Masoud Daneshtalab

Masoud Daneshtalab is a Professor at Mälardalen University, leading the Heterogeneous System research group (HERO). He previously held roles as a European Marie Curie Fellow at KTH Royal Institute of Technology (2014) and as a university lecturer and group leader at the University of Turku, Finland (2012-2014). His research focuses on interconnection networks, hardware/software co-design, deep learning acceleration, and evolutionary optimization. He specializes in fault-tolerant DNN accelerators, time-sensitive networking (TSN), and embedded systems. His work bridges theoretical advancements with practical implementations, emphasizing reliability and efficiency in edge computing and AI applications.

Research interests include:

• Network-on-Chip (NoC) architectures and congestion prediction
• Fault resilience in deep neural networks (DNNs)
• Optimization of federated learning and homomorphic encryption for edge AI
• Integration of TSN with 5G and automotive systems
• Hardware acceleration techniques for computational efficiency

Recent publications emphasize advancements in robust AI architectures, fault tolerance mechanisms, and TSN-based communication protocols. His work often addresses practical challenges in deploying machine learning models on resource-constrained devices. He actively contributes to interdisciplinary projects in autonomous systems, healthcare monitoring via FMCW radar, and neural architecture search for embedded applications.

Labs/Teams: Leads the HERO group at Mälardalen University, focusing on heterogeneous computing systems and real-time embedded systems.