Thesis supervisor: Péter Babarczi
Location of studies (in Hungarian): Távközlési és Médiainformatikai Tanszék Abbreviation of location of studies: TMIT
Description of the research topic:
Owing to the even growing complexity of communication networks instantaneous reaction to sudden environmental or traffic changes become humanly impossible, which would be essential to avoid noticeable performance degradation by the applications and the end users. Hence, the adaptation of resources within milliseconds to the changed conditions requires automated control mechanisms, and humans need to be replaced with sensors, computers and artificial intelligence. Therefore, the need for intelligent network algorithms which analyze, control and adjust resources in communication networks on a sufficiently small time-scale is increasing. Although the meaning of intelligence cannot be easily defined and has been considered in several different ways in different fields, the recent advances point towards that maximizing the number of possible reactions to certain challenges captures best its intuitive usage. For example, intelligent communication networks are prepared to the unknown future, e.g., with redundant resources against link failures, or with efficient bandwidth usage against resource scarcity. Although some existing network algorithms can be considered to be intelligent based on this definition, they require task-dependent fine-tuned objective functions. Therefore, it is essential to improve our understanding of the intrinsic motivation of intelligent network algorithms, generalize this knowledge towards task-independent objectives, and develop new practical and theoretical insights of intelligence in communication networks.
Open problems:
- Analyze different theoretical models of intelligence, define its meaning in communication networks based on the physical and information-theoretical models proposed in robotics and other fields.
- Identify and generalize the intrinsic motivation of intelligent algorithms in communication networks, including routing algorithms, resource allocation methods and consistent flow migration approaches.
- Analyze the applicability of intelligent algorithms in the automated control of self-driving networks.
- Investigate the effect of new technologies on the intelligent behaviour, including but not limited to the effect of network coding on the complexity of the consistent flow migration problem in special and general network topologies.
- Support claims with large scale simulations, develop software prototype.
Required language skills: angol Number of students who can be accepted: 1
Deadline for application: 2020-06-15
2024. IV. 17. ODT ülés Az ODT következő ülésére 2024. június 14-én, pénteken 10.00 órakor kerül sor a Semmelweis Egyetem Szenátusi termében (Bp. Üllői út 26. I. emelet).