Nirupama Bulusu

Ear-phone: an end-to-end participatory urban noise mapping system

A noise map facilitates monitoring of environmental noise pollution in urban areas. It can raise citizen awareness of noise pollution levels, and aid in the development of mitigation strategies to cope with the adverse effects. However, state-of-the-art techniques for rendering noise maps in urban areas are expensive and rarely updated (months or even years), as they rely on population and traffic models rather than on real data. Participatory urban sensing can be leveraged to create an open and inexpensive platform for rendering up-to-date noise maps. In this paper, we present the design, implementation and performance evaluation of an end-to-end participatory urban noise mapping system called Ear-Phone. Ear-Phone, for the first time, leverages Compressive Sensing to address the fundamental problem of recovering the noise map from incomplete and random samples obtained by crowdsourcing data collection. Ear-Phone, implemented on Nokia N95 and HP iPAQ mobile devices, also addresses the challenge of collecting accurate noise pollution readings at a mobile device. Extensive simulations and outdoor experiments demonstrate that Ear-Phone is a feasible platform to assess noise pollution, incurring reasonable system resource consumption at mobile devices and providing high reconstruction accuracy of the noise map.

Adaptive beacon placement

Beacon placement strongly affects the quality of spatial localization, a critical service for context-aware applications in wireless sensor networks; yet this aspect of localization has received little attention. Fixed beacon placement approaches such as uniform and very dense placement are not always viable and will be inadequate in very noisy environments in which sensor networks may be expected to operate (with high terrain and propagation uncertainties). We motivate the need for empirically adaptive beacon placement and outline a general approach based on exploration and instrumentation of the terrain conditions by a mobile human or robot agent. We design, evaluate and analyze three novel adaptive beacon placement algorithms using this approach for localization based on RF-proximity. In our evaluation, we find that beacon density rather than noise level has a more significant impact on beacon placement algorithms. Our beacon placement algorithms are applicable to a low (beacon) density regime of operation. Noise makes moderate density regimes more improvable.

Node localization using mobile robots in delay-tolerant sensor networks

We present a novel scheme for node localization in a delay-tolerant sensor network (DTN). In a DTN, sensor devices are often organized in network clusters that may be mutually disconnected. Some mobile robots may be used to collect data from the network clusters. The key idea in our scheme is to use this robot to perform location estimation for the sensor nodes it passes based on the signal strength of the radio messages received from them. Thus, we eliminate the processing constraints of static sensor nodes and the need for static reference beacons. Our mathematical contribution is the use of a robust extended Kalman filter (REKF)-based state estimator to solve the localization. Compared to the standard extended Kalman filter, REKF is computationally efficient and also more robust. Finally, we have implemented our localization scheme on a hybrid sensor network test bed and show that it can achieve node localization accuracy within 1 m in a large indoor setting.

Self-configuring localization systems: Design and Experimental Evaluation

Embedded networked sensors promise to revolutionize the way we interact with our physical environment and require scalable, ad hoc deployable and energy-efficient node localization/positioning.This paper describes the motivation, design, implementation, and experimental evaluation (on sharply resource-constrained devices) of a self-configuring localization system using radio beacons. We identify beacon density as an important parameter in determining localization quality, which saturates at a transition density. We develop algorithms to improve localization quality by (i) automating placement of new beacons at low densities (HEAP) and (ii) rotating functionality among redundant beacons while increasing system lifetime at high densities (STROBE).

The design and evaluation of a hybrid sensor network for Cane-Toad monitoring

This paper investigates a wireless, acoustic sensor network application --- monitoring amphibian populations in the monsoonal woodlands of northern Australia. Our goal is to use automatic recognition of animal vocalizations to census the populations of native frogs and the invasive introduced species, the Cane Toad (see Fig. 1). This is a challenging application because it requires high frequency acoustic sampling, complex signal processing and wide area sensing coverage.We set up two prototypes of wireless sensor networks that recognize vocalizations of up to 9 frog species found in northern Australia. Our first prototype is simple and consists of only resource-rich Stargate devices. Our second prototype is more complex and consists of a hybrid mixture of Stargates and inexpensive, resource-poor Mica2 devices operating in concert. In the hybrid system, the Mica2s are used to collect acoustic samples, and expand the sensor network coverage. The Stargates are used for resource-intensive tasks such as Fast Fourier Transforms (FFTs) and machine learning.The hybrid system incorporates three algorithms designed to account for the sampling, processing and communication bottlenecks of the Mica2s (i) high frequency sampling, (ii) compression and noise reduction, to reduce data transmission by up to 90%, and (iii) sampling scheduling, which exploits the sensor network redundancy to increase the effective sample processing rate.We evaluate the performance of both systems over a range of scenarios, and demonstrate that the feasibility and benefits of a hybrid systems approach justify the additional system complexity.

