Institutional Repository [SANDBOX]
Technical University of Crete
EN  |  EL

Search

Browse

My Space

Scalable approximate query tracking over highly distributed data streams with tunable accuracy guarantees

Giatrakos Nikolaos, Deligiannakis Antonios, Garofalakis Minos, Keren Daniel, Samoladas Vasilis

Full record


URI: http://purl.tuc.gr/dl/dias/A16D0308-EF61-468D-8B05-3162D819231B
Year 2018
Type of Item Peer-Reviewed Journal Publication
License
Details
Bibliographic Citation N. Giatrakos, A. Deligiannakis, M. Garofalakis, D. Keren and V. Samoladas, "Scalable approximate query tracking over highly distributed data streams with tunable accuracy guarantees," Inf. Syst., vol. 76, pp. 59-87, Jul. 2018. doi: 10.1016/j.is.2018.05.001 https://doi.org/10.1016/j.is.2018.05.001
Appears in Collections

Summary

The recently proposed Geometric Monitoring (GM) method has provided a general tool for the distributed monitoring of arbitrary non-linear queries over streaming data observed by a collection of remote sites, with numerous practical applications. Unfortunately, GM-based techniques can suffer from serious scalability issues with increasing numbers of remote sites. In this paper, we propose novel techniques that effectively tackle the aforementioned scalability problems by exploiting a carefully designed sample of the remote sites for efficient approximate query tracking. Our novel sampling-based scheme utilizes a sample of cardinality proportional to N (compared to N for the original GM and its variants), where N is the number of sites in the network, to perform the monitoring process. Our extensive experimental evaluation and comparative analysis over a variety of real-life data streams demonstrates that our sampling-based techniques can significantly reduce the communication cost during distributed monitoring with controllable, predefined accuracy guarantees. In that, we manage to scale the monitoring of any given non-linear function on much higher network scales which had not been reached by any GM related method or variant so far.

Services

Statistics