A Crowd-Cloud architecture for Big Data analytics

Varun Mehta, Zoheb Shaikh, Kesav Kaza, H. D. Mustafa, S. N. Merchant

Research output: Chapter in Book/Report/Conference proceedingConference contribution

3 Citations (Scopus)


The 3 V's defining big data raises the need for non-conventional computing and communication architectures for its analytics and processing. In this paper we evolve a new architecture based on amalgamating benefits of Crowdsourcing and Cloud Computing. The 'Crowd-Cloud' architecture so formed presents newer domains for efficient analysis and processing of big data. By integrating the sensing and processing capabilities of the mobile dynamic cloud, the 'Crowd-Cloud' architecture provides an efficient and increased functionality in handling Big Data. The proposed architecture furthermore has two-fold benefits, minimizing the overall energy consumption and efficiently utilizing the available resources. In the context of energy constraints and dynamic nature of a mobile cloud, a job assignment problem is formulated and an algorithm is proposed to reach a priority based energy efficient solution. Simulations performed over several different scenarios using realistic energy consumption models, applicable for some of the commercially available mobile devices, favors practical implementation of the 'Crowd-Cloud' concept.

Original languageEnglish
Title of host publication2016 22nd National Conference on Communication, NCC 2016
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9781509023615
Publication statusPublished - 06-09-2016
Externally publishedYes
Event22nd National Conference on Communication, NCC 2016 - Guwahati, India
Duration: 04-03-201606-03-2016


Conference22nd National Conference on Communication, NCC 2016

All Science Journal Classification (ASJC) codes

  • Communication
  • Electrical and Electronic Engineering
  • Signal Processing
  • Computer Networks and Communications


Dive into the research topics of 'A Crowd-Cloud architecture for Big Data analytics'. Together they form a unique fingerprint.

Cite this