Reduction of mutual coupling between dual-polarized antenna elements using defected ground structures. - In: 17th European Conference on Antennas and Propagation (EuCAP 2023), (2023), insges. 5 S.
To accurately model a communication channel using channel sounding technique, having dual-polarized antenna elements and low mutual coupling between them is required. In this work a Defected Ground Structure (DGS) is proposed to reduce the mutual coupling between two dual polarized mmWave antenna elements in an array structure. The defected ground is made of etched C-shape slots. The presence of the DGS has a strong impact on H-plane coupling where a suppression of 35 dB is achieved, while the reduction of E-plane coupling is about 4 dB. The unwanted effect on the radiation patterns is negligible and the back radiation is not increased.
https://doi.org/10.23919/EuCAP57121.2023.10133636
Deep learning based anomaly detection for visual inspection on the industrial dataset. - Ilmenau. - 64 Seiten
Technische Universität Ilmenau, Masterarbeit 2023
Die Erkennung von Anomalien in visuellen Inspektionssystemen ist entscheidend um die Qualitätskontrolle und Sicherheit in verschiedenen Industriezweigen zu gewährleisten. Deep Learning basierende Techniken haben vielversprechende Ergebnisse bei der Erkennung anomaler Daten in industriellen Datensäetzen geliefert. In dieser Masterarbeit haben wir einen neuartigen Ansatz untersucht, der auf unüberwachtem Lernen basiert unter Verwendung von Convolutional Autoencodern. Das Ziel war die Erkennung von Anomalien im Montageprozess der Fertigungshalle. Zusäetzliche wurde Yolov5 verwendet um die Kameraaufnahmen füer eine bessere Anomalierkennung aufzubereiten. Die vorliegende Arbeit untersucht die Durchführbarkeit von unüeberwachtem Lernen auf dem industriellen Datensatz und vergleicht ihn mit überwachten Lerntechniken, die an gelabelte, beschriftete Daten gebunden sind und somit erhebliche Kosten verursachen. Die bearbeiteten Forschungsfragen lauteten: (1) Können Convolutional Autoencoder effektiv Anomalien erkennen, (2) was sind die zugrunde liegenden Mechanismen, die unüeberwachte Lerntechniken effektiver bei der Erkennung von Anomalien machen als üeberberwachte Lerntechniken (3) Was sind die möglichen Grenzen und Weiterentwicklungsmöglichkeiten des vorgeschlagenen Ansatzes ? Die Ergebnisse der Arbeit zeigen, dass die Verwendung von Convolutional Autoencoder für die Erkennung von Anomalien eine wirksame Strategie für die visuelle Inspektion ist. Die Überlegenheit des unüberwachten Lernens bei der Erkennung von Anomalien ist auf seine Fähigkeit zurückzuführen, die Struktur und Muster der Daten zu erfassen und auf neue Instanzen von anomalem Daten zu reagieren. Der vorgeschlagene Ansatz erzielte eine hohe Präzision, Recall und F1-Score. Die Einschränkungen und potenziellen Bereiche für Verbesserungsmöglichkeiten liegen in der weiteren Optimierung und Validierung mit größeren Datenmengen. Insgesamt hat der Ansatz ein enormes Potenzial zur Verbesserung bestehender Verfahren der Anomalieerkennung in visuellen Inspektionssystemen.
An efficient modeling approach of 1D-planar metamaterials in the high-frequency regime. - In: Compel, ISSN 2054-5606, Bd. 42 (2023), 3, S. 776-786
Purpose The purpose of this paper is to present an alternative modeling approach in terms of the determination of a physically equivalent circuit model for one-dimensional (1D) planar metamaterials in the high-frequency regime, without a postprocessing optimization procedure. Thereby, an efficient implementation of physical relationships is aimed. Design/methodology/approach In this paper, a method based on quasi-stationary simulations and mathematical conversions to derive the values for a physically equivalent circuit model is proposed. Because the electromagnetic coupling mechanisms are investigated in detail, a simplification for the considered multiconductor transmission line structure is achieved. Findings The results show that the proposed modeling approach is an efficient and physically meaningful alternative to classical full-wave simulation techniques for the investigated inhomogeneous transmission line structure in both the time domain as well as in the frequency domain. In the course of this, the effort is reduced while a comparable accuracy is maintained, whereby specific coupling mechanisms are considered in circuit simulations. Originality/value The process to obtain information about physically interpretable lumped element values for a given structure or to determine a layout for known ones is simplified with the aid of the proposed approach. An advantageous adaption of the presented procedure to other areas of application is well conceivable.
https://doi.org/10.1108/COMPEL-09-2022-0313
Various performance bounds on the estimation of low-rank probability mass function tensors from partial observations. - In: IEEE ICASSP 2023 conference proceedings, (2023), insges. 5 S.
