Volker Turau

Cross-Platform Development Tools for Smartphone Applications

Developers can use cross-platform mobile development tools to create smartphone apps that meet user expectations, but existing XMTs need improvement.

BGR: blind geographic routing for sensor networks

This paper introduces BGR, a new beacon-less geographic routing algorithm for wireless sensor networks. Data packets are forwarded toward the destination, and nodes, which hear the packet compete for becoming the next hop. A recovery strategy is provided for the case of empty forwarding areas. The main innovation is a strategy to avoid simultaneous forwarding situations, which would otherwise cause packet failures. It is confirmed by simulation that BGR sends very few packets and is reliable infields with sufficiently high node density. Furthermore, BGR is compared with similar algorithms after developing a general framework for beacon-less algorithms with an integrative terminology. Different kinds of forwarding areas are compared.

Multicasting multimedia streams with active networks

Active networks allow protocol processing code to be loaded dynamically into network nodes at run-time. This code can perform tasks specific to a stream of packets or even a single packet. In this paper we compare two active network architectures: the active node transfer system (ANTS) and the messenger system (M0). We have implemented a robust audio multicast protocol and a layered video multicast protocol with both active network systems. We discuss the differences of the two systems, evaluate architectural strengths and weaknesses, compare the runtime performance, and report practical experience and lessons learned.

Prediction accuracy of link-quality estimators

The accuracy of link-quality estimators (LQE) is missioncritical in many application scenarios in wireless sensor networks (WSN), since the link-quality metric is used for routing decisions or neighborhood formation. Link-quality estimation must offer validity for different timescales. Existing LQEs describe and approximate the current quality in a single value only. This method leads to a limited accuracy and expressiveness about the presumed future behavior of a link. The LQE developed in this paper incorporates four quality metrics that give a holistic assessment of the link and its dynamic behavior; therefore, this research is an important step to achieving a higher prediction accuracy including knowledge about the short- and long-term behavior.

Linear self-stabilizing algorithms for the independent and dominating set problems using an unfair distributed scheduler

This paper presents distributed self-stabilizing algorithms for the maximal independent and the minimal dominating set problems. Using an unfair distributed scheduler the algorithms stabilizes in at most max{3n-5,2n} resp. 9n moves. All previously known algorithms required O(n^2) moves.

A framework for automatic generation of web-based data entry applications based on XML

This paper presents a framework for web-based data entry applications. It introduces a method for the conceptional and the navigational design based on a textual specification in the form of an XML-application. This forms the input to a code generation environment allowing for real automated prototyping. The environment produces fully functional skeletons for the web pages. Together with the framework classes they can be utilized for testing and for requirements review. They also form the starting point for the work of the presentation design. The main advantage of the framework is a clear separation between the presentation and the business logic. This allows work on each aspect to proceed in parallel along relatively independent but cooperating tracks. The framework has been implemented using Java Servlets and Java Server Pages.

Efficient Dynamic Look-Up Strategy for Multi-Methods

In object-oriented programming languages, multiple-dispatching provides increased expressive power over single-dispatching by guiding method look-up using the values of all arguments instead of the receiver only. There have been several programming languages supporting this mechanism and they demonstrate its usefulness. However, efficient implementation of multi-methods is still critical with regard to its success as a standard. In this paper, we present a new mechanism for implementing multi-methods dynamic lookup based on automaton techniques. Analysis and experimental results show that our strategy is time and space efficient. The presented result can provide the basis for designing new object-oriented paradigms based on multi-methods.

Perpetual Data Collection with Energy-Harvesting Sensor Networks

A sustainable, uniform, and utility-maximizing operation of energy-harvesting sensor networks requires methods for aligning consumption with harvest. This article presents a lightweight algorithm for online load adaptation of energy-harvesting sensor nodes using supercapacitors as energy buffers. The algorithm capitalizes on the elementary relationship between state of charge and voltage that is characteristic for supercapacitors. It is particularly designed to handle the nonlinear system model, and it is lightweight enough to run on low-power sensor node hardware. We define two energy policies, evaluate their performance using real-world solar-harvesting traces, and analyze the influence of the supercapacitor’s capacity and imprecisions in harvest forecasts. To show the practical merit of our algorithm, we devise a load adaptation scheme for multihop data collection sensor networks and run a 4-week field test. The results show that (i) choosing a duty cycle a priori is infeasible, (ii) our algorithm increases the achievable work load of a node when using forecasts, (iii) uniform and steady operation is achieved, and (iv) depletion can be prevented in most cases.

Efficient transformation of distance-2 self-stabilizing algorithms

Self-stabilizing algorithms for optimization problems can often be solved more easily using the distance-two model in which each vertex can instantly see the state information of all vertices up to distance two. This paper presents a new technique to emulate algorithms for the distance-two model on the distance-one model using the distributed scheduler with a slowdown factor of O(m) moves. Up until now the best transformer had a slowdown factor of O(n^2m) moves. The technique is used to derive improved self-stabilizing algorithms for several graph domination problems. The paper also introduces a generalization of the distance-two model allowing a more space efficient transformer.

Towards ubiquitous indoor location based services and indoor navigation

Outdoor navigation services have become ubiquitously available due to small handheld devices such as GPS enabled mobile phones or dedicated mobile navigation systems. Two main drivers were necessary in order to provide widespread location based services: Acquiring positioning information with a certain degree of precision and widespread (mobile) access to computer networks to use the position information with a large information basis such as the Internet. Envisioned is a future, where indoor navigation and location based services are used as naturally as outdoor location based services are now. To achieve this long term vision where users seamlessly navigate for example from work desk to departure gate at the airport and use location based services on the way, various challenges have to be solved. In this paper, these challenges and open issues are discussed. The paper proposes an architecture which abstracts from different mobile devices and localization technologies. Additionally, it sketches the use of indoor topology information to increase the accuracy of indoor localization. These two presented concepts form a basis for a transition period until standards for ubiquitous indoor location based services have emerged.