Open AccessNon‐stationarity of body area networks for sleep monitoring
Author(s) -
Chaganti V.,
Hanlen L.,
Smith D.
Publication year - 2013
Publication title -
electronics letters
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.375
H-Index - 146
ISSN - 1350-911X
DOI - 10.1049/el.2013.1511
Subject(s) - sleep (system call) , computer science , artificial intelligence , operating system
The stationarity of the body area network (BAN) channel when the subject is sleeping is examined. The authors question the validity of first‐ and second‐order distribution fits to BAN channel gain. Wide sense stationarity tests have an upper bound at 70% probability for a maximum data length of 5 s. For data lengths greater than 5 s, the probability of stationarity rapidly diminishes. It is shown that the BAN channel is inherently non‐stationary, and independent of the subject, sensor location and activity.