmental mapping
KW - extended Kalman filter
KW - millimeter wave
KW - mobile radar
KW - probability hypothesis density
U2 - 10.1109/PIMRC50174.2021.9569630
DO - 10.1109/PIMRC50174.2021.9569630
M3 - Conference contribution
AN - SCOPUS:85118471678
T3 - IEEE International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC
BT - 2021 IEEE 32nd Annual International Symposium on Personal, Indoor and Mobile Radio Communications, PIMRC 2021
PB - IEEE
T2 - IEEE Annual International Symposium on Personal, Indoor and Mobile Radio Communications
Y2 - 13 September 2021 through 16 September 2021
ER -