GUHA, R. Modelling of UAV Communications in Integrated Terrestrial and Non-Terrestrial Networks [online]. Brno: Vysoké učení technické v Brně. Fakulta elektrotechniky a komunikačních technologií. 2025.

Posudky

Posudek vedoucího

Kirubakaran, Balaji

The presented thesis is relevant to current trends in mobile network architecture, especially in the context of integrating UAVs with terrestrial and non-terrestrial networks. The topic is timely and aligns with ongoing industry and research efforts aimed at improving network coverage, flexibility, and user experience through the use of aerial platforms. The student has applied a solid combination of techniques—clustering (DBSCAN + Mean Shift), constrained K-means, and Genetic Algorithms—which are well-suited for optimizing UAV deployment and user association. The use of Python simulations along with geospatial datasets adds practical value and makes the work applicable to real-world scenarios. This type of modeling has strong research potential, especially for future applications in 6G and integrated NTN systems. That said, grammar throughout the thesis can be improved, and LaTeX formatting needs better attention. The simulation part could also be enhanced—for instance, by including 3D visualizations of the network setup to clearly demonstrate UAV-user distributions and movement.

Navrhovaná známka
C
Body
75

Posudek oponenta

Možný, Radek

This thesis tackles a contemporary and important problem: the integration of UAVs into hybrid terrestrial and non-terrestrial networks to enhance Quality of Experience (QoE). The topic is well contextualized within the current 5G/6G evolution, and the theoretical groundwork is comprehensive. The applied methods DBSCAN + Mean Shift clustering, constrained K-means, and Genetic Algorithms are appropriate for the optimization tasks and well-motivated. The thesis is logically structured, with a good state-of-the-art section and a clear problem formulation. On the practical side, the work is adequate in scope and demonstrates sufficient depth for a Master's thesis. The simulations are implemented using Python and include real-world geospatial scenarios, which adds realism. However, several formal shortcomings reduce the overall quality: poor figure resolution, formatting inconsistencies (e.g. large blank spaces, single words on lines), repeated or missing abbreviation explanations, and generally weak typesetting. Full code listings included at the end would have been better placed in an appendix or submitted as a separate file. Furthermore, referencing is inconsistently applied, and grammatical errors are frequent.

Navrhovaná známka
B
Body
83

Otázky

eVSKP id 167446