Joseph Wenninger

Energy profiling technique for network-level energy optimization

This work presents an energy profiling technique to optimize energy consumption in networks, with special focus in wireless sensor networks, where energy consumption is usually a major constraint. The approach aims to organize data obtained from power simulation in order to provide useful information to designers at high level abstraction layers, so that they are able to observe the impact of different design decisions, and therefore optimize their designs in terms of energy. Other energy profiling techniques deal only with local scope profiling, in order to optimize the node architecture. However, the energy profiling technique proposed here takes into account network-wide information expanding the scope of energy profiling to energy aware network design and optimization. Finally, the approach is further developed and its possibilities are explored by implementing on a wireless sensor network simulator and using it to optimize energy consumption at the network level in a pilot scenario.

Smart demand response scenarios

Automated demand response has the potential to be an essential future tool for maintaining the balance of supply and demand in electrical energy systems with a very high density of generation from renewable sources. Although this scenario can become true in the near future, only very few actual implementations of automated demand response can be found in Europe. This paper is tackling this by analyzing current demand response implementations, placing them into a matrix of different aspects and strategies, with the goal to provide a systematical basis of current application scenarios of demand response, highlighting barriers and starting points for further development decisions. Recommendations for important and possible near future application scenarios of demand side energy management concepts in Austria are the result of this analysis by a multidisciplinary team of researchers. This paper describes related and important future work for Austria in demand response in context of a defined scenario kit. Aspects and strategies of each scenario are described in the following subsections.

Response-surface-based design space exploration and optimisation of wireless sensor nodes with tunable energy harvesters

In an energy harvester powered wireless sensor node, the energy harvester is often the only energy source, therefore it is crucial to configure the microcontroller and the sensor node so that the harvested energy is used efficiently. This paper presents a response surface model (RSM) based design space exploration and optimisation of a complete wireless sensor node system. In our work the power consumption models of the microcontroller and the sensor node are defined based on their digital operations so that the parameters of the digital algorithms can be optimised to achieve the best energy efficiency. In the proposed technique, SystemC-A is used to model the systems analogue components as well as the digital control algorithms implemented in the microcontroller and the sensor node. A series of simulations are carried out and a response surface model is constructed from the simulation results. The RSM is then optimised using MATLABs optimisation toolbox and the results show that the optimised system configuration can double the total number of wireless transmissions with fixed amount of harvested energy. The great improvement in the system performance validates the efficiency of our technique.

The effect of smart appliances and smart gateways on network loads

Smart appliances and smart gateways (between private households and energy providers) are a hot topic, due to the promises of more convenience, more energy saving, less cost. This paper focuses on the impact of such devices on the load of networks, i.e. the view of energy providers. This is achieved by implementation of a simulator framework for sets of households (arranged in big buildings or even small villages) which considers social and environmental factors. The framework is presented in detail.

Assembler through the looking glass: Understanding digital systems

There are two ways of teaching students how computer systems work, in particular how the gap between a written program and the processor hardware is bridged, i.e. the assembler stage done by compilers. The traditional way is teaching by textbooks, which is very static. The other way is teaching and learning through interactive experiments, which is more dynamic and allows students to dive into the subject matters more easily. This paper focuses on how to give students an understanding of an assemblers code generation and on some experiences gained on the teachers side while using a self explaining, interactive graphical assembly language translator. The assembler visualization tool shown here is part of a whole suite of applications and hardware components used within various lectures and practical labs.1

System refinement design flow based on semi-symbolic simulations

Range based system simulations are increasingly preferred in recent years to cope with the performance issues inherent with standard multi-run simulations. Traditionally, deviations of nominal system parameters are considered statistically by steadily varying the system parameters and simulating all of these parameter space realizations in multiple simulation runs. In range based or semi-symbolic simulations, deviations of system parameters are modeled in continuous ranges expressed as symbolic quantities. Therefore, a range based system simulation achieves in one simulation run the result for all the modeled parameter deviations, thus significantly reducing the computation effort. This work uses a semi-symbolic simulation environment to obtain a range based system response. Subsequently the symbolic labeling of ranges is used to trace back the resulting sub-ranges of the system response to their respective parameter origin to identify potential refinement candidates. Based on this capability a refinement design flow is introduced which allows by refinements to improve the robustness and accuracy of cyber physical systems. Identifying refinement candidates is manifold and is demonstrated in an example by assessing a communication receiver SNR metric.

Power optimization of wireless sensor networks at design time

In the ever growing area of wireless sensor networks it is very important to estimate the energy usage of sensor nodes, especially if they should be powered by a battery. Often the accumulated energy consumption of a message delivery across the whole network is of interest too, in order to estimate battery life times, since message forwarding needs energy too. For rapid development cycles, fast but accurate simulations are needed. The approach presented in this paper uses transaction level modeling (TLM) to speedup simulation of wireless communication systems. The fundamental idea is that air communication can be modeled similary to a multi master bus. Nodes are simulated by employing a multi-threaded instruction set simulator (ISS) and models of the peripheral components. The energy usage for sensing, receiving and transmitting can be estimated in the simulation and this information can then be attached to TLM messages for accumulation related to messages. The example use case throughout this paper is a tyre pressure monitoring system. The modeling and simulation is done in SystemC. 1

Simulation of Ultra-Low Power Sensor Networks

This chapter gives an introduction into the importance of power analysis and power harvesting in wireless sensor networks. It additionally gives an overview over simulation techniques for this topic and an overview of related tools and methodologies, but mainly focusing on SystemC and SystemC AMS and extension libraries.

Model Based Design of Smart Appliances

As motivated in the Foreword, power management is a multi-disciplinary and complex issue. Designing and optimizing any part of it requires considering and optimizing as well the overall system. This can best be done by modelling and simulation. This chapter describes methods that allow the overall system simulation of smart buildings, including wireless networks at different layers of implementation, and energy using products at mixed levels of abstraction.

Watching a processor at work: A self-explanatory simulator and illustrator for the MC8 microprocessor

At the faculty of electrical engineering at the Vienna University of Technology there is a basic course called “Digital Systems”. Ambitious efforts are made to teach our students its contents in several ways: beside traditional courses to an audience of about 250 students using textbooks, there is also practical work in small classes of about 25 as well as laboratory work that has to be done on an individual basis by each student. The number of students is rather high and still increasing. Several years ago, the amounts of weekly hours of these courses have been reduced from 6.5 ECTS points to 4 ECTS points due to the “Bologna Process”. Now the laboratory is included in the practical work and has to be attended by all students during a very short time frame. In order to manage this new situation, e-learning becomes more and more important: instead of using real hardware like evaluation boards, power supplies and logic analyzers, the students can download a simulation and illustration tool that simulates all parts usually used and handled in the laboratory. This gives us the possibility to support many more students in the same time. The tool shown here reduces effort for both the students and the lecturers.