Expanding Rural Cellular Networks with Virtual Coverage
Authors:
Kurtis Heimerl and Kashif Ali, University of California, Berkeley; Joshua Blumenstock, University of Washington; Brian Gawalt and Eric Brewer, University of California, Berkeley Awarded Community Award!
Abstract:
The cellular system is the world’s largest network, providing service to over five billion people. Operators of these networks face fundamental trade-offs in coverage, capacity and operating power. These trade-offs, when coupled with the reality of infrastructure in poorer areas, mean that upwards of a billion people lack access to this fundamental service. Limited power infrastructure, in particular, hampers the economic viability of wide-area rural coverage.
In this work, we present an alternative system for implementing large-scale rural cellular networks. Rather than providing constant coverage, we instead provide virtual coverage: coverage that is only present when requested. Virtual coverage powers the network on demand, which reduces overall power draw, lowers the cost of rural connectivity, and enables new markets.
We built a prototype cellular system utilizing virtual coverage by modifying a GSM base station and a set of Motorola phones to support making and receiving calls under virtual coverage. To support the billions of already-deployed devices, we also implemented a small radio capable of adding backwards-compatible support for virtual coverage to existing GSM handsets. We demonstrate a maximum of 84% power and cost savings from using virtual coverage. We also evaluated virtual coverage by simulating the potential power savings on real-world cellular networks in two representative developing counties: one in sub-Saharan Africa and one in South Asia. Simulating power use based on real-world call records obtained from local mobile operators, we find our system saves 21-34% of power draw at night, and 7-21% during the day. We expect evenmore savings in areas currently off the grid. These results demonstrate the feasibility of implementing such a system, particularly in areas with solar or otherwise intermittent power sources.
USENIX is committed to Open Access to the research presented at our events. Papers and proceedings are freely available to everyone once the event begins. Any video, audio, and/or slides that are posted after the event are also free and open to everyone. Support USENIX and our commitment to Open Access.
BibTeX
@inproceedings {180311,
author = {Kurtis Heimerl and Kashif Ali and Joshua Blumenstock and Brian Gawalt and Eric Brewer},
title = {Expanding Rural Cellular Networks with Virtual Coverage},
booktitle = {10th USENIX Symposium on Networked Systems Design and Implementation (NSDI 13)},
year = {2013},
isbn = {978-1-931971-00-3},
address = {Lombard, IL},
pages = {283--296},
url = {https://www.usenix.org/conference/nsdi13/technical-sessions/presentation/heimurl},
publisher = {USENIX Association},
month = apr
}
This paper addresses a practical real-world challenge: making cellular network service cost-effective in sparsely populated rural areas, and particularly in locations without access to grid power. GSM base stations are now available at modest cost. They can use IP networks for backhaul, including long-distance directional WiFi. The key to expanding their deployment is reduce the cost to power them.
Some of the GSM base stations are programmable using open software toolkits, making them appealing as platforms for research in the NSDI community. The authors have have derived some good insights from tinkering with an open cellular platform that is suitable for deployment in rural areas. They also have an understanding of the needs of rural deployments, backed by real-world measurement data including call records from carriers in developing regions. Their paper presents those insights in an engaging way, and will beaccessible and informative to the entire NSDI community. A key contribution of the work is an open-source software release with potential for near-term positive impact on real people who are a long way from NSDI.
The focus of the work is to add support for programmatic power management to the base station. In unserved regions providing the power often requires an up-front capital cost, e.g., for solar panels and batteries. One way to reduce that cost is to reduce the totalenergy usage. In sparse deployments the base station will have substantial idle time, particularly at night when the sun is not shining. The idea of Virtual Coverage is to power the base station only when it is in use. That raises the problem of how to wake it on demand, so that the network it is available for use immediately when it is needed, e.g., for emergency communications. The authors address this problem in part by creative use of (in essence) a remote garage door opener.
The paper shows that the Virtual Coverage idea can yield a modest energy savings even in existing deployments for which trace data is available. Of course, those deployments exist only in areas that have enough people to recover the cost of an always-on base station. More importantly, the idea can facilitate deployment in areas that are toosparsely populated to recover that cost, and so are currently unserved. Thus a key contribution of this research is to reduce the barriers to providing cellular service for a billion or so people who do not have it today.
connect with us