Dissecting Muni Wi-Fi's DNA

Municipal wireless started in the basements of open source aficionados. How did this underground movement become all the rage among commercial companies looking to make a profit?

by / February 28, 2007
Long before major corporations like EarthLink and AT&T offered to build networks in metropolitan areas nationwide, talented groups of geeks and activists were busy building systems that pushed the envelope of what was technically possible.

While the history of this movement dates back decades -- to the freenets of the 1980s and 1990s, ham radio in the 1960s and 1970s, amateur radio enthusiasts of the 1920s, and independent telephone tinkers in the 1880s and 1890s -- the past five years have seen an explosion of interest in low-cost communications and the community organizations leading the charge to bring broadband connectivity to all.

Several key projects shaped this movement, and thus shaped the development of open source communications technologies.

Innovation Unveiled
Open source wireless development is incredibly decentralized, with myriad projects spanning scores of countries, hundreds of locations and thousands of people. Any history is therefore a glimpse into a labyrinthine network of sometimes affiliated and allied projects, their spin-offs, and concurrently emerging alternatives, options and competitors.

While setting a start date is rather arbitrary, 1997 was an important year for wireless technological development. That July the Institute of Electrical and Electronics Engineers adopted a little known standard known as 802.11.

A few months later, 802.11a and 802.11b standards were adopted. These open standards created a platform for the development of cheap, unlicensed wireless devices, and would facilitate a communications revolution.

Early adopters worked to find innovative ways to interconnect these wireless devices -- aiming to build wireless local area networks (wLANs) within their homes and businesses. Pioneering groups like the Bay Area Wireless Users Group (BAWUG) in San Francisco and the Champaign-Urbana Community Wireless Network (CUWiN) Foundation in Urbana, Ill., began experimenting with larger- scale networking ideas, and by 2000, both BAWUG and CUWiN were exploring ways to build metropolitan-scale wireless networks using off-the-shelf 802.11 devices.

While the United States was a hotbed for early exploration of community wireless technologies, a European group called LocustWorld was one of the first to develop a commercial product using an open source platform for mesh wireless networking.

LocustWorld's MeshBox helped inspire a new generation of community wireless aficionados, who realized that off-the-shelf hardware combined with open source software could create the holy grail of community wireless -- ubiquitous, low-cost, mesh wireless networking.

Let the Hacking Begin!
By 2002, CUWiN had deployed its first open source mesh wireless prototypes and started giving away its software online. LocustWorld was selling its technologies to customers worldwide, and a large and growing community wireless movement was taking shape. Research and development, however, was still in its infancy.

In June 2003 -- five years after the first consumer-grade wireless devices hit the shelves -- the 802.11g standard was ratified. This new generation of equipment drove down the prices of off-the-shelf equipment and spurred renewed interest among programmers worldwide. Groups from Seattle to New York City to Montreal to Vienna renewed their efforts developing useful technologies for community wireless networking.

By 2004, two groups were leading the research and development charge -- CUWiN in the United States and Freifunk in Germany. In August 2004, CUWiN hosted the first National Summit for Community Wireless Networks, bringing together hundreds of developers and implementers to discuss technologies, telecommunications policy and deployment ideas. In September 2004, the Djursland International Institute of Rural Wireless Broadband held an international gathering of open source wireless developers to work on coordination efforts and share information about their up-and-coming technologies.

Freifunk continued to pioneer the OpenWrt wireless access point -- a Linux-based mesh wireless system that could be "reflashed" onto off-the-shelf hardware by replacing the manufacturer's software with Freifunk's technology based on optimized link state routing (OLSR).

CUWiN's NetBSD-based system relied on an entirely new routing protocol called hazy-sighted link state -- commonly known as HSLS -- and was soon to be deployed in multiple locations worldwide.

In 2005, a new player entered the open source mesh wireless arena -- the Massachusetts Institute of Technology (MIT). The MIT Roofnet project combined raw talent with a larger budget than had ever been seen in this realm, and it quickly developed an open source mesh system based on the earlier work done by Freifunk.

These efforts burned brightly, but did not last long, with a majority of the Roofnet team taking a "sabbatical" to commercialize their products through a new company called Meraki Networks. Around the same time, community wireless networking veterans started up a company called Metrix Communications to sell mesh wireless hardware to a growing number of communities interested in do-it-yourself networking.

At the same time, the continued development and deployment of community wireless technologies spurred allied projects, especially those working on authentication, capture portals, voice over Internet protocol and other synergistic technologies. The California-based NoCat project, with its mission to bring "Infinite Bandwidth Everywhere for Free," helped pioneer a new generation of freely available network management tools.

Ile Sans Fil in Montreal took NoCat to the next level, adding a suite of functionality to its WiFiDog authentication/portal technologies.

View From the Bleeding Edge
Today, the three main open source mesh wireless research and development efforts are CUWiN, Freifunk/FunkFeuer and MIT/Meraki.

Though all three are allied in their pursuit of low-cost broadband, they each have strengths and focuses for future development. CUWiN is based on the Berkeley Software Distribution cross-platform solution, which is focused on ad hoc peer-to-peer networking -- so there's no central server, single point of failure or user authentication.

In addition, CUWiN pursued a development agenda that integrates multimedia services into the local wireless network, creating a communitywide wLAN that runs independently of the network's Internet connectivity. CUWiN consistently partners with research laboratories and institutes around the world, shoring up its in-house development with cutting-edge research.

Freifunk/FunkFeuer focuses on building a rock-solid implementation of its routing system and developing OLSR RFC 3626 -- a standards-based routing protocol. In Europe, the group picked up a lot of steam by interlinking several active open source development communities. As a Linux-based system, Freifunk/FunkFeuer already has a large user base to draw from. Both CUWiN and Freifunk/FunkFeuer have explicit goals to develop highly scalable meshing technologies, and both have had tremendous successes with their work.

Roofnet/Meraki found a third way to shore up their research and development efforts -- they have taken technologies first developed at MIT and spun off a for-profit company to commercialize them. While their commitment to open source development has been questioned by numerous developers who have found their technologies increasingly difficult to access, the Meraki Mini -- a mesh wireless router the size of a deck of cards that currently costs $50 -- has been a useful resource for numerous communities building their own networks.

Ile Sans Fil and the WiFiDog project have been busy building community networks and improving their authentication and Web portal technologies. While focused on hotspots rather than communitywide mesh architectures, the information dissemination and multimedia technologies they are developing form a natural extension to the community networks of CUWiN, Freifunk and Roofnet.

The interlacing among these three projects is likewise growing.

Whether it's Meraki's hiring of a core CUWiN developer, the growing exchange of information among Freifunk and CUWiN developers, CUWiN's porting of its software to the Meraki Mini, or integration efforts to add WiFiDog to existing mesh wireless networks, a convergence is beginning that may lead to "best-of-breed" technologies from each of these projects becoming integrated into a single turn-key open source, community wireless networking solution.

At the end of January, FunkFeuer hosted a gathering of OLSR developers in Vienna to help coordinate future development of their technologies. Meanwhile, after running the National Summit for Community Wireless Networks since 2004, CUWiN is gearing up to host the first International Summit for Community Wireless Networks in May 2007.

The wild card is simply when these innovative projects will receive adequate support to create a Linux/Firefox/Apache alternative to current proprietary systems.

In the final analysis, the future of open source wireless looks incredibly bright. Adoption rates are taking off, the technologies are maturing and the features sets and functionality rivals -- and often exceed -- those of proprietary products.
Sascha D. Meinrath Contributing Writer