Authors:
Zhiming Ding
1
;
Qi Yang
2
and
Limin Guo
1
Affiliations:
1
Chinese Academy of Sciences, China
;
2
National Center of ITS Engineering & Technology, China
Keyword(s):
Sensor Data Management, Spatial-temporal Data, Massive Data Processing, Cloud Data Management.
Related
Ontology
Subjects/Areas/Topics:
Data Communication Networking
;
Data Manipulation
;
Enterprise Information Systems
;
Internet of Things
;
Multi-Sensor Data Processing
;
Sensor Data Fusion
;
Sensor Networks
;
Software Agents and Internet Computing
;
Software and Architectures
;
Telecommunications
Abstract:
Massive sensor data management is an important issue in large-scale sensor based systems such as the Internet/web of Things. However, existing relational database and cloud data management techniques are inadequate in handling large-scale sensor sampling data. On the one hand, relational databases can not efficiently process frequent data updates caused by sensor samplings. On the other hand, current cloud data management mechanisms are largely key-value stores so that they can not support complicated spatial-temporal computation involved in sensor data query. To solve the above problems, we propose a Relational Data-Base and Key-Value store combined Cloud Data management (“RDB-KV CloudDB”) framework, in this paper. The experimental results show that the RDB-KV CloudDB can provide satisfactory query processing and sensor data updating performances in large scale sensor-based systems.