DocumentCode :
2135737
Title :
Intelligent protocols based on sensor signal change detection
Author :
Reznik, Leon ; Von Pless, Gregory ; Al Karim, Tayeb
Author_Institution :
Dept. of Comput. Sci., Rochester Inst. of Technol., NY, USA
fYear :
2005
fDate :
14-17 Aug. 2005
Firstpage :
443
Lastpage :
448
Abstract :
Embedded into the communication protocols, the signal change detection will allow data compression for improving network efficiency. It enhances reliability and security also. The proposed change detector employs a neural network function prediction in order to determine if the sensor outputs have changed In addition to the change detection system, a modification to a standard neural network function predictor is proposed that allows the change detection system to quickly learn how to accurately predict next sensor outputs. The parameter choice and the relationship between the threshold values and false alarm and change missing rates are studied The protocol is implemented and tested in real life environments with sensor networks built from Crossbow MICA-2 motes. Sensor network change detection system, which is designed to become a protocol core utility, is described The test results are analyzed and recommendations on applications are derived.
Keywords :
distributed sensors; neural nets; protocols; signal detection; Crossbow AFCA-2 motes; change missing rates; communication protocols; data compression; false alarm rates; intelligent protocols; network efficiency; neural network function prediction; parameter choice; reliability; security; sensor networks; sensor outputs; sensor signal change detection; threshold values; Communication system security; Data compression; Data security; Detectors; Intelligent sensors; Neural networks; Protocols; Sensor systems; Signal detection; Telecommunication network reliability; reg;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Systems Communications, 2005. Proceedings
Print_ISBN :
0-7695-2422-2
Type :
conf
DOI :
10.1109/ICW.2005.52
Filename :
1515562
Link To Document :
بازگشت