Motivation and design of a content distribution architecture for rural areas

  • Authors:
  • Arvind Mahla;Deepak Martin;Ishani Ahuja;Quamar Niyaz;Aaditeshwar Seth

  • Affiliations:
  • Indian Institute of Technology Delhi, India;Indian Institute of Technology Delhi, India;Indian Institute of Technology Delhi, India;Indian Institute of Technology Delhi, India;Indian Institute of Technology Delhi, India

  • Venue:
  • Proceedings of the 2nd ACM Symposium on Computing for Development
  • Year:
  • 2012

Quantified Score

Hi-index 0.00

Visualization

Abstract

Decreasing prices of digital cameras, phones and video cameras have made these devices accessible to low-income communities and to development organizations working in these communities. As a result, the social media revolution around user generated content on YouTube and Facebook has found equivalents in rural areas; here, community radio, community video, and interactive voice based systems are used for social media instead of Internet websites to host user generated content. We observe that there is much benefit to be gained from sharing this content across different rural locations, but it is hard due to unavailable or flaky Internet access in these areas. In this paper, we analyze a 1000+ video content production and consumption dataset from a nonprofit organization that specializes in participatory video production about agricultural best practices, and observe that solutions to connectivity in rural areas could greatly benefit from caching of content since much production and consumption tends to be local. Based on this insight, we propose a delay tolerant network architecture for content distribution, that recognizes content objects as first class entities cachable at different nodes in the network, and uses an always-on control channel on GPRS/EDGE connections to assist in the routing of data. Finally, we simulate the dataset in accordance with our architectural design to study the performance of different routing and caching algorithms in terms of delivery latency and other metrics for 6 different districts. We find that our proposed architecture is suitable to provide content distribution services with minimal investments in IT infrastructure, and we plan to do a small-scale field deployment shortly.