Small logo of ETH main building ETH Zurich : Computer Science : Pervasive Computing : Distributed Systems : Events : : Workshop Contributions




Summer School on Wireless Sensor Networks and Smart Objects

August 29 - September 3, 2005, Schloss Dagstuhl, Germany

Workshop Contributions

Session 1, Tue, 9 - 10.30

Julia Downes: Wireless Environmental Sensor Networks; Protocols, configuration and Inter-Connectivity

Wireless Environmental Sensor Networks provide an effective method of collecting and disseminating data from large numbers of sensor devices. They allow detailed environmental data to be distributed in near-real time through mediums such as the Internet. A common requirement of these networks is the use of low-cost, low power devices, which impose constraints on the suitable protocols and network configurations. This presentation looks at a collaborative project between the University of North Florida and the National Weather Service USA to provide real-time mesoscale weather data from a network of nodes throughout North Florida. The ensuing Road Weather Information System is designed to provide drivers with detailed up to date weather information accessible from road side kiosks or the Internet. Research showed that the IEEE 802.15.4 protocol is ideally suited to such networks and when used with a suitable routing protocol can form the basis for a network solution. This and other options will be discussed, along with the possibility of enabling interconnection between such devices and networks - moving towards true ubiquitous computing.

Reinhard Bischoff: Wireless Force Monitoring of Bridge Stay Cables

The aging of the road and railway infrastructure and the demand for higher traffic intensity and loads require an increasing management, inspection and maintenance effort. The current praxis based on periodic, visual inspections may become too expensive in the near future. Hence, the demand for autonomous monitoring of civil engineering structures using physical sensors is rising worldwide. The goal is to optimize the period of visual inspections and to provide up-to-date information about the performance of the structure.
The objective of this talk is to present a prototype wireless sensor network for monitoring of civil structures. An application is the monitoring of the cable tensile force of cable-stayed bridges. Each stay cable is equipped with a node. It measures the accelerations of the cable that are induced by traffic or wind. Since data communication requires a much higher power demand than data processing, the signals acquired by the sensor have to be processed on the node for achieving a significant data reduction. Because of this, the filtering of signals, the estimation of natural frequencies and the algorithm to compute the force is applied to the dataset on the node. The final result, one value per dataset, is then transmitted to the base station of the network. A status monitor visualizes the data and performs a pre-diagnostic for identifying abnormal states. Additionally, the collected data is stored in a database that can be investigated by a query tool to get specific data for further investigation.
The talk presents laboratory tests performed with the prototype on a scaled cable stayed bridge.
This work is partially funded by the EC within the FP6 Integrated Project “Sustainable Bridges”.

Antonio Miraglia: Anti Fire detection system with wireless sensors network

Often commercial anti-fire systems reports false positive events when environmental signals show quick fluctuations: deploying an error-free anti-fire system, is generally expensive because the system needs a high number of sensors and complex connections between sensors and other system entities.
New technology solutions provide networks of intelligent nodes in order to filter false-alarms and sends data in digital mode on bus connections or wireless broadcasting. Commercial solutions cannot easily be integrated in larger contexts like anti-intrusion, or video-survillance systems or any other building automation systems.
The correlation of temperature and humidity data with the detection of smoke presence, enables to detect fire-event removing false-alarms and trasmission errors; the use of wireless sensors networks allows a non invasive and high density deployment of sensor nodes in the rooms monitored with the aim to obtain a precise fire-detection.
A fire detect wireless node in a wireless sensors network is composed of node provided of temperature and humidity sensors and a smoke detector connected on a port of the same node.
The informations flow spreads in the sensors network like single events:
• reading events (reading of temperature and humidity sensors) ;
• allert events (contact open or closed on node port connected to smoke detectors);
This events are collected by our security monitoring application (SMS); SMS is a simple client of events with the aim to listen reading and allert events from sensors network and correlate them in order to detect a fire event.
A finite state machine was designed in order to manage and check the occurred events and the security state of the monitored environment; this machine is able to recognize fire event when it happen. A fire event is a particular condition that happen in a confined room linked to a rapid rise of temperature and a rapid decrease of humidity.
When a fire event happen, the temperature grows like an experimental parametric characteristic defined in ISO Standard 834 (international fire figthers standard) or there is a smoke event and temperature is higher of a limit value.
The parametric characteristic ISO 834 and the fire work load are assumed in order to respect international security standards and security italian law about fire detection.
This application is able to manage some rooms in a office or house environment, display and record the security state and the temperature and humidity trend for each monitored room.
Finally the SMS application is able to manage a Network camera in order to give a visual check of occorred event in a single room;

Christian Frank: Algorithms for Generic Role Assignment in Wireless Sensor Networks.

