Daniel Minder

FlexCup: a flexible and efficient code update mechanism 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 adaptation of the sensor nodes but should also operate in an energy and time efficient manner. In this paper, we present FlexCup, a flexible code update mechanism that minimizes the energy consumed on each sensor node for the installation of arbitrary code changes. We describe two different versions of FlexCup and show, using a precise hardware emulator, that our mechanism is able to perform updates up to 8 times faster than related code update algorithms found in the literature, while consuming only an eighth of the energy.

TinyCubus: a flexible and adaptive framework sensor networks

With the proliferation of sensor networks and sensor network applications, the overall complexity of such systems is continuously increasing. Sensor networks are now heterogeneous in terms of their hardware characteristics and application requirements even within a single network. In addition, the requirements of currently supported applications are expected to change over time. All of this makes developing, deploying and optimizing sensor network applications an extremely difficult task. In this paper, we present the architecture of TinyCubus, a flexible and adaptive cross-layer framework for TinyOS-based sensor networks that aims at providing the necessary infrastructure to cope with the complexity of such systems. TinyCubus consists of a data management framework that selects and adapts both system and data management components, a cross-layer framework that enables optimizations through cross-layer interactions, and a configuration engine that installs components dynamically. Furthermore, we show the feasibility of our architecture by describing and evaluating a code distribution algorithm that uses application knowledge about the sensor topology in order to optimize its behavior.

Meeting lifetime goals with energy levels

In this paper we present Levels, a programming abstraction for energy-aware sensor network applications. Unlike most previous work it does not try to maximize network lifetime but rather helps to meet user-defined lifetime goals while maximizing application quality. Levels is targeted to applications where there is no redundancy and no node should fail early. With our programming abstraction the application developer defines so-called energy levels. These energy levels form a stack and can be deactivated from top to bottom if the lifetime goal cannot be met otherwise. Each code block within an energy level contains information about its energy consumption, which can be obtained from simulation tools without much effort. The runtime system then uses the data about the energy consumption of the different levels to compute an optimal level assignment for the time remaining. As we show in the evaluation, applications using Levels can accurately meet given lifetime goals and offer good application quality. In addition, the runtime overhead of our system is almost negligible.

Prototyping sensor-actuator networks for home automation

Integrating actuators into sensor networks is often considered to be the next logical step in the evolution of wireless sensor networks. However, few practical examples of such sensor and actuator networks have been demonstrated so far. In this paper, we present a prototype system that supports the easy prototyping of such applications in the area of home automation. We demonstrate the utility of this system with a simple light control application built on top of it. We also report first experiences and insights gained with the help of real-world experiments.

Removing the memory limitations of sensor networks with flash-based virtual memory

Virtual memory has been successfully used in different domains to extend the amount of memory available to applications. We have adapted this mechanism to sensor networks, where, traditionally, RAM is a severely constrained resource. In this paper we show that the overhead of virtual memory can be significantly reduced with compile-time optimizations to make it usable in practice, even with the resource limitations present in sensor networks. Our approach, ViMem, creates an efficient memory layout based on variable access traces obtained from simulation tools. This layout is optimized to the memory access patterns of the application and to the specific properties of the sensor network hardware. Our implementation is based on TinyOS. It includes a pre-compiler for nesC code that translates virtual memory accesses into calls of ViMems runtime component. ViMem uses flash memory as secondary storage. In order to evaluate our system we have modified nontrivial existing applications to make use of virtual memory. We show that its runtime overhead is small even for large data sizes.

TinyCubus: An Adaptive Cross-Layer Framework for Sensor Networks TinyCubus: Ein Adaptives Cross-Layer Framework für Sensornetze

Summury With the proliferation of sensor networks and sensor network applications, the overall complexity of such systems is continuously increasing. Sensor networks are now heterogeneous in terms of their hardware characteristics and application requirements even within a single network. In addition, the requirements of currently supported applications are expected to change over time. All of this makes developing, deploying, and optimizing sensor network applications an extremely difficult task. In this paper, we present the architecture of TinyCubus, a flexible and adaptive cross-layer framework for TinyOS-based sensor networks that aims at providing the necessary infrastructure to cope with the complexity of such systems. TinyCubus consists of a cross-layer framework that enables optimizations through cross-layer interactions, a configuration engine that distributes components efficiently by considering the roles of the sensor nodes and provides support to install components dynamically, and a data management framework that selects and adapts both system and data management components. Finally, relevant research challenges associated with the development of each framework are identified and discussed in the paper.

Adaptation and cross-layer issues in sensor networks

An intrinsic characteristic of current projects in the area of sensor networks is the heterogeneity of hardware and application requirements. In addition, the requirements of current applications are expected to change over time. This makes developing, deploying, and optimizing sensor network applications an extremely difficult task. In this paper, we present the architecture of TinyCubus, a flexible and adaptive cross-layer framework for TinyOS-based sensor networks that aims at providing the necessary infrastructure to cope with the complexity of such systems. TinyCubus consists of three parts: a data management framework that selects and adapts both system and data management components, a cross-layer framework that enables optimizations through cross-layer interactions, and a configuration engine that installs components dynamically. We show the feasibility of our architecture by describing and evaluating a code distribution algorithm that optimizes its behavior by using application knowledge about the sensor topology.

TinyXXL: Language and Runtime Support for Cross-Layer Interactions

In the area of wireless sensor networks, cross-layer interactions are often preferred to strictly layered architectures. However, architectural properties such as modularity and the reusability of components suffer from such optimizations. In this paper we present TinyXXL that provides programming abstractions for data exchange, a form of cross-layer interaction with a large potential for optimizations. Our approach decouples components providing and using data, and it allows for automatic optimizations of applications composed of reusable components. Its runtime representation is efficient regarding memory consumption and processing overhead

An Analysis of Cross-Layer Interactions in Sensor Network Applications

In the field of sensor networks cross-layer interactions are favored over strict layering of components and regarded as a way to provide the optimization capabilities required by sensor network applications. Despite their importance, developers tend to devise specific solutions for the application at hand, instead of designing more general primitives that can be used across applications. The contribution of this paper is twofold: We analyze several typical sensor network applications and provide a classification of the types of cross-layer interactions found in their code. Based on this classification, we propose TinyXXL, an extension to the nesC language that defines primitives for seamless cross-layer data exchange.

Versatile support for efficient neighborhood data sharing

Many applications in wireless sensor networks rely on data from neighboring nodes. However, the effort for developing efficient solutions for sharing data in the neighborhood is often substantial. Therefore, we present a general-purpose algorithm for this task that makes use of the broadcast nature of radio transmission to reduce the number of packets. We have integrated this algorithm into TinyXXL, a programming language extension for data exchange. This combined system offers seamless support both for data exchange among the components of a single node and for efficient neighborhood data sharing. We show that compared to existing solutions, such as Hood, our approach further reduces the work of the application developer and provides greater efficiency.