The development of a reliable large-scale wireless sensor networks (WSNs) is very difficult because of their stringent resource constraints, harsh energy budget, and demanding application requirements. We identify that three OS features – OS protection, virtual memory, and preemptive scheduling – will significantly improve the reliability of WSN systems and facilitate developing complex WSN software. However, due to the limitation of hardware, it is impossible to implement these features with traditional OS design techniques. To solve this problem, we design a new OS kernel, the tkernel, to perform extensive load-time code modification and enhance the system abstraction visible to programmers. After the modification, the application and OS work in a collaborative way supporting the aforementioned features. Having implemented the t-kernel on MICA2 motes with an 8-bit processor and 4KB RAM, we evaluate its performance by measuring the overhead and execution speed. We analyze the CPU utilization in sensor network applications, and verify that, though CPU-bound computation tasks may slow down 0.5–4 times, the performance of applications under typical workloads does not degrade. The t-kernel significantly enhances developers ’ ability to design sophisticated applications and protects WSNs from accidental programming errors. To the authors ’ best knowledge, the t-kernel is unique in the follow ways: it performs efficient binary translation on highly resource constrained sensor nodes with only 4KB RAM, it provides software based virtual memory without repeatedly writable swapping devices, and it protects OS from application error without memory protection or privileged execution hardware. 1

Recent gains in energy-efficiency of new-generation NAND flash storage have strengthened the case for in-network storage by data-centric sensor network applications. This paper argues that a simple file system abstraction is inadequate for realizing the full benefits of high-capacity low-power NAND flash storage in data-centric applications. Instead we advocate a rich object storage abstraction to support flexible use of the storage system for a variety of application needs and one that is specifically optimized for memory and energy-constrained sensor platforms. We propose Capsule, an energy-optimized log-structured object storage system for flash memories that enables sensor applications to exploit storage resources in a multitude of ways. Capsule employs a hardware abstraction layer that hides the vagaries of flash memories for the application and supports energy-optimized implementations of commonly used storage objects such as streams, files, arrays, queues and lists. Further, Capsule supports checkpointing and rollback of object states to tolerate software faults in sensor applications running on inexpensive, unreliable hardware. Our experiments demonstrate that Capsule provides platform-independence, greater functionality, more tunability, and greater energy-efficiency than existing sensor storage solutions, while operating even within the memory constraints of the Mica2 Mote. Our experiments not only demonstrate the energy and memory-efficiency of I/O operations in Capsule but also shows that Capsule consumes less than 15% of the total energy cost in a typical sensor application.

Abstract. We introduce and study the minimum-backlog problem (MBP). The MBP arises in sensor networks and is related to the classic k-server problem. It can be understood as a 2-person game played on a graph G = (V, E). The “player ” moves along the edges of the graph; the opponent is the “adversary. ” The game proceeds in timesteps. In each timestep the adversary pours a total of one unit of water into “cups ” that are located on the vertices of the graph, arbitrarily distributing the water among the cups. The player then moves from her current vertex to an adjacent vertex and empties the cup at that vertex. The player’s objective is to minimize the maximum amount of water (the backlog) in any cup at any time. We show that the competitive ratio of any algorithm for the MBP has a lower bound of Ω(∆), where ∆ is the diameter of the graph. Thus, we focus on determining a strategy for the player that guarantees a uniform upper bound on the backlog. In general graphs, the deamortization analysis of Dietz and Sleator gives a bound of O( ∆ ln |V |). Our main result is that in geometric settings (e.g., sensor fields), one can obtain substantially better bounds on the maximum backlog. In particular, for a 2-dimensional n-by-n grid, we achieve a backlog of O(n √ ln ln n), improving the O(n ln n) upper bound for general graphs, and coming close to the naive Ω(n) lower bound. Then, in a model of continuous motion of the player and continuous pouring by the adversary, for cups placed at m points in the plane we show that the backlog can be bounded by O(D √ ln ln m), where D is the diameter of the point set. Our methods apply also to higher (fixed) dimensions. We study also the variant of the MBP in which the adversary has a location within the graph and must act locally (filling cups) with respect to his position, just as the player acts locally (emptying cups) with respect to her position. We prove that deciding the value of this game is PSPACE-hard.

This paper presents PRESTO, a novel two-tier sensor data management architecture comprising proxies and sensors that cooperate with one another for acquiring data and processing queries. PRESTO proxies construct time-series models of observed trends in the sensor data and transmit the parameters of the model to sensors. Sensors check sensed data with model-predicted values and transmit only deviations from the predictions back to the proxy. Such a model-driven push approach is energyefficient, while ensuring that anomalous data trends are never missed. In addition to supporting queries on current data, PRESTO also supports queries on historical data using interpolation and local archival at sensors. PRESTO can adapt model and system parameters to data and query dynamics to further extract energy savings. We have implemented PRESTO on a sensor testbed comprising Intel Stargates and Telos Motes. Our experiments show that in a temperature monitoring application, PRESTO yields one to two orders of magnitude reduction in energy requirements over on-demand, proactive or model-driven pull approaches. PRESTO also results in an order of magnitude reduction in query latency in a 1 % duty-cycled five hop sensor network over a system that forwards all queries to remote sensor nodes.