We consider configuration of wireless sensor networks, where certain functions must be automatically assigned to sensor nodes, such that the properties of a sensor node (e.g., remaining energy, network neighbors) match the requirements of the assigned function. Essentially, sensor nodes take on certain "roles" in the network as a result of configuration. To help developers with such configuration tasks for a variety of applications, we propose "generic role assignment" as a programming abstraction, where roles and rules for their assignment can be easily specified using a configuration language. We present such a role specification language and distributed algorithms for role assignment according to such specifications. We evaluate our approach and show that efficient and robust generic role assignment is practically feasible for wireless sensor networks.

Matthias Gauger: FlexCup - Flexible and Efficient Code Updates for Sensor Networks

The ability to update the program code installed on wireless sensor nodes plays an import role in the highly dynamic environments sensor networks are often deployed in. Such code update mechanisms should support flexible reconfiguration and adaption of the sensor nodes but should also operate in an energy and time efficient manner. In my talk, I will present FlexCup, a flexible code update mechanism that minimizes the energy consumed on each sensor node for the installation of arbitrary code changes. The main concept of FlexCup is the ability to exchange individual components instead of replacing the complete code image. FlexCup dynamically integrates these components into the existing program code on the sensor nodes. I will present the main concepts of FlexCup and a short evaluation using a comparison with two related approaches.

Eiko Yoneki: Bridging WSNs to the Internet: Issues on Event Filtering, Aggregation and Correlation

Mobile devices play an important role for collecting sensor data over ad hoc networks and conveying it to the Internet backbone nodes, where no network infrastructure is supported in remote locations. When constructing such dynamic ad hoc networks, a robust data routing protocol is an important issue, but data semantics should also be integrated for efficient and accurate information dissemination. The sensed data need to be filtered, correlated, and managed when they flow over heterogeneous network environments. These mobile devices are resource-constrained, and a detectable/implementable event detection mechanism is required. Event management will be a multi-step operation from event sources to final subscribers. I would discuss issues to be addressed on event filtering, aggregation and correlation semantics for wireless sensor data in such global computing environments.

Session 2, Tue, 10.45 - 12.15

Yann-Ael Le Borgne: Polling Strategy fo Predictions in Wireless Sensor Networks

Strong spatio-temporal dependencies are often observed between readings in a sensor network. Given a query involving a set of sensors, it may be possible to only solicit a subset of them, and to predict the readings of the remaining ones within some user defined accuracy. In this talk, we adress the problem of varying the sensor subset used for predictions, so as keep on collecting readings from all sensors, and to evenly distribute energy consumption.

Michael McHugh: Event detection in audio sensor networks

A brief talk discussing why we feel that event detection in audio sensor networks is useful, and the approaches that we are taking. We'll cover the data we've captured so far, the next steps we plan to take, and what we hope to gain from the research.

Aylin Aksu: Joint Sensor Selection and Data Routing in Sensor Networks

We propose a new joint sensor selection and routing algorithm, which selects a set of sensor nodes (sensing nodes) in a sensor network to take measurements, and determines a set of paths connecting the sensing nodes to the sink node. Our objective is to maximize the network lifetime, while satisfying the data precision required by the user. We first develop a multi-objective optimization model for this problem and design the near-optimal OPT-RE algorithm based on this model for network lifetime maximization. Next, we design a low complexity heuristic called SP-RE. SP-RE first labels the links between the nodes with a metric which trades off the residual energies of the transmitting and receiving nodes with the required transmission and reception energy. Then, SP-RE calculates the shortest paths from all nodes to the sink, and identifies the node which is closest to the sink as a sensing node. This process is repeated until the required data precision is satisfied. We demonstrate by simulations that SP-RE and OPT-RE can increase the network lifetime several orders of magnitude compared to naive approaches.

