A new system for controlled testing of sensor network applications: architecture, prototype and experimental evaluation
Section snippets
Introduction and related work
Sensor network application software needs to achieve acceptable performance while operating with limited energy and other scarce resources available on each node. Application software is usually comprised of a set of algorithms, protocols, data structures, etc. Since there are multiple algorithms and protocols that provide similar functionality, various versions of application software (hereafter referred to as an application) that serve the same purpose but operate differently can be built. To
Motivation
The primary purpose of a wireless sensor network is to collect and process data. We can, therefore, consider the network as a collective computing machine that runs an application that operates on an input or stimuli and produces some output , which can be represented as:
Consider a scenario where we are to evaluate two slightly different applications, say and , and pick a winner depending on some property of the output.
Also consider two separate experiments that have been
Proposed architecture
The proposed architecture allows testing sensor network applications by providing stimuli to and allowing control over the operation of each sensor node in the network. The stimuli is provided by another sensor node (hereafter referred to as the control node), which generates the stimuli and routes it and control signals to the sensor node through a wired interface. When testing an application on real hardware, the sensor nodes are usually deployed in a large area to match the desired topology
Prototype
The design goal of building a prototype was to use hardware and software that had open architectures and implement the stimuli system using the fewest and simplest components possible. We built the CSPs using MICAz [15] nodes from Crossbow Technologies Inc. and the open source TinyOS [16] operating system. TinyOS was used to build a simple test application and the software for the control node. A photograph of the prototype highlighting the pins relevant to stimuli generation and routing is
Evaluation
To evaluate the stimuli system, we designed experiments to test the accuracy of stimuli generation and the operation of CSPs as a network. Exp. 1 tests the accuracy of the system in recreating a pre-recorded set of values and Exp. 2 tests the accuracy of generating an arbitrary set of values. Exp. 3 tests the accuracy of the system in recreating values that are read as a stream by the sensor node (such as an accelerometer). To evaluate the control and measurement capabilities, we designed a
Control data dissemination
In this section we first show how the problem of disseminating control packets through the control network is a facility location problem. We then introduce the variables needed to formulate the problem and include an example. Later we introduce the problem objective and constraints and also show solutions for some simple instances of the problem.
Conclusion and future work
In this paper we presented an architecture which pairs two sensor nodes. One node can provide the stimuli and exercise control over the other node which runs the test application. This architecture combines the benefits of using hardware-based stimuli for fidelity and software-based programmability for ease of control.
A prototype implementation of our architecture for generating analog stimuli was also presented. The prototype was evaluated as a networked system, presenting experimental results
Acknowledgements
This work was supported in part by NSF Award CNS-0626548. Any opinions, findings, and conclusions or recommendations expressed in this material are those of the author(s) and do not necessarily reflect the views of the National Science Foundation.
We would like to thank Jeffrey W. Wildman II, graduate student in the Department of Electrical and Computer Engineering, College of Engineering, Drexel University for his input during the design of the stimuli system. We also extend our thanks to Dr.
Anbu Elancheziyan is a Ph.D. candidate in Department of Electrical and Computer Engineering at Drexel University. His research interests include developing architectures for distributed embedded systems and their implementation, simulating network protocols and analysis of large scale networks.
References (25)
- et al.
Dynamic multi-commodity capacitated facility location: a mathematical modeling framework for strategic supply chain planning
Computers and Operations Research
(2006) - P. Levis, N. Lee, M. Welsh, D. Culler, TOSSIM: accurate and scalable simulation of entire TinyOS applications, in:...
- L. Girod, T. Stathopoulos, N. Ramanathan, J. Elson, D. Estrin, E. Osterweil, T. Schoellhammer, A system for simulation,...
- L. Luo, T. He, G. Zhou, L. Gu, T.F. Abdelzaher, J.A. Stankovic, Achieving repeatability of asynchronous events in...
- T. Clouser, R. Thomas, M. Nesterenko, Emuli: Emulated Stimuli for Wireless Sensor Network Experimentation, Tech. Rep.,...
- et al.
MoteLab: a wireless sensor network testbed
- et al.
Kansei: a high-fidelity sensing testbed
IEEE Internet Computing
(2006) - et al.
EmPro: an environment/energy emulation and profiling platform for wireless sensor networks
SECON ’06
(2006) - et al.
Deployment support network – a toolkit for the development of WSNs
- et al.
Next-generation prototyping of sensor networks
A local facility location algorithm for large-scale distributed systems
Journal of Grid Computing
Cited by (8)
Nonlinear vibration energy harvester: Design and oscillating stability analyses
2019, Mechanical Systems and Signal ProcessingCitation Excerpt :With new concepts like the Industry 4.0 or the Internet of Things, which bring about the idea of multiple independent electronic units and sensors cooperating with intelligent processing algorithms, also comes the question of their energy supply. Energy harvesting has become more important in the last decade as a possible power source for the so-called wireless sensor networks (WSNs) [1], which can be applied to environment monitoring [2], structural health monitoring [3], or the IoT in general [4]. Compared to other possible sources, e.g. batteries, energy harvesting offers longer operation time without maintenance.
A multifunctional road-compatible piezoelectric energy harvester for autonomous driver-assist LED indicators with a self-monitoring system
2019, Applied EnergyCitation Excerpt :Piezoelectric energy-harvesting technology has attracted much attention because of its high energy density and its use of mechanical energy [10–12]. A piezoelectric energy harvester installed on an actual road generates electricity from the axial loads of vehicles moving along the road and can power safety-related structures in conjunction with a wireless sensor network [13–15]. However, as is the case for most actual road energy harvesters, the output power is not sufficient [16,17].
Feasibility study of impact-based piezoelectric road energy harvester for wireless sensor networks in smart highways
2017, Sensors and Actuators, A: PhysicalCitation Excerpt :The energy harvesting technology related to roads including bridge has been particularly spotlighted because of its advantages such as applicability and high input energy [19–22]. Moreover, it can be used as a power source for wireless sensor networks [23–25] in smart highways, traffic control [26], structural monitoring [27–29]. A piezoelectric road energy harvester is buried beneath the road, and energy is generated using the axial load from the vehicle moving on the road, as the input.
Design and optimization of piezoelectric impact-based micro wind energy harvester for wireless sensor network
2015, Sensors and Actuators, A: PhysicalCitation Excerpt :In recent years, wireless sensor networks (WSNs) have been widely used at home, environment monitoring, healthcare, industrial monitoring, structural health monitoring, and traffic control [1–4].
A Bending-Type Piezoelectric Energy Harvester with a Displacement-Amplifying Mechanism for Smart Highways
2018, Journal of the Korean Physical SocietyA Survey of Energy Harvesting Technologies
2018, International Conference on Innovations in Control, Communication and Information Systems, ICICCI 2017
Anbu Elancheziyan is a Ph.D. candidate in Department of Electrical and Computer Engineering at Drexel University. His research interests include developing architectures for distributed embedded systems and their implementation, simulating network protocols and analysis of large scale networks.
Jaudelice C. de Oliveira received her Ph.D. degree in Electrical and Computer Engineering from the Georgia Institute of Technology in May 2003. She is currently an associate professor in the Department of Electrical and Computer Engineering at Drexel University. Her research interests include the development of new protocols that improve the performance of ad hoc, sensor and computer networks.
Steven Weber (M’03) received his M.S. and Ph.D. degrees from The University of Texas at Austin in 1999 and 2003 respectively. He joined the Department of Electrical and Computer Engineering at Drexel University in 2003 where he is currently an associate professor. His research interests are centered around mathematical modeling of computer and communication networks, specifically streaming multimedia and ad hoc networks.