FASTBEE: A Fast and Self-Adaptive Clustering Algorithm Towards to Edge Computing

With the rapid development of data explosion era, the cloud computing has been unable to process the massive data efficiently which forced on a urgent need for edge computing. Edge computing refers to a new type of computing model that performs computations on a large amount of data at the edge of a...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Veröffentlicht in:2018 5th IEEE International Conference on Cyber Security and Cloud Computing (CSCloud) 2018 4th IEEE International Conference on Edge Computing and Scalable Cloud (EdgeCom) S. 128 - 133
Hauptverfasser: Du, Haizhou, Zeng, Shuqin, Dou, Tingting, Fang, Wei, Wang, Yi, Zhang, Chenlu
Format: Tagungsbericht
Sprache:Englisch
Veröffentlicht: IEEE 01.06.2018
Schlagworte:
Cloud computing
Clustering algorithms
Computational modeling
Conferences
Data models
Density based Clustering
Density collision
Edge computing
Heuristic algorithms
Online-Zugang:Volltext
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
Beschreibung
Zusammenfassung:With the rapid development of data explosion era, the cloud computing has been unable to process the massive data efficiently which forced on a urgent need for edge computing. Edge computing refers to a new type of computing model that performs computations on a large amount of data at the edge of a network. In order to improve the operating efficiency in the network, we put forward that applying a fast and self-adaptive clustering algorithm to the edge computing which helps the edge devices to distribute different types of data clustered to the cloud computing center. In our paper, we proposed the FASTBEE algorithm which is suitable for edge computing. The FASTBEE algorithm makes improvements on density collision and dynamic determination of density threshold by using gradient descent method to update the sum of squared errors formula. The proposed algorithm is extensively tested on several well-known datasets. The results proved the performance of our approach that its accuracy is 36 percent higher and it runs much faster compared with the CFSFDP algorithm and the DBSCAN algorithm.
DOI:10.1109/CSCloud/EdgeCom.2018.00031