Muneeb Ali: MMAC: A Mobility-Adaptive, collision-Free MAC Protocol for Wireless Sensor Networks

Mobility in wireless sensor networks poses unique challenges to the medium access control (MAC) protocol design. Previous MAC protocols for sensor networks assume static sensor nodes and focus on energyefficiency. In this talk, we present a mobility-adaptive, collision-free medium access control protocol (MMAC) for mobile sensor networks. MMAC caters for both weak mobility (e.g., topology changes, node joins, and node failures) and strong mobility (e.g., concurrent node joins and failures, and physical mobility of nodes). MMAC is a scheduling-based protocol and thus it guarantees collision avoidance. MMAC allows nodes the transmission rights at particular timeslots based on the traffic information and mobility pattern of the nodes. Simulation results indicate that the performance of MMAC is equivalent to that of TRAMA in static sensor network environments. In sensor networks with mobile nodes or high network dynamics, MMAC outperforms existing MAC protocols, like TRAMA and S-MAC, in terms of energyefficiency, delay, and packet delivery.

Ioannis Krontiris: Secure Communication and In-network Processing

We will consider the problem of securing communication between sensor nodes in large-scale sensor networks. A scalable, distributed and deterministic key management protocol designed to satisfy authentication and confidentiality, without the need of a key distribution center will be described. We will then extend our presentation on how a security protocol can utilize aggregator nodes which are responsible for data aggregation and command dissemination. We will describe the requirements for this case and how they can be satisfied without compromising the security level. We will finally present some research challenges currently existing in this area.

Robert Eigner: Ideas for an efficient, adaptive and scalable information exchange in mobile ad-hoc networks

The first part of the talk gives an introduction to the project Network on Wheels (NOW). NOW is a project dealing with mobile ad-hoc networks connecting vehicles. The main constraint in this field of networking is not electrical power but network load, namely bandwidth and contention time limitations. Approaches so far have relied on "geocasting", a transmission scheme that addresses network nodes by their geographic position rather than by some specific node ID.
Simulations have shown that geocasting does not scale in larger networks. Therefore, enhancements to geocast will be presented alongside with a new message system that is able to calculate the relevance of information, taking into accout distance from message origin, message age and message type. After receiving a message, the relevance is again calculated by the receiver. Network load then is reduced by sending only messages whose relevance is above a certain threshold. The threshold can be placed adaptively depending on network load. Thus, only information with a large prospected benefit will be kept inside the network. Open issues and problems such as storage and physical transport of information within vehicles, information renewal/invalidation and data aggregation will be addressed.

Session 3, Wed, 9 - 10.30

Selene Mota: Automated Posture Analysis for Detecting Learner’s Interest Level

A system for recognizing naturally occurring postures and associated affective states is presented. The affective states covered are those related to a child’s interest level while performing a learning task on a computer. Postures are gathered using two matrices of pressure sensors mounted on the seat and back of a chair. Subsequently, posture features are extracted using a mixture of four gaussians, and input to a 3-layer feedforward neural network. The neural network classifies nine postures in real time and achieves an overall accuracy of 87.6% when tested with postures coming from new subjects. A set of independent Hidden Markov Models (HMMs) is used to analyze temporal patterns among these posture sequences in order to determine three categories related to a child’s level of interest, as rated by human observers. The system reaches an overall performance of 82.3% with posture sequences coming from known subjects and 76.5% with unknown subjects.

Pasi Välkkynen: PPhysical Browsing

Physical browsing is a user interaction paradigm for associating physical objects with digital information in an ambient intelligence setting. In this presentation, I will review the research we have carried out at VTT Technical Research Centre of Finland. Our research evolves from designing ambient intelligence scenarios with physical browsing as one of the basic user interaction patterns. We have defined physical selection methods for pointing, touching and scanning and built a prototype system for studying the user interaction.

Urs Bischoff: Cooperative Artefacts

Cooperative Artefacts (CAs) are physical (everyday) objects instrumented with embedded computing, wireless communication, sensors and actuators. CAs can reason about the state of the world based on basic domain knowledge, observations and shared knowledge. They do not rely on any external infrastructure. This talk will present some scenarios and discuss a conceptual framework. Furthermore, it will address continuing and future research directions.

