Enhanced-KNN (M-KNN) based outlier detection and sensor data aggregation for large data streams in the IoT-cloud
by Y.R. Sampath Kumar; H.N. Champa
International Journal of Internet Protocol Technology (IJIPT), Vol. 15, No. 3/4, 2022

Abstract: IoT is a technology that facilitates several applications of the modern age. The pervasive deployment of Sensor Nodes (SN) affects the sensor data acquisition. The IoT-data requires utilisation of Dimensionality Reduction (DR) and classification techniques. An advanced Data Aggregation (DA) technique decreases the number of total data transmissions and facilitates data accuracy. Here, a data analysis framework is suggested for DA and Outlier Detection (OD) by employing a modified K-Nearest Neighbour (M-KNN) algorithm. Further, the recovery error rate as well as Energy Consumption (EC) of proposed methodology is analysed and compared with the existing Recursive Principal Component Analysis (R-PCA) technique. The techniques are compared by varying the cluster size and the results depict that the proposed M-KNN attains the lowest relative error and lowest EC for variable number of cluster size for both Intel and NDBC-TAO data. The proposed model is crucial for exploring the sensor data and predicting future events based on observed sensor data analysis.

Online publication date: Wed, 05-Oct-2022

The full text of this article is only available to individual subscribers or to users at subscribing institutions.

 
Existing subscribers:
Go to Inderscience Online Journals to access the Full Text of this article.

Pay per view:
If you are not a subscriber and you just want to read the full contents of this article, buy online access here.

Complimentary Subscribers, Editors or Members of the Editorial Board of the International Journal of Internet Protocol Technology (IJIPT):
Login with your Inderscience username and password:

Username:        Password:          Forgotten your password?

Want to subscribe?
A subscription gives you complete access to all articles in the current issue, as well as to all articles in the previous three years (where applicable). See our Orders page to subscribe.

If you still need assistance, please email subs@inderscience.com