At the Yamamoto Laboratory, we conduct a wide range of research on network-related themes. Our current main research themes are as follows.

Master's Program

• High temporal resolution of network traffic by time-series data imputation using Transformer
• Network resource allocation based on user interest in video streaming services
• Mahjong winning tile prediction using an incomplete information estimation model based on Transformer
• Improving the explainability of anomaly detection in IDS using machine learning
• Improving embedding performance in virtual network embedding
• Autonomous anomaly detection and attack countermeasures in IoT networks

Undergraduate Program

• Task offloading optimization based on total energy consumption in MEC environments using genetic algorithms
• Hybrid packet loss concealment by deep learning and time-series prediction targeting ensemble sound sources
• Evaluating the impact of ad insertion positions and memorability on QoE in video streaming services
• Cooperative delivery methods for heterogeneous UAV swarms in dynamic environments
• For other main research themes, please refer to the Research Overview.

Regarding the selection of research themes, we basically encourage students to decide based on their interests through continuous paper reading after assignment, though we may assign a theme in some cases. However, regarding research, we respect students' autonomy and focus on independent thinking. We also provide guidance to help students acquire the manuscript writing and presentation skills necessary for presenting their research results.

Interviews are generally mandatory for laboratory assignment. We also place more emphasis on your autonomy to actively acquire new knowledge and the communication skills to convey your thoughts to others, rather than solely on academic ability.

Please note that we are currently not accepting research students due to laboratory capacity limits.

Undergraduate and graduate students are being recruited as usual.