Abstract
Due to the wide availability of 5G mobile networks, joint communication and radar sensing (JCRS) receives increasing attention by research communities. Here, radar sensing can be done as a side product of communication without additional hardware costs. In contrast to dedicated radar systems, the maximum range as well as the range resolution of these systems are limited. In this paper, we have investigated the limitations of radar systems through a classification problem, recognizing 10 digit-shaped foil balloons. For this purpose, we have recorded a dataset using a 77-GHz frequency modulated continuous wave (FMCW) radar. Furthermore, we have created multiple datasets with different quality levels by reducing the range resolution and the snapshot rate of the recorded measurements. Finally, we have analyzed the behaviours of two machine learning (ML) approaches, random forests (RF) and multilayer perceptron (MLP) to understand the limitations of restricted systems.
Original language | English |
---|---|
Title of host publication | 2022 56th Asilomar Conference on Signals, Systems, and Computers |
Pages | 564-568 |
Number of pages | 5 |
DOIs | |
Publication status | Published - 2 Nov 2022 |
Event | 2022 56th Asilomar Conference on Signals, Systems, and Computers - Pacific Grove, CA, USA Duration: 31 Oct 2022 → 2 Nov 2022 |
Conference
Conference | 2022 56th Asilomar Conference on Signals, Systems, and Computers |
---|---|
Period | 31/10/22 → 2/11/22 |
Keywords
- Radar measurements
- Target recognition
- Training data
- Radar
- Bandwidth
- Reflection
- Classification algorithms