Authors:
Aikaterini Vlachaki
;
Ioanis Nikolaidis
and
Janelle Harms
Affiliation:
University of Alberta, Canada
Keyword(s):
Wireless Sensor Networks (WSNs), Cognitive Networking, Sample Cross-correlation, Received Signal Strength Indicator (RSSI), Channel State.
Related
Ontology
Subjects/Areas/Topics:
Aggregation, Classification and Tracking
;
Artificial Intelligence
;
Biomedical Engineering
;
Biomedical Signal Processing
;
Data Manipulation
;
Data Quality and Integrity
;
Distributed and Collaborative Signal Processing
;
Environmental Impact Reduction
;
Health Engineering and Technology Applications
;
Human-Computer Interaction
;
Methodologies and Methods
;
Neurocomputing
;
Neurotechnology, Electronics and Informatics
;
Obstacles
;
Pattern Recognition
;
Physiological Computing Systems
;
Sensor Networks
;
Signal Processing
;
Soft Computing
Abstract:
We consider a wireless sensor network in an urban environment and attempt to characterize the interference found in the communication channel by means of empirically collected Received Signal Strength Indicator (RSSI) values over Industrial, Scientific and Medical (ISM) and non-ISM bands. We assume a node-based interference classification scheme exists and examine whether nodes that classify the channel as belonging to the same class also exhibit strong cross-correlation in terms of the RSSI time series they independently observe. In effect, we are studying how the agreement of nodes, e.g., via consensus, on the class of a channel can be linked to the cross-correlation statistic and to what extent. We find that the particular class impacts the degree to which we can confidently claim that the channel observed independently by each node, and classified to belong to the same class, indeed behaves the same way.