Probability mass function (PMF) estimation using a low-rank model for the PMF tensor has gained increased popularity in recent years. However, its performance evaluation relied mostly on empirical testing. In this work, we derive theoretical bounds on the attainable performance under this model assumption. We begin by deriving the constrained Cramér-Rao Bound (CCRB) on the low-rank decomposition parameters, and then extend the CCRB to bounds on the mean square error in the resulting estimates of the PMF tensor’s elements, as well as on the mean Kullback-Leibler divergence (KLD) between the estimated and true PMFs. The asymptotic tightness of these bounds is demonstrated by comparing them to the performance of the Maximum Likelihood estimate in a small-scale simulation example.
https://doi.org/10.1109/ICASSP49357.2023.10095206
Rate splitting and precoding strategies for multi-user MIMO broadcast channels with common and private streams. - In: IEEE ICASSP 2023 conference proceedings, (2023), insges. 5 S.
In this paper, we present a precoder design for multi-user multiple-input multiple-output (MU-MIMO) broadcast systems with rate splitting at the transmitter. The proposed scheme applies to both underloaded and overloaded communication systems and supports the transmission of multiple common and private streams. We show how the generalized singular value (GSVD) and multilinear generalized singular value (ML-GSVD) decompositions can be used to define the number of common and private streams and adjust the message split. Additionally, we present transmit precoding and receive combining designs that allow the simultaneous transmission of common and private streams but do not require successive interference cancellation (SIC) at the receivers and can be used in cases where the total number of streams does not exceed the number of transmit antennas.
https://doi.org/10.1109/ICASSP49357.2023.10095138
Equivalence of aperture reduction in element space and constrained combination of DFT beams in beamspace. - In: IEEE ICASSP 2023 conference proceedings, (2023), insges. 5 S.
In this paper, we present an analytical proof of equivalence of the signal processing in the reduced aperture element space and in beamspace produced by the combination of multiple adjacent DFT beams with a subsequent constraining of the resulting magnitudes. This link finds applications in millimeter wave (mmWave) communications and radars that are typically equipped with a small number of RF chains and employ hybrid beamforming with analog phase shifters. This result unifies the transceiver designs, reduces complexity, and proves the applicability of state-of-the-art beamspace-based methods. It has a special implication for channel estimation at the initial stage when terminals acquire coarse estimates of the Sectors-of-Interest (SoIs). We show that the constrained groups of beams are equivalent to DFT beamformers of a smaller size aperture and present a closed-form expression of the corresponding effective aperture length as a function of the number of beams. We also derive an approximation of this expression to find the indices of the active array elements in a closed form. Finally, we verify this theory and analyze the accuracy of the proposed approximation using numerical simulations.
https://doi.org/10.1109/ICASSP49357.2023.10096523
Multi-environment based meta-learning with CSI fingerprints for radio based positioning. - In: 2023 IEEE Wireless Communications and Networking Conference (WCNC), (2023), insges. 6 S.
Radio based positioning of a user equipment (UE) based on deep learning (DL) methods using channel state information (CSI) fingerprints have shown promising results. DL models are able to capture complex properties embedded in the CSI about a particular environment and map UE’s CSI to the UE’s position. However, the CSI fingerprints and the DL models trained on such fingerprints are highly dependent on a particular propagation environment, which generally limits the transfer of knowledge of the DL models from one environment to another. In this paper, we propose a DL model consisting of two parts: the first part aims to learn environment independent features while the second part combines those features depending on the particular environment. To improve transfer learning, we propose a meta learning scheme for training the first part over multiple environments. We show that for positioning in a new environment, initializing a DL model with the meta learned environment independent function achieves higher UE positioning accuracy compared to regular transfer learning from one environment to the new environment, or compared to training the DL model from scratch with only fingerprints from the new environment. Our proposed scheme is able to create an environment independent function which can embed knowledge from multiple environments and more effectively learn from a new environment.
https://doi.org/10.1109/WCNC55385.2023.10118753
Framework for simulation models and algorithms in ISAC networks. - In: 2023 IEEE 3rd International Symposium on Joint Communications & Sensing (JC & S), (2023), insges. 6 S.
