Integrated Sensing and Communication (ISAC) and refers to the integration of radio communication and sensor technology within a standardized framework. It is a promising approach for vehicular applications, offering benefits in size, cost, power efficiency, and spectrum usage. This thesis will investigate a reinforcement learning (RL)-based ISAC system using MIMO automotive radar, inspired by the work in [1]. The system dynamically allocates antennas between sensing and communication tasks, adapting to unknown environments. The objective is to optimize sensing accuracy—measured via direction-of-arrival estimation—while ensuring reliable communication performance. The study involves RL algorithm design, antenna configuration strategies, and simulation-based evaluation.

In the course of the thesis, an understanding of the model and strategies used in [1] should be obtained, followed by implementation (MATLAB or other languages) and performance analysis for different scenarios.

Both the supervision and the writing of the thesis are possible in either German or English.

References:

[1] W. Zhai, X. Wang, M. S. Greco and F. Gini, "Reinforcement Learning based Integrated Sensing and Communication for Automotive MIMO Radar," 2023 IEEE Radar Conference (RadarConf23), San Antonio, TX, USA, 2023, pp. 1-6

With the introduction of solutions based on the Internet of Things (IoT) on various application areas, such as healthcare, industrial control and more, wireless data aggregation (WDA) becomes a promising solution for data collection from sensors with limited bandwidth. In essence, the superposition property of wireless multiple access channels (MAC) is utilized to accelerate our calculations, by letting the channels do the calculations for us. In practical networks, however, unfavorable signal propagation conditions for WDA prevail. To overcome this problem, an intelligent reflecting surface (IRS) is introduced to the network. The IRS is able to reflect incoming signals, while smartly tuning the phase shift at the individual reflect elements. This enables not only new but also controllable channel links improving the signal propagation conditions for WDA significantly.

References:

[1] T. Jiang and Y. Shi, "Over-the-Air Computation via Intelligent Reflecting Surfaces," 2019 IEEE Global Communications Conference (GLOBECOM), Waikoloa, HI, USA, 2019, pp. 1-6.

Also available at: https://arxiv.org/abs/1904.12475

Die Betreuung kann wahlweise in deutsch oder englisch erfolgen.

The concept of intelligent reflecting surfaces (IRSs) is considered as a promising technology for increasing the efficiency of mobile wireless networks. This is achieved by employing a vast amount of low-cost individually adjustable passive reflect elements, that are able to apply changes to the reflected signal. To this end, the IRS makes the environment real-time controllable and can be adjusted to significantly increase the received signal quality at the users by passive beamsteering. However, controlling each of the reflect elements individually entails a high computational complexity in practice. In order to alleviate this problem, we utilize the idea of resource bundling: By applying the same phase shifts on a block (tile) of adjacent reflect elements, the complexity of the system can be reduced significantly at the cost of a slight performance degradation.

References:

[1] A. Sezgin, B. Bandemer, A. Paulraj and E. A. Jorswieck, "Tile-based MIMO OFDM systems: The impact of inaccurate channel state information," 2008 42nd Asilomar Conference on Signals, Systems and Computers, 2008, pp. 1326-1329, doi: 10.1109/ACSSC.2008.5074633.

Also available at:  https://www.researchgate.net/publication/224505605_Tsile-based_MIMO_OFDM_systems_The_Impact_of_inaccurate_channel_state_information

[2] K. Weinberger, A. A. Ahmad, and A. Sezgin, “On Synergistic Benefits of Rate Splitting in IRS-assisted Cloud Radio Access Networks,” 2020, submitted to ICC2021

Also available at: https://arxiv.org/abs/2011.14763

Die Betreuung kann wahlweise in deutsch oder englisch erfolgen.

