Satellite Radio Company Tries to Kill Wi-Fi
Internetnews.com reports on a potential FCC ruling limiting 2.4 GHz radiated signal: this is the most frightening development to hit Wi-Fi in its history. If the attempt by satellite radio firm Sirius succeeds, public space, wISP, and community Wi-Fi could suffer crushing blows.
Concourse Launches First Airport
The big news of the day (until the above story came in), as I break in today's New York Times (3rd item), is that Concourse Communications launches its Minneapolis/St. Paul International Airport (MSP) service on March 20. Concourse's approach is to not deal with the back-office at all. They're a pure infrastructure company which is contracting with airport authorities, negotiating the intricacies of that world, handling contractors for installation, and letting iPass resell access to their system.
The MSP airport will have about 70 percent coverage initially, increasing rapidly to near 100 percent service. The installation includes Bluetooth access point and printer stations. The first 30 days after the network is lit will be free. Afterwards, service is through iPass, an iPass partner or affiliate (which could include other aggregators), or at a walk-up rate of $7.95 per 24 hours, which is rapidly becoming the de facto day rate.
Concourse has the contract for the New York metro airports (Newark, Kenndy, and LaGuardia), and has already wired two terminals at Newark and Kennedy. The Port Authority of New York and New Jersey suffered huge losses in the World Trade Center attack, as it many of the administrative staff worked out of the towers, which were in fact built by and leased out by the Port Authority. Their personal and professional losses coupled with the downturn in air traffic and increased security all led to delays in Concourse's deployment.
Concourse expects to offer some limited service by mid-year in Newark and Kennedy and expand based on their experience in MSP.
DoCoMo to launch Wi-Fi trial: integrated billing might be an incentive for use.
Philips and Viewsonic introduce separate Wi-Fi displays: the Philips unit (press release in German) is detachable from a base station; Viewsonics is entirely independent and can be used like a tablet PC.
The Register reports that Britain and the Netherlands will be able to use 802.11a devices by July: the first in a wave of approvals for unlicensed 5 GHz devices using 802.11a in Europe, albeit over a smaller set of channels than available in the U.S. (But still more clear, nonoverlapping domains than 802.11b.)
Extremely technical, but precise overview of the challenges facing 802.11g finalization and rollout: mixed 802.11b and g networks could reduce overall throughput to only slightly above solo 802.11b networks because, as the author puts it, the data from b radios reduces the amount of time available for g encoding, which reduces overall throughput.
Several days ago, the Austrian wISP Metronet started service in Vienna: they assert that they are the first public WLAN company in Europe (disputation, anyone?). The network has about 40 access points with another 100 coming online shortly. Service extends to Graz and Linz soon. (Es ist sehr interessant, dass auf der Site so viel Englisch gebraucht wird! Englisch muss noch für Technologie wie Boxing und Sport die "mot juste" haben.)
Phillipe Langlois wrote in to announce a new wireless security list: Discussion in this group relates to wireless security, wardriving, GSM networks, GPRS, 802.11b and .11a security, WEP, hiperlan2, spread spectrum, WAP, bluetooth, UMTS and 3G, UWB. This list is created in the spirit of BugTraq to provide a forum for full disclosure on wireless security. Phillipe also runs a firm that makes security products, WaveSecurity. Subscribe at email@example.com or visit the archives and/or subscribe.
Let's talk about throughput
Lots of fellow reporters and Web site bloggers have been using throughput to mean maximum speed (in the sense of maximum theoretical throughput). But that falls far from the mark: users care about real throughput, or the net Net speed they can see in day-to-day use.
The IEEE defines throughput broadly as, the amount of work that can be performed by a computer system or component in a given period of time. The U.S. government's telecommunications glossary more exactly defines it in terms of bandwidth: The number of bits, characters, or blocks passing through a data communication system, or portion of that system....Throughput may vary greatly from its theoretical maximum.
The 802.11a spec may have a maximum raw bandwidth of 54 Mbps, but a practical throughput of as low as 23 Mbps, according to reports. So let's be exact: when you're talking about the potential of a system, it's fine to use the theoretical maximum number of bits; but when you're talking about the realities, you must subtract overhead (framing, addressing, etc.), error correction, and other bits that don't contain actual data in transmission.
(I'm not even going to start on latency: this two-part, four-year-old article says it all.)