DocumentCode
168455
Title
FlashTrack: A Fast, In-network Tracking System for Sensor Networks
Author
Hao Jiang ; Jiannan Zhai ; Hallstrom, Jason O.
Author_Institution
Sch. of Comput., Clemson Univ., Clemson, SC, USA
fYear
2014
fDate
26-28 May 2014
Firstpage
51
Lastpage
58
Abstract
We revisit the classic object tracking problem with a novel and effective, yet straightforward distributed solution for resource-lean devices. The difficulty of object tracking lies in the mismatch between the limited computational capacity of typical sensor nodes and the processing requirements of typical tracking algorithms. In this paper, we introduce an in-network system for tracking mobile objects using resource-lean sensors. The system is based on a distributed, dynamically-scoped tracking algorithm which alters the event detection region and reporting rate based on object speed. A leader node records the detected samples across the event region and estimates the object´s location in situ. We study the performance of our tracking implementation on an 80-node test bed. The results show that it achieves high performance, even for very fast objects, and is readily implemented on resource-lean sensors. While the area is well-studied, the unique combination of algorithmic features represents a significant addition to the literature.
Keywords
object tracking; target tracking; wireless sensor networks; event detection region; in-network tracking system; mobile object tracking; object tracking problem; resource-lean devices; resource-lean sensors; sensor networks; tracking algorithms; Heuristic algorithms; Radar tracking; Sensors; Signal processing algorithms; Synchronization; Target tracking; Distributed Algorithm; Network Protocol; Object Tracking; Wireless Sensor Networks;
fLanguage
English
Publisher
ieee
Conference_Titel
Distributed Computing in Sensor Systems (DCOSS), 2014 IEEE International Conference on
Conference_Location
Marina Del Rey, CA
Type
conf
DOI
10.1109/DCOSS.2014.24
Filename
6846145
Link To Document