File(s) under permanent embargo
Anomaly detection models for detecting sensor faults and outliers in the iot-a survey
conference contribution
posted on 2019-01-01, 00:00 authored by Anuroop GaddamAnuroop Gaddam, Tim WilkinTim Wilkin, Maia Angelova TurkedjievaMaia Angelova Turkedjieva© 2019 IEEE. Over the past few years, the Internet of Things (IoT) has gained significant recognition to become a novel sensing paradigm to interact with the physical world. The sensors within the Internet of Things are indispensable parts and are the first port to capture the raw data. As the sensors within IoT are usually deployed in environments which are harsh, which inevitably make the sensors venerable to failure and malfunction. Beside sensor faults and malfunctions, the inherent environment where the sensors are usually installed could also make the sensor to fail prematurely. These conditions will make the sensors within the IoT to generate unusual and erroneous data, often known as outliers. Outliers detection is very crucial in IoT to detect the high probability of erroneous reading or data corruption, thereby ensuring the quality of the data collected by sensors. Data anomalies, abnormal data or outliers are considered to be the sensor data streams that are significantly distinct from the normal behavioural data. In this paper, we present a comprehensive survey that can be used as a guideline to select which outlier model is adequate for the application in the IoT context.
History
Event
Sensing Technology. Conference (2019 : 13th : Sydney, New South Wales)Pagination
1 - 6Publisher
IEEELocation
Sydney, New South WalesPlace of publication
Piscataway, N.J.Publisher DOI
Start date
2019-12-02End date
2019-12-04ISSN
2156-8065eISSN
2156-8073ISBN-13
9781728146317Language
engPublication classification
E1 Full written paper - refereedTitle of proceedings
ICST 2019 : Proceedings of the 13th International Conference on Sensing TechnologyUsage metrics
Categories
No categories selectedKeywords
Licence
Exports
RefWorks
BibTeX
Ref. manager
Endnote
DataCite
NLM
DC