Clemens Holzmann: Orientation-Aware Artifacts

This talk deals with everyday artifacts which are aware of their spatial orientation. First, I will give a quick overview of the technological state-of-the-art and point out the potentials of orientation information for pervasive computing. I will then present some of our work in this field, which investigates the use of orientation-aware artifacts for gesture-based interaction.

Dikaios Papadogkonas: Solving the Navigation Problem in Ambient Intelligent Landscapes

People use maps and signing to guide in exhibition areas in order to find what mostly interests them. This usually doesnt work that well because of flaws at the navigation system. In order to investigate what is going wrong we need to know the exact trail of a number of users and by analyzing those trails we will be able to identify patterns and calculate statistics in order to improve it.

Kasper Hallenborg: UWB technology for location services

The UWB technology is a promising new approach for indoor positioning and smart spaces. It requires less instrumentation of the environment than most other technologies and has an accuracy of 15 cm for rather large cell-sizes (approx. 30m diameter). The talk will outline the principles of the technology and demonstrate an installation at our institute.

Session 4, Wed, 10.45 - 12.15

Paul Bowman: A wireless sensor network for telecare

In these days of ever decreasing resources – personnel, coupled with an aging population, the continuous assessment of personal well-being for the elderly will become harder to provide. BT, in collaboration with a number of Universities in the UK, have been engaged in a joint research project, with the DTI, to examine how a measure of well-being might be derived from data gathered from a sensor network in the home. To this end, a sensor network was designed and installed in two homes in the UK.

The presentation will briefly look at the motivation for this work, the constraints imposed and some insight into the preliminary conclusions.

Patrik Spiess: Executing Business Processes using Business Logic on Smart Items

To be able to compete in a globalized market, today's businesses rely heavily on automated, IT-backed intra and inter-company processes. These processes, modeled in the enterprise software, sometimes fail to match the real world situation because they are triggered by data keyed in by humans. We believe that using smart items equipped with sensors, actuators and business logic can mitigate the discrepancies between the real and the modeled process thereby increasing its fidelity. This is achieved by delegating execution of part of the business logic out of the backend into smart items, and feeding the remote results back to the enterprise system using a service-oriented architecture.

David Merrill: Invisible Media: Attention-sensitive informational augmentation for physical objects

In this talk I will be describing a new platform for creating attention-sensitive physical interaction scenarios. Our hardware/software platform supports the augmentation of physical objects with media and informational content in a way that can both inform, and respond to, the focus of a user’s visual attention and task context in a personalized, multimodal manner. I will also describe the design of, and recent features added to Engine-Info, an example application built on the platform that teaches the parts of a physical internal combustion engine.

Elisabetta Farella: Design and Implementation of WiMoCA Node for a Body Area Wireless Sensor Network

We present the design and implementation of a wireless sensor node for a Motion Capture system with Accelerometers (WiMoCA). It is composed by a tri-axial integrated accelerometer, a microcontroller and a wireless transceiver. The use of a single integrated tri-axial accelerometer allows to overcome inaccuracies of orthogonally mounted biaxial accelerometers and to build a smaller device compared to other sensors presented in literature. WiMoCA nodes have been exploited to build a Wireless Body Area Sensor Network (WBASN) that allows to implement a wireless/ wearable distributed gesture recognition system where nodes are mounted on many parts of the human body. We describe the hardware architecture and all the software layers supporting the recognition system. We also show characterization experiments on WiMoCA nodes that highlight how their performance and power consumption levels make them suitable to HCI applications.

Timo Vuorela: Energy consumption aware software optimization of embedded systems

As the functionality of portable devices is increasing, the energy consumption of these devices is reaching beyond the limit of batteries. In order to avoid the decrease of operating hours some ways to reduce the energy consumption are needed. One way is to optimize the energy consumption of executed software. In many research projects around the world it has been shown that different instructions in a microcontroller or in a microprocessor consume a different amount of energy. Therefore, savings in the energy consumption of portable or embedded applications can be gained through optimizing the written software according to the instruction level energy consumption. In this presentation some ways to reduce the energy consumption of Atmega8515 microcontroller are illustrated. The target of our research is to write an energy aware compiler, which can automatically optimize the software according to its energy consumption.

ETH ZurichLast updated June 20 2023 01:45:14 PM MET    kr