PROJECT TITLE :

Datum: Managing Data Purchasing and Data Placement in a Geo-Distributed Data Market - 2018

ABSTRACT:

This Project studies two design tasks faced by a geodistributed cloud information market: which information to buy (information purchasing) and where to put/replicate the information for delivery (knowledge placement). We show that the joint downside of knowledge getting and information placement within a cloud knowledge market will be viewed as a facility location problem and is so NP-hard. But, we tend to provide a provably optimal algorithm for the case of a data market created of one information center and then generalize the structure from the only data center setting so as to develop a close to-optimal, polynomial-time algorithm for a geo-distributed information market. The ensuing design, Datum, decomposes the joint getting and placement drawback into 2 subproblems, one for data purchasing and one for data placement, using a transformation of the underlying bandwidth costs. We tend to show, via a case study, that Datum is near optimal (inside 1.half dozen%) in sensible settings.