Objectives

• Study the use of Knowledge Distillation (KD) for predicting network performance in wireless and IoT networks.
• Compare global reduced models and link-specific models for predictive accuracy and computational cost.

Methodology

• Conduct a literature review on KD applied to IoT and wireless networks.
• Collect network performance data (reliability, latency, retransmissions) in various scenarios and configurations.

Expected Contributions

• A validated prototype demonstrating the effectiveness of KD in improving model performance.
• Recommendations for deploying KD techniques in constrained environments.

For applications, please send your CV and cover letter to lemia.louail@univ-lorraine.fr.