Enabling Network Connectivity in Developing Nations

The Internet is so seamlessly integrated into modern life that it can be easily taken for granted. But when Jay Chen — now an assistant professor of Computer Science at NYU Abu Dhabi — stepped outside of the "bubble" of his US upbringing and got on a plane to Kerala, India, to participate in a Microsoft research internship while working toward his PhD, he experienced "a series of culture shocks" about network connectivity in the country that made a lasting impression.

"I went to this university and nothing worked," Chen, who is a researcher at NYUAD's Center for Technology and Economic Development (CTED), said. "The network was down for four hours in the afternoon — why? No idea. I went to a school and they had no Internet for a week — why? Because somebody was digging somewhere, building a road. I went to another school and they had Internet access in the form of one of those USB modems that the teacher was paying out of pocket for to get one day of access a week — why? Because they couldn’t afford it and there was no infrastructure, so that was his only option."

Improved access to technology and information will go a long way in accelerating economic productivity in rural and developing areas and will play an important role in educating the workforce so that it can compete on the international stage. But bridging the technology gap is no simple feat. While computer hardware donations distributed in rural areas are well intentioned, a critical gap remains: "The technology that exists is typically built for the developed world where everything works," Chen explains. "You have reliable power, you have reliable Internet access, maintenance is easy, and everyone knows how to use technology — there are numerous problems with those assumptions when you're taking your laptop to a place that has nothing."

This is why Chen is using his computer science expertise to research real issues on the ground in developing countries, with the aim of developing unified system architecture solutions that have the potential to be reapplied across different geographic locations that face similar challenges.

One of his first projects at CTED involved providing valuable electronic information services to educational organizations within the constraints of an extremely limited and slow network connection. In certain rural areas at the frontier of Internet service provider (ISP) locations, where connectivity may be limited to one USB modem shared across a classroom, offline Internet access has proven to be an effective solution. By using a teacher's course syllabus as a guide, Chen and his research team developed a focused web crawler to search relevant topics and automatically download pertinent web pages relating to those subjects. After being stored on an external hard drive, the information within the system architecture is then integrated as a cache for the web browser and can be accessed and produced through search, removing the need to go through a congested local area network link. The offline webpage solution has been well received at schools in India and Kenya, and is being more widely deployed.

The technology that exists is typically built for the developed world where everything works. There are numerous problems with those assumptions when you're taking your laptop to a place that has nothing.

Jay Chen, NYUAD Assistant Professor of Computer Science

Another solution that Chen has explored in the effort to support organizations with exceptionally stretched network capacity is the concept of asynchronous web browsing. This model was developed for a university in Kerala that did in fact have a decent 2-megabit connection through fiber optic cable, but when shared by more than 400 users, speed was reduced to less than one-tenth the speed of a dial-up connection. As the connection degrades, users may find that pages sometimes do not load; but with an asynchronous web browsing model, if the system detects that this situation is occurring, user requests can simply be cued up to download at a later point, without causing time-outs that might otherwise occur, thus providing a more seamless user experience.

"We put a local proxy on the user side that sits on the local area network so that it's well connected to the client's machine. That local proxy is then paired with a remote proxy across the network link, which is well connected to the Internet in a different location. The user communicates directly with the local proxy as if that were the Internet, and the local proxy serves the clients as well as it can, often using cached pages," Chen explained. "When the connection is up, but maybe slow, the local proxy will communicate with the remote proxy to get the pages that it needs and the remote proxy will communicate with servers, get the pages, filter, and compress them — so everything is pushed away from that bottleneck network link. Instead, you have these two agents on either side that negotiate with the clients and Internet servers respectively."

This innovative solution solved the research problem of reestablishing connectivity in some form, but it was not perfect during implementation, Chen admitted. Other infrastructural limitations created a new set of drawbacks, and as he explained, "Once you chop that end-to-end connection from the user to the actual servers, you need to do a lot to make it seamless again."

While Chen continues to specialize in areas of computer science and networking, he has also been involved in a range of other CTED projects, including using mobile phone text messaging as an efficient tool for gathering and collecting information for practical purposes. In many rural areas, data collection is still paper-based, using significant time and financial resources. Given that most of the world has at least a 2G connection, SMS is one of the "least common denominators" of mobile communication. Chen has been involved in developing tools, such as easy-to-use, form-based applications, as well as automated response to text-based search, in order to leverage mobile phone technology for tasks traditionally associated with computing. He has also contributed to the development of an anti-counterfeiting system that utilizes unique barcodes to track and verify the authenticity of goods as they travel through the supply chain, making it prohibitively more difficult and expensive for counterfeiters to circumvent official channels.

As Chen continues his work of creating practical technology solutions that enable greater access to information for those in the developing world, he remains mindful of the fact that as the Internet becomes faster and webpages become more complex, the technology gap is growing at an accelerating pace.

"It's always a matter of frontiers," he said. "We're always going to be getting better speeds and the frontier for connection is going to keep pushing further and further out geographically. The question is: Can we push past that frontier so that those who are way out there with no connection can get access to information?"