Microsoft’s WhiteFi: Wi-Fi Using Whitespaces

Long-range, low-cost wireless Internet could soon be delivered using “whitespaces” — radio spectrum once reserved for use by TV stations. This week at ACM SIGCOMM 2009, a communications conference held in Barcelona, Spain, a computer network that uses “white spaces” in a fashion somewhat similar to Wi-Fi was outlined.

Networking over UHF white spaces is fundamentally different from conventional Wi-Fi, explains Ranveer Chandra of Microsoft Research, a main contributor to the paper. Their WhiteFi approach is a Wi-Fi like system constructed on top of UHF white spaces. WhiteFi incorporates a new adaptive spectrum assignment algorithm to handle spectrum variation and fragmentation in unused television channesl, and proposes a low overhead protocol to handle temporal variation.

Most of the prior research in UHF white spaces has focused on accurately detecting the presence of incumbent RF signals, says Chandra in the paper (pdf). Researchers are now beginning to look at the problem of establishing a wireless link between white space devices. Their research pushes the state-of-art to the next level by going beyond a single link.

Matt Welsh, a Professor of Computer Science at Harvard University, and one of the authors of the paper (pdf), tells Dailywireless that the paper was presented at the conference by Rohan Murty at Harvard University, one of his Ph.D. students who contributed to the work along with Ranveer Chandra of Microsoft Research. The paper (pdf) is a joint effort between researchers at Microsoft Research and at Harvard.

WhiteFi focuses primarily on the problem of setting up a Wi-Fi like network, consisting of an Access Point (AP) with multiple associated clients. It uses a new spectrum assignment algorithm with a new mechanism that quickly discovers APs operating anywhere in the 180 MHz white space, using any arbitrary channel width. They also described a new technique for handling disconnections where clients signal to the AP without interfering with ongoing wireless microphone transmissions.

They estimate UHF spectrum fragmentation in 3 settings: urban, suburban and rural (population less than 6000). In all 3 settings there is at least one locale in which there is a fragment of 4 contiguous channels available, that is, 24 MHz of spectrum. In rural areas fragments of up to 16 channels are expected.

Microsoft researchers tested the new protocol, called White Fi, in the Redmond campus. It uses UHF white spaces and adaptively configures itself to operate in the most efficient part of the available white spaces. TV spectrum could provide good long-range connectivity in rural areas, and help fill in gaps in city networks.

The spectrum between 512 megahertz and 698 megahertz was originally allotted to analog TV channels from 21 to 51. It offers a longer range than conventional Wi-Fi, which operates at 2.4 gigahertz. “Imagine the potential if you could connect to your home [Internet] router from up to a mile,” says Ranveer Chandra, a member of the Networking Research Group at Microsoft Research behind the project.

The FCC ruled last November that companies could build devices that transmit over white spaces but also gave strict requirements that this should not interfere with existing broadcasts, both from TV stations and from other wireless devices that operate within the same spectrum. Chandra and his colleagues say their “White Fi,” protocol can successfully navigate the tricky regulatory and technical obstacles involved with using white spaces.

A second approach, which is being considered by the IEEE 802.22 working group, involves an explicit channel renegotiation protocol between clients and APs when they detect a wireless mic. This approach, however, assumes that control messages will not induce audible interference on the wireless mic.

The White Spaces Coalition consists of eight large technology companies that plan to deliver high speed broadband internet access in unused television frequencies between 54-698 MHz (TV Channels 2-51). The coalition expects speeds of 80 Mbps and above, and 400 to 800 Mbps for white space short-range networking. The group includes Microsoft, Google, Dell, HP, Intel, Philips, Earthlink, and Samsung Electro-Mechanics. The White Spaces Database Group maps out available spectrum.

On February 27 2009, the National Association of Broadcasters (NAB) and the Association for Maximum Service Television, Inc. (MSTV) asked a Federal court to shut down the FCC’s authorization of white space wireless devices. The plaintiffs allege that portable, unlicensed personal devices operating in the same band as TV broadcasts have been “proven” to cause interference despite FCC tests to the contrary.

In (somewhat) related news, researchers at Harvard University and BBN have developed CitySense, an urban scale sensor network testbed. CitySense will consist of 100 wireless sensors deployed across a city. Each node will consist of an embedded PC, 802.11a/b/g interface, and various sensors for monitoring weather conditions and air pollutants.

Harvard’s Sensor Networks Lab has also deployed three wireless sensor networks on active volcanoes.

Read Write Web reviews on Citysense and WikiCity, iPhone applications that integrate sensor networks with social networks. A recent W3C Workshop on the Future of Social Networking, held in Barcelona in January, reviews the trend of sensors mixing with social networks and offers some real-world applications.

Both Social Networks and Sensors information can be modeled using Semantic Web technologies, says the paper. They can be connected in an interoperable and straightforward way. The W3C’s Resource Description Framework (RDF), is an open Web standard that can be freely used by anyone.

By combining social networks and social activities to Semantically-Interlinked Online Communities, the Semantic Web can map existing applications in new and innovative ways.

Products involving Web-connected devices include flood gauges, air pollution monitors, stress gauges on bridges, and mobile heart monitors. CardioSign hopes to commercialize a wearable blood pressure sensor.

Sensor Web XML-based specifications were created in consideration of Semantic Web technology, which allows data from various sources to be used with a common semantics for the data.

Read Write Web says mobile apps like Citysense or Brightkite (their pick last December for Most Promising App for 2009), promise great power and flexibility and may find favor in both popular mobile applications as well as in areas like healthcare.

Dailywireless also wrote about CitySense some 2 years ago.

Posted by Sam Churchill on .

Leave a Reply