Technology-aware Asymmetric 3D-Inteconnect Architectures: Templates and Design Methods
In this project funded by the DFG, the chair for hardware-oriented technical computer science aims to develop architectural templates and design methods for 3D-interconnect architectures for heterogeneous 3D-SoCs. New production methods enable the design of heterogeneous 3D-System-on-Chips (3D-SoCs), which consist of stacked silicon dies manufactured with different technologies. In contrast to homogeneous SoCs, this allows to adjust the technological characteristics of each die to the specific requirements of the components placed in each layer. Heterogeneous 3D-SoCs provide unprecedented integration possibilities for embedded and high performance systems. To exploit that potential, powerful, flexible, and scalable communication infrastructures are required. Yet, current interconnect architectures (IAs) tacitly assume a multilayer homogeneous 3D-SoC and do not consider the influence of different technology parameters on the topology, architectural, and micro-architectural level of the IA.
We target two main innovations: First, we will exploit the specific technology characteristics of individual chip layers in heterogeneous 3D-SoCs. Therefore, we will re-evaluate and extend existing approaches for heterogeneous and hybrid 2D-interconnect architectures. Second, we aim at discovering new interaction mechanisms among components, which may be spatially distributed even at the micro-architectural level, to exploit their diverse features when manufactured in different technologies. The combination of these aspects leads to technology asymmetric 3D-interconnect architectures (TA-3D-IAs), as defined in this proposal for the first time.
Adaptive Data Management in Evolving Heterogeneous Hardware/Software
In this DFG funded project the chair of hardware-oriented computer science works on concepts to integrate Co-Processors such as FPGAs and GPUs in adaptive data base systems. We work on optimization strategies not only exploiting individual device-specific features but also the inherent cross-device parallelism in multi-device systems. Thereby we focus on operators from the relational and graph domain to derive concepts not limited to a certain application domain. To achieve the project goals, interfaces and abstraction concepts for operators and processing devices are defined. Furthermore, operator and device characteristics are made available to all system layers such that the software layer can account for device specific features and the hardware layer can adapt to the characteristics of the operators and data. The availability of device and operator characteristics is especially important for global query optimization to find a suitable execution strategy. Therefore, we also analyze the design space for query processing on heterogeneous hardware, in particular with regards to functional, data and cross-device parallelism. To handle the enormous complexity of the query optimization design space incurred by the parallelism, we follow a distributed optimization approach where optimization tasks are delegated to the lowest possible system layer. Lower layers also have a more precise view on device-specific features allowing to exploit them more efficiently. To avoid interferences of optimization decisions at different layers, a focus is also set on cross-layer optimizations strategies. These incorporate learning- based techniques for evaluating optimization decisions at runtime to improve future optimization decisions.
Broadband GPR for the localization and identification of landmines
Landmines are located and identified using tomographic images that require broadband microwave illumination up to 4 GHz. Different broadband antenna systems are investigated and tested with regard to their lateral resolution. Signal processing methods are developed that optimize the depth-dependent resolution and minimize the ambiguity of identification.
MEMS components for broadband RF data transmission
Micro-Electro-Mechanical Switches (MEMS), mechanical switches with dimensions in the micrometer range, are actuated by means of electrostatic forces. Due to their size MEMS elements can be integrated with semiconductor components. They offer a low-noise replacement for semiconductor switches. Selected MEMS structures are modeled, fabricated, measured and integrated into RF circuits. Modelling includes appropriate RF equivalent circuit diagrams, electromechanical models for the actuation mechanism and noise models. MEMS elements are manufactured in the university's clean room. The obtained models are verified metrologically. Tunable filters, VCO’s and controllable phased array antennas using MEMS elements are set up and measured.
Capacitive and ohmic microelectromechanical switches with spring steel bridge structures, in particular for high-frequency applications
The aim of the project is to design, manufacture, optimize and characterize electrostatically actuated capacitive and resistive microelectromechanical switches (MEMS switches) using spring steel bridge structures. The moveable bridges are to be produced in a monolithic manner by sputtering spring steel cathodically and in a hybrid manner by connecting the substrate with a structured spring steel foil. In both approaches, the capacitive or ohmic MEMS bridges are to be modified by using additional electrically and thermally highly conductive metal layers. At least for the monolithic approach, switch structures should also be considered, in which the bridge is made up of three parts consisting of a spring area (made of spring steel), a contact area (e. g. using silver) and a spring area (made of spring steel). Using these MEMS switches, tunable and reconfigurable RF filter and antenna structures are to be realized.
The Department of Neuro-Information Technology (NIT) is involved in a number of research projects on the federal funded program "ZWANZIG20 - Partnerschaft für Innovation ", which is a million-strong funding program for Eastern Germany (selection):
• 3D gesture interaction and fusion of 3D images
• Contact-free camera-based measurement of vital parameters with improved immunity to interference
• Active line camera systems for fast and high-resolution 3D measurement of large surfaces
Within the framework of the innovation alliance, the department is focussed on questions of "human-machine interaction" in current and future research. The goal of the Alliance is to fundamentally change the interaction between man and machine.
In a transdisciplinary and intersectoral research approach, the development of a new generation of 3D technologies for image acquisition, image processing and visualization, as well as the interpretation of complex scenarios in real time, will be carried out.
Among other things, the safety of the human in manufacturing processes is to be increased, the mobility in urban and rural areas to become more independent of health and age-related impairments and the possibilities for health care will be improved by identification of abnormalities and dangers. Research will be integrated into the fields of cognitive and neurosciences, social, labor as well as information science.
Automatic Pain Recognition based on Facial Expression and Psychobiological Parameters
The assessment of acute pain is one of the basic tasks in clinics. To this day, the common practice is to rely on the utterance of the patient. For mentally affected patients this is little reliable and valid. For non-vigilant people or newborns it cannot be used at all. However, there are several characteristics t hat indicate pain. These include specific changes in the facial expression and in psychobiological parameters like heart rate, skin conductance or electrical activity of skeletal muscles.
We are working towards an automatic system, which can distinguish whether a patient feels pain or not, and can assess the intensity of the pain. Based on experiences in facial expression recognition our system can distinguish facial expressions of pain from others and rate the intensity of the expression. In the current comprehensive study, we investigate the relations between pain, the facial expressions and the psychobiological feedback. Further, the data recorded in the study, which is named BioVid Heat Pain Database, is available to the scientific community. In the project, we collaborate with the Emotion Lab of the University of Ulm.