UC Berkeley

The last billion people to join the online world, are likely to face at least one of two obstacles:

Part I: Rural Internet Access

Rural, sparsely populated, areas make conventional infrastructure investments unfeasible: Big

corporations attempt to address this challenge via the launch of Low-Earth-Orbiting (LEO) satellite

constellations, fleets of high-altitude balloons, and giant solar-powered drones; although these

grandiose initiatives hold potential, they are costly and risky. At the same time, small local operators,

Wireless Internet Service Providers (WISPs), are growing in numbers, in subscribe base and

in territory covered. WISPs can play a major role in serving a growing number of rural communities,

as well as offer real competition to incumbent operators in urban and semi-urban markets,

leading to better service at reduced costs.

The key motivation for this work is to lower the barriers-to-entry for small rural WISPs, and to

sustainably grow their operation — this has been my research focus for over 15 years.

The core of this work is based on a case-study of a WISP, FurtherReach — which we have

built from the ground up. This WISP brings broadband Internet service to hundreds of subscribers

at the south coast of Mendocino county in California. Through designing, deploying and operating

this venture, we learn about the real challenges faced by WISPs, develop technical solutions as

well as business models, operational methodologies and deployment strategies. The FurtherReach

case study is presented in chapter 2. Chapter 1 introduces the overall WISP ecosystem. In chapter

3 we discuss the potential of Software Defined Networks (SDN) to aid in WISP operations. Finally,

chapter 4 discusses the often overlooked computer-security concerns that are unique to rural

communities, especially in developing countries.

Part II: Dissent Networking

Oppressive regimes censor and restrict information flow. Sadly, Internet censorship, in some

countries, does not seem to be going away, and presents a growing challenge. The degree and effectiveness

of censorship varies greatly, as does the risk of getting caught circumventing it. Similarly,

the technologies to aid dissenters vary accordingly. My work in this field predates smart-phones,

which I believe could be made to offer safe and effective solutions even in the most dangerous

of countries. Should we consider these technologies Internet access? Once again the degree of

connectivity, and especially interactivity, from behind a censoring firewall, varies greatly.

In chapter 5, I present our attempt at defining the threats and narrating the exceptionally challenging

problem space. I find this chapter quite discouraging as it dictates exceptional restrictions

on the design space, yet ignoring these constraints may put users of the technology in greater risks

than without it. It makes us question if technology can help at all?

I continue in chapter 6, to present Rangzen, our initial attempt at designing a solution that

adheres to the strict constraints presented in 5. While limited in functionality, especially given

its delay-tolerant approach, which is incompatible with many Internet applications that expect

real-time interactivity, it does successfully follow our design guidelines for dissent technologies

as presented in chapter 5. We have built an Android app, Rangzen, based on these design specifications,

which undergoes a beta testing program at the time of this writing. The app will be

distributed freely on the Android store in January 2016, and the code is open source and available

to the public.