Design and evaluation of a hybrid sensor network for cane toad monitoring

This paper investigates a wireless, acoustic sensor network application-monitoring amphibian populations in the monsoonal woodlands of northern Australia. Our goal is to use automatic recognition of animal vocalizations to census the populations of native frogs and the invasive introduced species, the cane toad. This is a challenging application because it requires high frequency acoustic sampling, complex signal processing and wide area sensing coverage. We set up two prototypes of wireless sensor networks that recognize vocalizations of up to 9 frog species found in northern Australia. Our first prototype is simple and consists of only resource-rich Stargate devices. Our second prototype is more complex and consists of a hybrid mixture of Stargates and inexpensive, resource-poor Mica2 devices operating in concert. In the hybrid system, the Mica2s are used to collect acoustic samples, and expand the sensor network coverage. The Stargates are used for resource-intensive tasks such as fast Fourier transforms (FFTs) and machine learning. The hybrid system incorporates three algorithms designed to account for the sampling, processing and communication bottlenecks of the Mica2s (i) high frequency sampling, (ii) compression and noise reduction, to reduce data transmission by up to 90%, and (iii) sampling scheduling, which exploits the sensor network redundancy to increase the effective sample processing rate. We evaluate the performance of both systems over a range of scenarios, and demonstrate that the feasibility and benefits of a hybrid systems approach justify the additional system complexity.

A communication paradigm for hybrid sensor/actuator networks

This paper investigates an anycast communication service for a hybrid sensor/actuator network, consisting of both resource-rich and resource-impoverished devices. The key idea is to exploit the capabilities of resource-rich devices (called micro-servers) to reduce the communication burden on smaller, energy, bandwidth and memory constrained sensor nodes. The goal is to deliver sensor data to the nearest micro-server, which can (i) store it (ii) forward it to other micro-servers using out-of-band communication or (iii) perform the desired actuation. We propose and evaluate a reverse tree-based anycast mechanism tailored to deal with the unique event dynamics in sensor networks. Our approach is to construct an anycast tree rooted at each potential event source, which micro-servers can dynamically join and leave. Our anycast mechanism is self-organizing, distributed, robust, scalable, routing-protocol independent and incurs very little overhead. Simulations using Network Simulator (ns-2) show that: our anycast mechanism when added to Directed Diffusion can reduce the network’s energy consumption by more than 50%; can reduce both the mean end-to-end latency of the transmission and the mean number of transmissions by more than 50%; achieves 99% data delivery rate for low and moderate micro-server mobility rate; and handles network dynamics reasonably well.

Automatic Collection of Fuel Prices from a Network of Mobile Cameras

It is an undeniable fact that people want information. Unfortunately, even in todays highly automated society, a lot of the information we desire is still manually collected. An example is fuel prices where websites providing fuel price information either send their workers out to manually collect the prices or depend on volunteers manually relaying the information. This paper proposes a novel application of wireless sensor networks to automatically collect fuel prices from camera images of road-side price board (billboard) of service (or gas) stations. Our system exploits the ubiquity of mobile phones that have cameras as well as users contributing and sharing data. In our proposed system, cameras of contributing users will be automatically triggered when they get close to a service station. These images will then be processed by computer vision algorithms to extract the fuel prices. In this paper, we will describe the system architecture and present results from our computer vision algorithms. Based on 52 images, our system achieves a hit rate of 92.3% for correctly detecting the fuel price board from the image background and reads the prices correctly in 87.7% of them. To the best of our knowledge, this is the first instance of a sensor network being used for collecting consumer pricing information.

DHV: A Code Consistency Maintenance Protocol for Multi-hop Wireless Sensor Networks

Ensuring that every sensor node has the same code version is challenging in dynamic, unreliable multi-hop sensor networks. When nodes have different code versions, the network may not behave as intended, wasting time and energy. We propose and evaluate DHV, an efficient code consistency maintenance protocol to ensure that every node in a network will eventually have the same code. DHV is based on the simple observation that if two code versions are different, their corresponding version numbers often differ in only a few least significant bits of their binary representation. DHV allows nodes to carefully select and transmit only necessary bit level information to detect a newer code version in the network. DHV can detect and identify version differences in O (1) messages and latency compared to the logarithmic scale of current protocols. Simulations and experiments on a real MicaZ testbed show that DHV reduces the number of messages by 50%, converges in half the time, and reduces the number of bits transmitted by 40-60% compared to DIP, the state-of-the-art protocol.

A performance comparison of data dissemination protocols for wireless sensor networks

This paper reports on the results of a simulation comparison made by an independent researcher using the ns-2.26 simulator for the WSN protocols: directed diffusion (DD), two-tier data dissemination (TTDD), and gradient broadcast (GRAB). Our performance study provides useful insights for the network designer - such as which protocols (and design choices) control the traffic well, improve data delivery or reduce overall energy consumption. We observe and identify a number of statically configured parameters for each protocol which reduce performance and suggest how they should be dynamically configured in response to measured network state.