Title :
Detecting Faulty Percepts in a Context-Aware Ubiquitous System
Author :
Ahmed, Bilal ; Lee, Young-Koo ; Lee, Sungyoung
Author_Institution :
Dept. of Comput. Eng., Kyung Hee Univ., Seoul
Abstract :
Perception of a ubiquitous system is based solely upon sensor and device readings. The system decides to provide adequate services based upon these readings. We define a system of fault detection in the perception mechanism of a ubiquitous system based upon Bayesian networks and the combination of beliefs from independent networks. The proposed scheme is distributed and adequately suits the functioning of a large multi-domain ubiquitous system. Eradicating faulty percepts and identifying the malfunctioning device would improve the quality of service, and make the system more reliable.
Keywords :
belief networks; computer network reliability; fault diagnosis; quality of service; ubiquitous computing; Bayesian networks; context-aware ubiquitous system; faulty percepts detection; large multi-domain ubiquitous system; quality of service; Bayesian methods; Computer vision; Context modeling; Embedded computing; Fault detection; Fault diagnosis; Fault tolerance; Minimization methods; Pervasive computing; Sensor systems; Bayesian Networks; Context-Awareness; Fault-Detection; Ubiquitous Systems;
Conference_Titel :
Advanced Communication Technology, The 9th International Conference on
Conference_Location :
Gangwon-Do
Print_ISBN :
978-89-5519-131-8
DOI :
10.1109/ICACT.2007.358358