Energy management is a critical concern in wireless sensornets. Despite its importance, sensor network operating systems today provide minimal energy management support, requiring applications to explicitly manage system power states. To address this problem, we present ICEM, a device driver architecture that enables simple, energy efficient wireless sensornet applications. The key insight behind ICEM is that the most valuable information an application can give the OS for energy management is its concurrency. Using ICEM, a low-rate sensing application requires only a single line of energy management code and has an efficiency within 1.6 % of a hand-tuned implementation. ICEM’s effectiveness questions the assumption that sensornet applications must be responsible for all power management and sensornets cannot have a standardized OS with a simple API.

Recent advances in sensor networks permit the use of a large number of relatively inexpensive distributed computational nodes with camera sensors linked in a network and possibly linked to one or more central servers. We argue that the full potential of such a distributed system can be realized if it is designed as a distributed search engine where images from different sensors can be captured, stored, searched and queried. However, unlike traditional image search engines that are focused on resource-rich situations, the resource limitations of camera sensor networks in terms of energy, bandwidth, computational power, and memory capacity present significant challenges. In this paper, we describe the design and implementation of a distributed search system over a camera sensor network where each node is a search engine that senses, stores and searches information. Our work involves innovation at many levels including local storage, local search, and distributed search, all of which are designed to be efficient under the resource constraints of sensor networks. We present an implementation of the search engine on a network of iMote2 sensor nodes equipped with low-power cameras and extended flash storage. We evaluate our system for a dataset comprising book images, and demonstrate more than two orders of magnitude reduction in the amount of data communicated and up to 5x reduction in overall energy consumption over alternate techniques.

This paper presents the design, implementation, and evaluation of EnviroMic, a novel distributed acoustic monitoring, storage, and trace retrieval system. Audio represents one of the least exploited modalities in sensor networks to date. The relatively high frequency and large size of audio traces motivate distributed algorithms for coordinating recording tasks, reducing redundancy of data stored by nearby sensors, filtering out silence, and balancing storage utilization in the network. Applications of acoustic monitoring with EnviroMic range from the study of mating rituals and social behavior of animals in the wild to audio surveillance of military targets. EnviroMic is designed for disconnected operation, where the luxury of having a basestation cannot be assumed. We implement the system on a TinyOS-based platform and systematically evaluate its performance through both indoor testbed experiments and a preliminary outdoor deployment. Results demonstrate up to a 4-fold improvement in effective storage capacity of the network compared to uncoordinated recording. Index Terms Sensor networks, applications, acoustics, distributed storage, group management

Reports of NAND flash device testing in the literature have for the most part been limited to examination of circuit-level parameters on raw flash chips or prototypes, and systemlevel parameters on entire storage subsystems. However, there has been little examination of system-level parameters of raw devices, such as mean latency and endurance values. We report the results of such tests on a variety of devices. Read, program, and erase latency were found to align closely with manufacturer’s specified “typical ” values in almost all cases. Program/erase endurance, however, was found to exceed specified minimum values, often by as much as a factor of 100. In addition significant performance changes were found with wear. These changes may be used to track wear, and in addition have significant implications for system performance over the lifespan of a device. Finally, random write patterns which incur performance penalties on current flashbased memory systems were found to incur no overhead on the devices themselves. 1.

Persistent storage offers multiple advantages for sensor networks, yet the available storage systems have been unwieldy because of their complexity and device-specific designs. We present the Coffee file system for flashbased sensor devices. Coffee provides a programming interface for building efficient and portable storage abstractions. Unlike previous flash file systems, Coffee uses a small and constant RAM footprint per file, making it scale elegantly with workloads consisting of large files or many files. In addition, the performance overhead of Coffee is low: the throughput is at least 92 % of the achievable direct flash driver throughput. We show that network layer components such as routing tables and packet queues can be implemented on top of Coffee, leading to increased performance and reduced memory requirements for routing and transport protocols.

Abstract. In this paper, we present Microsearch, a search system suitable for small devices used in ubiquitous computing environments. Akin to a desktop search engine, Microsearch indexes the information inside a small device, and accurately resolves user queries. Given the very limited hardware resources, conventional search engine designs and algorithms cannot be used. We adopt information retrieval techniques for query resolution, and propose a space efficient algorithm to perform top-k query on limited hardware resources. Finally, we present a theoretical model of Microsearch to better understand the tradeoffs in system design parameters. By implementing Microsearch on actual hardware for evaluation, we demonstrate the feasibility of scaling down information retrieval systems onto very small devices. 1