In Integrated Sensing and Communications (ISAC), radar and communications functionalities share the same radio channel and radio resources. This sharing allows the deployment of both functionalities at a low cost without additional frequency bands and infrastructure. Because radar and communications have different radio resource requirements, new radio access and radio resource management (RRM) methods are required in a shared system. To develop and evaluate new ISAC methods, this paper presents a modular simulation framework for the system level and link level of ISAC networks. The main components of the framework are the data storage, the simulators, and the algorithms. This paper presents the framework's architecture, the integrated simulation models for channel and signal simulation, and the integrated algorithms for multiple radar target localization and data transmission.
https://doi.org/10.1109/JCS57290.2023.10107507
DeModulation Reference Signal (DM-RS) densification in 5G-NR downlink. - Ilmenau. - 84 Seiten
Technische Universität Ilmenau, Masterarbeit 2023
Demodulation Referenzsignale (DM-RS) ermöglichen die kohärente Demodulation von physikalischen Kanälen in Systemen der fünften Generation (5G NR). Für multi-Layer-Übertragungen von physikalischen Kanälen ist jeder Layer assoziiert mit einem DMRS-Port, der im Zeit-Frequenzgrid des Transportblocks des physikalischen Kanals eingebettet ist. Ein Ziel dieser Arbeit ist die Erhöhung der Anzahl der verfügbaren DMRS-Ports eines 5G-Systems ohne die Erhöhung des DMRS-Overheads. Dieses erlaubt einer Basisstation (gNodeB) mehr Nutzer zu bedienen und erhöht die spektrale Effizienz von Uplink- oder Downlink-Übertragungen. In 5G Rel.15 werden bis zu 12 orthogonale DMRS-Ports unterstützt. In 3GPP Rel.18 wird die Erhöhung der DMRS-Ports bei gleichbleibendem Overhead untersucht. In dieser Arbeit werden verschiedene Methoden zur Erhöhung der Anzahl der DMRS-Ports untersucht sowie bezüglich ihrer Leistungsfähigkeit in Link- und Systemlevel-Simulationen miteinander verglichen. Zwei Methoden werden hierzu miteinander verglichen: längere Orthogonale Cover Codes (FD-OCC) als in 3GPP Rel. 15 und dünn-besetzte DMRS Resource-Allokationen im Frequenzbereich. Das FD-OCC ermöglicht die Orthogonalisierung von verschiedenen DMRS-Ports in der Code-Domain, wobei die DMRS-Sequenzen zu gleichen Zeit-Frequenz-Ressourcen gemappt werden. Für FD-OCC werden Hadamard- und Diskrete-Fourier-Transforms (DFT)-Ansätze untersucht. Ergebnisse aus Linklevel-Simulationen belegen, dass die erweiterten FD-OCC Ansätze eine ähnliche Performance wie die vorhandenen Rel. 15 Strukturen bzgl. der Kanalschätzung aufweisen. Eine dünnbesetzte Allokation der DMRS führt zu einer Verringerung der Leistungsfähigkeit bei Kanälen mit hohem Delay-Spread. Der Vergleich verschiedener FD-OCC-Längen für verdichtete DMRS-Strukturen mit Rel.15-Länge-2-OCC-Sequenz über verschiedene Interferenzniveaus hinweg wird ebenfalls in der Arbeit vorgestellt, um den Einfluss der Interferenz auf die verwendete OCC-Art und -Länge zu analysieren.
VANET-based traffic light management for an emergency vehicle. - In: Ubiquitous networking, (2023), S. 129-137
Wireless communications have affected our lifestyle in the last decades. It is helpful to improve quality of life for communities. Communications among vehicles usually take place in vehicular ad-hoc networks (VANETs). Vehicle-to-vehicle (V2V) and vehicle-to-infrastructure (V2I) communications are aspects of communications in the transportation which are growing rapidly. They can play a pivotal role in the transportation field. Management of traffic lights (TLs) is crucial to control traffic flow especially at an intersection. The goal of this paper is to manage the TLs at an intersection when an emergency vehicle (EV) is approaching. First, we simulate an intersection which includes TLs via simulation of urban mobility (SUMO). Later, we simulate VANETs communication to manage the TLs at the intersection when the EV is coming with the help of objective modular network testbed in C++ (OMNeT++) and vehicles in network simulation (Veins). Finally, the impact of V2I communication on delivery efficiency of the emergency service is investigated. Simulation results show an improvement in delivery efficiency of the emergency service.