In sensor networks, the location of a sensor making an observation is a vital piece of information to allow accurate data analysis. GPS is an established technology to enable precise position information. Yet, resource constraints and size issues prohibit its use in small sensor nodes that are designed to be cost efficient. Instead, most positions are estimated by a number of algorithms. To date, however, the focus was on individual accuracy of each sensor's position estimates in isolation to the complete network. In this paper, we propose a novel positioning algorithm called nQUAD to enable representative topology information based on relative accuracy. nQUAD makes use of relative distances to determine the quadrant it resides in and refines its estimation according to neighbor provided information and reports a level of certainity along with estimates. Our experiment results suggest significant improvements in individual accuracy in comparison to lateration based alternatives. At top of all, drastic improvements are achieved in the overall topology to enable accurate analysis of data.

A major concern in sensor networks is to reduce the data traffic and the energy consumption to extend the lifetime of the network. In this regard, radio transmission is the dominant factor in energy consumption. Therefore, the routing policy has a significant impact on the overall lifespan of the sensor network. For instance, pure multi-hop routing, e.g., tree-based routing, has the drawback that nodes closest to the data collection point (sink node) consume their energy and die first. In this paper, we propose an opportunistic routing mechanism that aims to extend network lifespan by balancing energy consumption. Our approach distributes the power consumption in the network by altering the communication range when applicable. Our experiment results suggest significant improvements when compared to tree-based routing pure multihoping and LEACH based routing.

While sensor networks continue to attract significant interest in various research communities, high impact applications still have a long list of challenges to be addressed. An individual sensor system can provide important observations within a local area. However, local observations alone are not sufficient for some applications that require planetary scale coverage. Monitoring volcanic activity, nuclear disasters, magnetic field changes, migration patterns of species, pandemic disease spread patterns are some examples to such applications. These applications require a close interaction between different sensor networks with in-situ and remotely sensed observations. In this paper we describe our PLASMA (PLAnetary Scale Monitoring Architecture) project to motivate the challenges that need to be addressed at such scale. These include approximations in spatiotemporal attributes due to resource constraints and also multi-attribute visualization to enable a real-time user interface to the system.

Advances in wireless technology have resulted in an emerging broadcast-capable infrastructure. A major debate for such infrastructures is whether we should use push or pull to support large client populations. To date, push was suggested to provide a scalability that is not possible with pull. In this paper we conclude otherwise, based on simulation and experimental prototype results. A major contribution of this paper is to show that pull-based algorithms outperform push-based ones. Contrary to the conventional wisdom, this is true even for environments where request processing and scheduling overheads are a significant factor.

In this paper we study a power-efficient event update dissemination system in a data broadcast environment for time-critical observations. We propose an on-line algorithm that can produce efficient broadcast schedules to disseminate event updates in linear time. Our algorithm assigns clients to a particular channel from where they can monitor updates of interest to reduce the channel monitoring energy consumption. Our simulation results suggest that our algorithm drastically improves power efficiency compared to previous algorithms.

Sensors have a constrained lifetime for various reasons such as limited battery power, malfunctioning components, or deactivation due to natural incidents. Therefore, there is a continuous need to deploy new sensors or to change the current topology of the sensors to ensure the functionality of the network. In this paper, we study on-demand communication mechanisms between smart and reconfigurable sensors. We present an automatic registration protocol for publish/subscribe wireless sensor deployments which allows a sensor to register with neighboring sensors upon activation, creates a schedule online without any initial global knowledge of the topology, limits the amount of transmissions made by a sensor, and permits modification to subscription patterns. Our simulation results suggest that our protocol significantly decreases the energy consumption while providing a competitive scheduling in comparison to the offline greedy vertex-coloring algorithm.

Broadcast is becoming an increasingly attractive data-dissemination method for large client populations. In order to effectively utilize a broadcast medium for such a service, it is necessary to have efficient on-line scheduling algorithms that can balance individual and overall performance and can scale in terms of data set sizes, client populations, and broadcast bandwidth. We propose an algorithm, called R/spl times/W, that provides good performance across all of these criteria and can be tuned to trade off average and worst-case waiting time. Unlike previous work on low overhead scheduling, the algorithm does not use estimates of the access probabilities of items, but rather, it makes scheduling decisions based on the current queue state, allowing it to easily adapt to changes in the intensity and distribution of the workload. We demonstrate the performance advantages of the algorithm under a range of scenarios using a simulation model and present analytical results that describe the intrinsic behavior of the algorithm.

Advances in telecommunications have enabled the deployment of broadcast-based wide-area information services that provide on-demand data access to very large client populations. In order to effectively utilize a broadcast medium for such a service, it is necessary to have efficient, on-line scheduling algorithms that can balance individual and overall performance and can scale in terms of data set sizes, client populations, and broadcast bandwidth. In this study we introduce a parameterized algorithm that provides good performance across all of these criteria and can be tuned to emphasize either average or worst case waiting time. Unlike previous work on low overhead scheduling, the algorithm is not based on estimates of the access probabilities of items, but rather, it makes scheduling decisions based on the current queue state, allowing it to easily adapt to changes in the intensity and distribution of the workload. We examine the performance of the algorithm using a simulation model.