Rural Internet Measurements:Main Page
From ACT4D Project Wiki
As of today there is very little understanding of the nature of Internet connectivity available in rural regions of the world. We intend to look at understanding the status of Internet connectivity at rural locations across India as a starting point. To this end, we intend to measure various parameters at offices and community centers located in rural regions of India.
While understanding network characteristics in rural regions is interesting in itself, knowing connectivity status can help move forward in several directions:
- How throughput varies in small and large time scales can help design better transport protocols
- Understanding of network up times can help make a case for delay/disruption tolerant applications. Semantic content prioritization can be explored as a consistent feature across such applications.
- Understanding the challenges of conducting measurements can help build a framework for conducting measurements in rural regions
- Understanding experience of users using such connectivity can help understand impact of the connectivity on users and potential directions of improving user experience.
- Bandwidth aggregation over multiple connections can be explored as a way of increasing uptime/performance.
What to Measure
Details of this discussion are present here.
For all the temp stuff
- Criteria for selecting device/box for rural internet measurements
- How to make it easy for untrained people to run the measurements
- How to choose service providers and connection technologies for measurements?
Most of the work related to Internet access in rural regions falls in one of the following categories:
- Internet connectivity infrastructure: This includes DTN and Wireless Long Distance connections related stuff.
- WWW access in developing regions: One or both of low throughput and high latency are identified as challenges. These are usually based on one or two sample points of measurements that are cyber cafes or universities. One study was based on an in network view of the connectivity patterns. One or more of the following approaches are used to address them:
- Content caching at a proxy
- Content prioritization at port/flow level to improve interactivity
- User experience of searching/web access as a result of poor connectivity.
It is interesting to note that no study was found that provided a larger coverage of connectivity patterns as observed from end users in rural regions.
In addition, there is a decent collection of work that studies network characteristics of home networks in developed world. Results from these studies can form our benchmark for comparison. These studies can also help us identify exactly what should be measured.
Detailed related work is present here.
This section contains all the experiments done till now including background, tests done, results, etc.