TY - JOUR

T1 - On detection and tracking of variant phenomena clouds

AU - Thai, M.T.

AU - Tiwari, R.

AU - Bose, R.

AU - Helal, Sumi

PY - 2014/1/1

Y1 - 2014/1/1

N2 - Phenomena clouds are characterized by nondeterministic, dynamic variations of shapes, sizes, direction, and speed of motion along multiple axes. The phenomena detection and tracking should not be limited to some traditional applications such as oil spills and gas clouds but also be utilized to more accurately observe other types of phenomena such as walking motion of people. This wider range of applications requires more reliable, in-situ techniques that can accurately adapt to the dynamics of phenomena. Unfortunately, existing works which only focus on simple and well-defined shapes of phenomena are no longer sufficient. In this article, we present a new class of applications together with several distributed algorithms to detect and track phenomena clouds, regardless of their shapes and movement direction. We first propose a distributed algorithm for in-situ detection and tracking of phenomena clouds in a sensor space. We next provide a mathematical model to optimize the energy consumption, on which we further propose a localized algorithm to minimize the resource utilization. Our proposed approaches not only ensure low processing and networking overhead at the centralized query processor but also minimize the number of sensors which are actively involved in the detection and tracking processes.We validate our approach using both real-life smart home applications and simulation experiments, which confirm the effectiveness of our proposed algorithms. We also show that our algorithms result in significant reduction in resource usage and power consumption as compared to contemporary stream-based approaches. © 2014 ACM 1550-4859/2014/01-ART32 15.00.

AB - Phenomena clouds are characterized by nondeterministic, dynamic variations of shapes, sizes, direction, and speed of motion along multiple axes. The phenomena detection and tracking should not be limited to some traditional applications such as oil spills and gas clouds but also be utilized to more accurately observe other types of phenomena such as walking motion of people. This wider range of applications requires more reliable, in-situ techniques that can accurately adapt to the dynamics of phenomena. Unfortunately, existing works which only focus on simple and well-defined shapes of phenomena are no longer sufficient. In this article, we present a new class of applications together with several distributed algorithms to detect and track phenomena clouds, regardless of their shapes and movement direction. We first propose a distributed algorithm for in-situ detection and tracking of phenomena clouds in a sensor space. We next provide a mathematical model to optimize the energy consumption, on which we further propose a localized algorithm to minimize the resource utilization. Our proposed approaches not only ensure low processing and networking overhead at the centralized query processor but also minimize the number of sensors which are actively involved in the detection and tracking processes.We validate our approach using both real-life smart home applications and simulation experiments, which confirm the effectiveness of our proposed algorithms. We also show that our algorithms result in significant reduction in resource usage and power consumption as compared to contemporary stream-based approaches. © 2014 ACM 1550-4859/2014/01-ART32 15.00.

KW - Distributed event processing

KW - Optimization

KW - Phenomena detection and tracking

KW - Sensor networks

KW - Detection and tracking

KW - Dynamic variations

KW - In-situ detections

KW - In-situ techniques

KW - Localized algorithm

KW - Resource utilizations

KW - Speed of motion

KW - Algorithms

KW - Automation

KW - Energy utilization

KW - Intelligent buildings

KW - Mathematical models

KW - Oil spills

KW - Tracking (position)

U2 - 10.1145/2530525

DO - 10.1145/2530525

M3 - Journal article

VL - 10

JO - ACM Transactions on Sensor Networks

JF - ACM Transactions on Sensor Networks

SN - 1550-4859

IS - 2

M1 - 34

ER -