Abstract
A low-complexity algorithm is presented that clusters sensor nodes based on similarity in the sensed signals. This feature makes it an enabler for distributed detection of events that are impossible to identify using information available to a single node. The algorithm does not require system training prior to deployment nor does it assume statistical knowledge of the signal. Experimental results confirm that clusters produced by our algorithm match signal patterns more closely than those formed by a comparatively simple algorithm that minimizes Euclidean distance between signals.
Original language | English |
---|---|
Title of host publication | IEEE SENSORS 2017 - Conference Proceedings |
Publisher | Institute of Electrical and Electronics Engineers Inc. |
Pages | 1-3 |
Number of pages | 3 |
Volume | 2017-December |
ISBN (Electronic) | 9781509010127 |
DOIs | |
Publication status | Published - 21 Dec 2017 |
Event | 16th IEEE SENSORS Conference, ICSENS 2017 - Glasgow, United Kingdom Duration: 30 Oct 2017 → 1 Nov 2017 |
Conference
Conference | 16th IEEE SENSORS Conference, ICSENS 2017 |
---|---|
Country/Territory | United Kingdom |
City | Glasgow |
Period | 30/10/17 → 1/11/17 |
Keywords
- event detection
- sensor clustering
- similarity metric