University of Patras
Advanced signal processing techniques for next generation mobile networks in the mmWave spectrum
Abstract
dc:descriptionThe ever-increasing need for wireless data transmission at a very high rate has resulted in the study of systems that can occur in the 5G and 6G mobile networks with frequencies up to over 60 GHz. A key feature of the millimeter frequency bands is that, due to the severe path loss, the channel is very sparse, not only in time but also in the frequency and angular space, and the received signals at the antennas are correlated. High attenuation and high sensitivity to blocking require architectures and protocols to address these challenges. Hence, massive hybrid antenna arrays have been proposed in the relevant literature for transmitting these signals. Their hybrid structure seems to be necessary since a purely digital structure has high implementation costs and power consumption in massive MIMO antenna arrays. This structure consists of an analogue and a digital part, which, in the transmitter, is called precoder and, in the receiver, combiner. The basic architectures for implementing the analogue part of the hybrid antenna systems are divided into fully or partially connected and sub-connected (the antennas are organized into non-overlapping groups), which can be implemented through either phase shifters or switches. Channel knowledge is required for the design of these structures. For this reason, this thesis focuses on channel estimation techniques to identify open issues and propose new approaches. The application of conventional channel estimation techniques is prohibitive due to the large number of transmitter-receiver antennas and required training symbols due to the high bandwidth and low SNR before transmitter-receiver alignment. The channel estimation in the mmWave spectrum can be achieved either by scanning the angle space (e.g., through an exhaustive, hierarchical search) to find the best pair of beams between transmitter-receiver or by formulating a sparse recovery problem. The propagation environment is typically modeled by employing a geometric channel model that includes the path loss and angles of departure (AoD) and arrival (AoA) on the transmitter and receiver sides, respectively. The basic idea of parametric modeling is to estimate the involved angles and the corresponding attenuations rather than estimating the entire impulse response of the channel. In array signal processing, various AoA estimation algorithms have been designed for classical ULAs and are of particular interest to researchers for several decades in, e.g., radar and smart antennas. In the first part of the dissertation, assuming a hybrid architecture antenna array at the receiver, a novel preprocessing scheme is suggested which enables the full snapshot recovery as if we had a ULA, using measurements from the radio-frequency (RF) chains of the hybrid structure. Then, any classical high-resolution power spectrum technique can be employed to estimate the involved AoAs and thus the channel. Also, a plethora of efficient sparse coding approaches have been suggested to directly estimate the geometric parameters by taking advantage of the sparsity properties of the mmWave channel to reduce the computational cost of the spatial scanning methods. Moreover, given that 5G networks will consist of small and dense cells, cooperative communication among proximal devices that share similar channel characteristics will be a promising technology. In wireless communications, sparse representations have been utilized to efficiently model and estimate a device channel with the help of sparse coding approaches, enhancing the spectrum and energy efficiency of fifth-generation (5G) and Internet of Things (IoT) networks. Hence, the second part of this thesis focuses on joint sparsity models for the representation of the channels of proximal devices and distributed sparse coding-based estimation methods to formulate a fully distributed voting-based (adaptive weights) cooperative approach suitable for networks of multiple tasks with partially overlapping parameters.
Degree
thesis:*- Grantor dc:publisher
- University of Patras
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Trigka, Maria
- Τρίγκα, Μαρία
Subjects
dc:subject × 18- Δίκτυα 5ης γενιάς/6ης γενιάς
- Χιλιοστομετρικό φάσμα
- Εκτίμηση καναλιού
- MUSIC/H-MUSIC
- Κατανεμημένη εκτίμηση
- Αραιές αναπαραστάσεις/κωδικοποίηση
- OMP
- 5G/6G
- mmWave
- Channel estimation
- Distributed estimation
- Sparse representations/coding
- Επιστήμη Ηλεκτρολόγου Μηχανικού, Ηλεκτρονικού Μηχανικού, Μηχανικού Η/Υ
- Επιστήμες Μηχανικού και Τεχνολογία
- Μηχανική και συστήματα επικοινωνιών, Τηλεπικοινωνίες
- Electrical Engineering, Electronic Engineering, Information Engineering
- Engineering and Technology
- Communication engineering and systems, Telecommunications
Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/52718