TUBITAK 2523
Network Framework for In-Aerial Network Computing over Layered Aerial Networks
This project focuses on optimizing semantic communication in both inter-vehicle and intra-vehicle links, balancing parameters like system complexity and data size. It aims for superior throughput, surpassing Shannon’s bound, while ensuring resilience against communication challenges. The objective is to establish efficient communication in Vehicular Ad-Hoc Networks (VANETs), contributing to a smarter mobility environment.
In order to overcome the difficulties associated with large-scale data processing in harsh environments, the purpose of this project is to revolutionize Federated Learning (FL) by utilizing High-Altitude Platforms (HAPs). Even in situations where there is a lack of infrastructure or it is unreliable, we will be able to conduct efficient FL training by utilizing HAPs as mobile edge servers. It is possible for HAPs to provide stable and reliable communication links for the purpose of data exchange between ground devices and the central server due to their comprehensive coverage and high altitude. Cross-layer orchestration will be accomplished through the utilization of Software-Defined Networking (SDN), which will allow us to optimize both the performance of the network and the allocation of its resources.
This project that is being carried out in collaboration between Istanbul Technical University and Hongik University, Seoul, South Korea, will involve theoretical research, simulations, and implementation in real world.