- Time Criticality
- Critical Instant
- Energy efficiency
- Artificial Intelligence
- Machine Learning
- Neural Networks
- Neural Networks
- Machine Learning
- Internet of Things
- Wireless Communication
- Functional Safety
- Single Point of Failure
Themes for Student Thesis:
Topic 1 – Medical Care Task Scheduling
Implementation of an ECG algorithm utilizing a Microcontroller and a set of sensors.
The objective of this Thesis is to determine the capabilities of a scheduling algorithm in the medical field. As there are several items to be considered, including criticality of the tasks, energy available and human conditions, it is important to analyze the timely and correct execution of the system.
We are looking for a responsible and proactive student who is willing to work, according to its experience level (Bachelor / Master), on an embedded System: M5Stack programmed with C/C++
Topic 2 – Scheduler Analysis
Analysis of Parameters for a Scheduling Algorithm
We are performing some tests in the scheduling area and developed a dynamic scheduling algorithm which could be used for a variety of scenarios. The problem is that we require the fine tuning of the parameters present in the algorithm.
The student (Bachelor/Master) will learn about scheduling algorithms, optimization and parameter selection over different techniques. The algorithm and test environment were created using C/C++, Matlab and FreeRTOS
Topic 3 – Energy Testing and Battery Analysis for a Multi-Criticality Scenario
Analysis of energy consumption and battery life in a multi-criticality tasks scenario
Our approach on a novel scheduler has proven to be energy efficient, however, we need to obtain more information to be sure that the low energy consumption is applicable on a variety of use cases. Therefore, it is necessary to identify, define, test and analyze the energy consumption in different configurations
The student (Bachelor, Master) should be interested in the study of energy consumption, furthermore, knowledge in C/C++ are beneficial
 Osvaldo Navarro, Jones Yudi Mori, Javier Hoffmann, Hector Gerardo Muñoz Hernandez, Michael Hübner:A Machine Learning Methodology for Cache Memory Design Based on Dynamic Instructions. ACM Trans. Embedded Comput. Syst. 19(2): 12:1-12:20 (2020)
 Fricke, Florian; Mahmood, Safdar; Hoffmann, Javier; Ali, Muhammad; Shahin, Keyvan; Hübner, Michael; Göhringer, Diana: (2020) Domain Adaptive Processor Architectures. In: Jasperneite J., Lohweg V. (eds) Kommunikation und Bildverarbeitung in der Automation. Technologien für die intelligente Automation (Technologies for Intelligent Automation), vol 12. Springer Vieweg, Berlin, Heidelberg
 Florian Fricke, Javier Eduardo Hoffmann, Safdar Mahmood, Michael Hübner - “A Tool to Ease Design-Space-Exploration Using the Tensilica LX7 ASIP” in CDNLive EMEA 2019
 J. Hoffmann: “Energy Aware Enhanced Earliest Deadline First Algorithm”. In: K. De Bosschere, Ed., “ACACES 2019: poster abstracts.” HiPEAC High-Performance Embedded Architecture and Compilation, Subiaco, Rome, 2019.
 J. Hoffmann, D. Kuschnerus, T. Jones and M. Hübner, "Towards a Safety and Energy Aware protocol for Wireless Communication," 2018 13th International Symposium on Reconfigurable Communication-centric Systems-on-Chip (ReCoSoC), Lille, 2018, pp. 1-6.
 J. Hoffmann: “An Energy-aware Dynamic Scheduler for Hard Deadline Tasks”. In: K. De Bosschere, Ed., “ACACES 2018: poster abstracts.” HiPEAC High-Performance Embedded Architecture and Compilation, Subiaco, Rome, 2018.
 Florian Kästner, Osvaldo Navarro Guzman, Benedikt Janßen, Javier Eduardo Hoffmann, Michael Hübner. Analysis of Hardware Implementations to Accelerate Convolutional and Recurrent Neuronal Networks. In International Journal on Advances in Software, vol 10 no 3 & 4, year 2017
 Javier Eduardo Hoffmann, Osvaldo Navarro Guzman, Florian Kästner, Benedikt Janßen, Michael Hübner, “A survey on CNN and RNN implementations,” in PESARO 2017, The Seventh International Conference on Performance, Safety and Robustness in Complex Systems and Applications, 2017
 Osvaldo Navarro, Jones Mori, Javier Hoffmann, Fabian Stuckmann, and Michael Hübner. 2017. A machine learning methodology for cache recommendation. In Proceedings of the International Symposium on Applied Reconfigurable Computing. Springer, 311--322.