Clusters divide data into groups so that we can gather beneficial and meaningful information. To obtain meaningful information clusters should capture the natural structure of data. In some scenarios cluster analysis is only a useful starting point for other purposes, such as data summarization . At a high-level Clustering algorithm are classified as Partition based (K-means), Hierarchical based (Birch), Density based(DBScan), Grid based and Model based (SOM). They all are having capabilities to classify the data efficiently. But it depends on the type of data-set some one is using.

This paper might help you:

Machine learning for Internet of Things data analysis: A survey

Thank you Kshira Sagar Sahoo for your massive explanation. I will go through the suggested paper.

I hope that you will benefit from your search

https://towardsdatascience.com/the-5-clustering-algorithms-data-scientists-need-to-know-a36d136ef68

https://ieeexplore.ieee.org/abstract/document/7976279/

Thank you Benrazek Alaeddine

LEACH - Low energy adaptive clustering hierarchy and its associated variants

K-means - based on distance / energy

Thank you dear Abhishek Tayal

Use incremental clustering algorithms

Thank you Adinarayana Salina

Dear

use fuzzy K- Medoids algorithm. It gives best results

Thank you Santosh Satapathy