Research at the Electronic Systems Group

The ES group is world-renowned for its design automation and embedded systems research. It is our ambition to provide a scientific basis for design trajectories of digital electronic circuits and embedded systems. Current highlights include:

• Computational modelling and model-driven design flows
  
  This includes dataflow analysis, scenario-aware dataflow, Pareto algebra, design-space exploration, multiprocessor mapping, schedule synthesis, software synthesis, high-level and logic synthesis. More information on some of these topics can be found on the following pages:
  • SDF3
  • scenario-based design
  • scenario aware dataflow (SADF)
  • pareto algebra
  • POOSL
  
  
• Parallel and multiprocessor architectures
  
  This includes application-specific multiprocessors, SIMD and VLIW architectures, GPUs, real-time systems including networks on chip, memory controllers, and systems on chip (SOCs). More information on the GPU research can be found on the Parallel Architecture Research Eindhoven (PARsE) pages. The outcomes of this research are also used in the Embedded Computer Architecture, Advanced Computer Architecture, and Processor Design courses. More information on the real-timeCompSOC platform, which is also used in the Embedded Systems Lab.
  
  
• Dependable networked embedded systems and sensor networks
  This topic includes system-level performance analysis, QoS and resource management, and optimized protocol stacks. More information can be found here.
  
  
• System-level design solutions
  We research cross-cutting ("non-functional") topics such as real-time systems, virtualization, mixed-criticality applications, adaptivity for energy, power, variability, and thermal constraints, system scenarios by themselves, but always apply them in the application domains given above.
  
  
• Deep-submicron integration
  Next to prototyping our systems on FPGAs, we also design real hardware to be able to take advantage of the silicon for (ultra) low-power design, performance optimization, and manufacturability.
  
  
• Multimedia and imaging solutions
  Although we research multiple application domains, we have a strong heritage in multimedia and imaging, including on application-architecture co-design, video processing, feature extraction, imaging for health and intelligent traffic solutions. We teach several related courses.
  

Tools and platforms

Our research results in tools and platforms such as

• SDF3 dataflow analysis tooling
• Octopus toolset for Model-Driven Design-Space Exploration
• POOSL parallel object-oriented specification language
• MAMPS real-time multiprocessor platform
• CompSOC real-time partitioned/composable multiprocessor platform
• DRAMPower for power estimation of DRAMs
• Pareto Calculator for compositional computation of sets of Pareto points

Some of these tools are in widespread and/or industrial use.

Research funding

The ES group has been very successful in attracting funding for its research through national and international projects and collaborations. Ongoing projects are:

• ASAM on automatic architecture synthesis and application mapping for heterogeneous multi-processor embedded systems.
• BENEFIC on real-time middleware for scalable applications and adaptive multicore architecture
• CARM on architecture design and design-space exploration for high-performance, nano-precision mechatronic systems
• Cobra on scenario-aware dataflow for homogeneous multicore architectures
• Flextiles on dynamically reconfigurable multiprocessors with virtualised hardware, including embedded FPGAs
• EVA on a generic Embedded Vision Architecture for embedded in electromechanical equipment for industrial inspection and production.
• MORPHEUS on motion analysis with contactless camera sensing for professional polysomnography and baby sleep-watching at home
• NEST on multicore streaming applications
• T-CREST on real-time multiprocessor systems, in particular virtualisation and real-time DRAM memory controller
• RECAP on remote health monitoring of patients and elderly people
• Robust Cyber-Physical Systems (RCPS), a multi-project research program on uniting the embedded systems and control disciplines. ES is involved in four projects:
  • Integrated scheduling and control for cyber-physical systems.
  • High-performance control for nano-precision systems.
  • Energy efficient computer-brain interaction.
  • Control based on data-intensive sensing.
• SenSafety on heterogeneous hierarchical sensor networks for safety and security in public places