Transmit diversity has been studied extensively as a method of combating detrimental effects in wireless fading channels because of its relative simplicity of implementation. One attractive approach to transmit diversity is space-time block coding (STBC), in which full diversity is achieved. The main characteristic of these codes is their orthogonality property, which comes at the cost of the transmission rate. With the introduction of a quasi-orthogonal design higher transmission rates can be provided while sacrificing the full diversity. In order to alleviate this problem, we aim to utilize an intelligent reflecting surface (IRS), which employs a vast amount of low-cost individually adjustable passive reflect elements, that are able to apply changes to the reflected signal. To this end, we can utilize the IRS to support the quasi-orthogonal design and even orthogonalize the design at the users, potentially achieving both, full diversity and higher transmission rates.

References:

[1] B. Badic, M. Rupp and H. Weinrichter, "Quasi-Orthogonal Space-Time Block Codes: Approaching optimality," 2005 13th European Signal Processing Conference, 2005, pp. 1-8.

Also available at https://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.332.4395&rep=rep1&type=pdf

Die Betreuung kann wahlweise in deutsch oder englisch erfolgen.

Intelligent Reflective Surfaces (IRS) is a promising technology to improve coverage and energy efficiency through intelligent control of the propagation environment. In practice, however, achieving the expected benefits of IRS requires accurate channel estimation.  Most existing works assume the availability of perfect channel state information (CSI) to design the precoding vectors in the BS and the phase shift matrix in the IRS.  However, it is highly unlikely that this assumption will hold in practice. This is because, unlike conventional communications systems where channels can be estimated by actively transmitting, receiving, and processing pilot symbols, the IRS has no radio resources for transmitting and receiving pilot symbols and no signal processing capability for estimating channels. Therefore, it is challenging to obtain accurate CSI.

References:

[1] Z. Wang, L. Liu and S. Cui, "Channel Estimation for Intelligent Reflecting Surface Assisted Multiuser Communications," 2020 IEEE Wireless Communications and Networking Conference (WCNC), 2020, pp. 1-6, doi: 10.1109/WCNC45663.2020.9120452.

Also available at: https://arxiv.org/pdf/2104.01221.pdf

Beyond the fifth generation (B5G) wireless communication networks are expected to be exposed to unprecedented amounts of data traffic, thanks to the tremendous increase in the number of efficient mobile communication devices. Edge caching in wireless networks brings the content closer to users and promises to be an efficient technique to reduce latency and network congestion, especially during peak-traffic communication. However, the efficiency of caching depends on the distribution of user's content preferences. The proactive scheme of recommendation is capable of reshaping the content request probabilities of different users. Through analysis and implementation of the recommendation and caching scheme, a deep understanding of its merits shall be obtained. Different numerical simulations will give useful insights on the characteristics of the schemes.

References:

[1] Y. Fu, L. Salaün, X. Yang, W. Wen and T. Q. S. Quek, "Caching Efficiency Maximization for Device-to-Device Communication Networks: A Recommend to Cache Approach," in IEEE Transactions on Wireless Communications, doi: 10.1109/TWC.2021.3075278.

Also available at www.researchgate.net/Caching_Efficiency_Maximization_for_Device-to-Device_Communication_Networks_A_Recommend_to_Cache_Approach

Im Rahmen dieser Bachelorarbeit soll eine neue Methode zur Extraktion der Vitalparametern aus den Kanalinformationen des WLan-Systems implementiert werden. Die neue Methode nutzt einen Referenzkanal mittels Koax-Kabel zur Synchronisierung um eine hohere Genauigkeit zu erreichen und Hardwareeffekte auszugleichen.
Die Methode ist in [1] im Detail beschrieben. Die neue Implementierung soll unter gleichen Bedingungen gegen das verfügbare System verglichen und ausgewertet werden. Zusätzlich soll der Algorithmus mit einem alternativen Synchronisierungsschema [2], mittels verteiltem LO clock in den Vergleich mit einbezogen werden. Nach der erfolgreichen Implementierung und Test des Algorithmus wird eine Messreihe uber eine Gruppe von Personen durchgeführt, um Trainingsdaten von verschiedensten Atmungszuständen zu erhalten. Mit diesen Trainingsdaten wird erneut ein Classifier zur Detektion trainiert und in der Performance gegen die bestehenden Methoden verglichen.

