BerlinRoofNet

The MIT Roof Net project has demonstrated that it is possible to provide an 802.11b-based wireless network backbone infrastructure in a city like Boston (MA). The Berlin Roof Net project tries to determine if a similar network can be setup in Berlin/Germany.

The Boston network consists of approximately 40 nodes that are mounted on the roofs of buildings (hence the name ’roof net’) and is operated by students and other volunteers. RoofNet nodes discover each other automatically and create a mesh network, which, with the help of proper routing protocols, allow for transmission of IP packets between any two nodes of the network, with data rates of about 1Mbps.

Berlin has unique properties that make the establishment of such a roof network challenging:

• Berlin has generally higher buildings with more metal structures than Boston.
• Distances between project participants tend to be larger in Berlin than in Boston.
• It is yet unknown whether the mounting of RoofNet nodes/antennas on buildings is practicable in Berlin.
• Radio interference may be more intense in Berlin due to more systems that are simultaneously operated at 2.4 GHz.
• The acceptance of radio-based communication by the general public may be less pronounced in Berlin than in Boston.

Frequently Asked Questions about the Berlin RoofNet Project

Q: What is Berlin Roof Net?

A: 'Berlin Roof Net' is a project run by volunteer students of the Computer Science Department at Humboldt University Berlin. The goal is to construct a network of nodes (access points) which share Internet access over wireless radio connections. The access points are run independently by the students with their own equipment.

The project was inspired by the MIT RoofNet project in Boston, USA.

Q: Which wireless communication technologies are used?

A: One technology that is economically feasible for our purpose is IEEE 802.11b, also known as WiFi or Wireless LAN. However, the communication range of this technology is rather limited, especially in cities with no line of sight (NLOS) between the participating nodes. We would like to augment our system with 802.16a (WiMax) technology, which provides longer communication ranges and higher data rates, as soon as such equipment becomes available at a reasonable price.

Q: Is it safe for my health?

A: WaveLAN operates at 0.25 watts in a bursty mode (i.e. only when data is actually transmitted). Cellular phones go up to 3 watts and microwave ovens leak even more - and they are generally considered 'safe enough'.

All devices on the US market are approved by the FCC, which has a good public document on RF Safety.

Q: What is the vision?

A: This is a research project. Researchers often take a simple good idea and push it to its logical extreme. From this they learn a lot and finally understand what's possible and what is not. Then, taking one step back from the extreme approach, real good practical systems can be built. Here is our extreme story:

Traditional communication networks are owned by operators; i.e. companies who invest money into the deployment of cables and devices, and who operate and maintain those resources in order to offer services to users. To recover their investments (and to make some profit) - they charge their customers.

This may change: Falling hardware prices combined with ever increasing functionality may make it feasible to outsource the function of an operator (company) to the service users. The software that normally runs on operator-owned computers may be split into components that are executed at end-user owned equipment (such as computers, access points or mobile phones). Cables may be replaced by wireless links - which are even free if operated in an unlicensed frequency spectrum, such as 2.4GHz or 5GHz .

Easy? Not so fast! There is still a lot of plumbing and configuration necessary to build and run a reliable network - just that it has to happen automatically now. And, in the event that something goes wrong, the system has to be abele to automatically recover from ANY failure by itself - there is no operator to help. ... and this is exactly where our research comes in. We want to see how far the idea of 'outsourcing the network operator to the end-user' can go with today's technology. If we are successful, those community owned networks may become a vital component in making Internet communication accessible to more people, for little or no money.

Is this the death of telephone operator companies? By no way! Certain people and businesses will always require high quality communication services - and they are willing to pay for that extra quality. Those services will continue to be provided by commercial operators - at least for now :-)

Further Information

Technical Papers:

• Programming the Linksys WRT54GS Wireless Broadband Router (Tutorial)
• WiMax's technology for LOS and NLOS environments (pdf)
• Using Standardized Development Environment for OpenWRT (Tutorial)
• Upgrading the Linksys WRT54GS antenna

General Press:

• Spiegel (German), March 2005: Utopien aus Sauerkrautdosen (pdf)
• Spiegel (German), March 2005: Breitband zwischen Deich und Düne (pdf)

Similar projects in other cities:

• MIT RoofNet in Boston
• New York Wireless
• Austin Wireless
• Seattle Wireless
• http://wavelan-berlin.de -- Freies Netz in Berlin
• http://freifunk.net -- Freie (Wlan)Netze in Dtl., Österreich und der Schweiz
• WlanHain (Wireless LAN in Berlin Friedrichshain)

Related Projects:

• Click Modular Router