[1] D. Zhang, Y. Hu, Y. Chen, and B. Zeng, "Breathtrack:Tracking indoor human breath status via commodity WiFi ," IEEE Internet of Things Journal, vol. 6, no. 2, pp. 3899{3911, 2019.

[2] S. Tewes and A. Sezgin, "WS-WiFi: Wired Synchronization for CSI Extraction on COTS-WiFi-Transceivers," in IEEE Internet of Things Journal, vol. 8, no. 11, pp. 9099-9108, 1 June1, 2021, doi: 10.1109/JIOT.2021.3058179.

In 5G networks, the amout of data that can be transmitted has increased significantely due to massive MIMO, larger frequencies and bandwidths and smaller cell areas. With a massive MIMO setup, beamforming strategies can be used to send different signal beams into different directions. In this thesis, it should be investigated how much different beamforming strategies can help in order to keep the location of the user equipment (UE) secret to (a) the base-station (BS) with a (possibly imperfectly) known channel and (b) a randomly-located adversary. Therefore, an according system model has to be implemented in Matlab. Afterwards, different precoding strategies have to be derived mathematically, which will then have to be tested in the simulations and analyzed.

The thesis can be written in English or German.

Literature:

https://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=9171875

In order to meet the rapidly increasing data rate demands, migration to Terahertz (THz) frequency bands becomes a key feature of future 6G wireless systems. These high fequency bands offer an ultra-broad bandwidth, however, this comes at a cost of high propagation loss and molecular absorption. Hence, massive MIMO antenna arrays are used to generate very narrow beams (pencil beams) in order to overcome the path loss. This makes the communication susceptible to channel blockage and beam misalignment, which can lead to frequent communication outages. To mitigate these effects and provide a reliable communication, a relay can be deployed, providing an alternative link to the user. In [1], the benefit of such a relaying strategy for joint blockage and beam misalignment mitigation is studied.

In the course of the thesis, an understanding of the model and strategies used in [1] should be obtained, followed by implementation (MATLAB) and performance analysis for different scenarios. The thesis can be done in either English or German.

Reference:

[1] Stratidakis, Giorgos, et al. "Relay-based blockage and antenna misalignment mitigation in THz wireless communications." 2020 2nd 6G Wireless Summit (6G SUMMIT). IEEE, 2020.

Also available at: https://arxiv.org/pdf/2003.03964.pdf

Title: Decentralized Radio Resource Management for Digital Twins and Sub-Metaverses: Synchronization in Multi-Cloud and Multi-MEC Networks

The aim of the master's thesis is to achieve a detailed understanding of the Multi-Cloud/Multi-MEC communication and computing architecture, the concept of Digital Twins and the Metaverse, and network decentralization. Methodologically, it should begin with a literature review of both references [1], [2], and other relevant works. Technically, the User-to-cloud association from [2] will be compared with the Metaverse and DT association from [1]. The system model from [2] will be extended to include the concepts of the Metaverse and Digital Twin (Synchronization, computation, etc.). As the work in [1] stops after the association, the extended system model from [2] will be further expanded to incorporate a decentralized radio resource management. The performance of the developed methods will be validated through various numerical simulations.

The work on this topic goes beyond the current state of research. The outcome of the master's thesis is an extended system model that combines concepts from [1] and [2], making a research contribution to the topic of decentralized resource management for Digital Twins and Sub-Metaverses.

References:

[1] O. Hashash et. al., 'Towards a Decentralized Metaverse: Synchronized Orchestration of Digital Twins and Sub-Metaverses,' 2023. [Online] https://arxiv.org/abs/2211.14686.

[2] R.-J. Reifert et. al., 'Distributed Resource Management in Downlink Cache-Enabled Multi-Cloud Radio Access Networks,' 2022. [Online] https://ieeexplore.ieee.org/document/9847048.