Broadband Wireless Internet Access (BWIA) News

Toshiba Home Audio Server Uses 802.11b Wireless

[It] costs too much for too little, but gotta start somewhere, and it has a wireless component. Toshiba announced the Magnia SG20, a server for the home intended to be a "digital hub for the entire home". To its credit, Toshiba at least recognizes that the market for a "digital catch-all for the home" exists. Many homes now have multiple computers and digital media files - electronic photos, digitized video, and digitized audio are all clogging individual computer's hard disks. It absolutely makes sense for such a household to have a home server with big, cheap EIDE/Ultra ATA disks in a redundant array to guard against individual disk failure. The Magnia SG20 is badly flawed in its bang-for-the-buck. First, the price - $1399. Urk! That high price is a deal-killer for an appliance that most don't yet recognize that they need. Second the incredibly paltry amount of hard disk storage that it ships with - 15 GB, when 80 GB drives are hovering at $100 (retail) and 160 GB drives are at $299. The high price would seem a lot more justified if it included a 160 GB drive to start with... but a mere 15 GB is a bit of a joke. Most of the PCs in households that can be expected to use such a server appliance (remember, only "technically advanced" households are going to realize that they have some need for this) will have bigger hard drives than the server. If that's the case... what's the point of a server? To its credit, Toshiba implements the wireless portion of the system as a PC card, theoretically allowing upgrades to newer wireless standards as they emerge, such as 802.11a, Ultrawideband, etc. In reality, because the Magnia SG20 is a proprietary system, the ability to use additional wireless cards is wholly dependent on Toshiba developing the requisite drivers for particular cards. Also to its credit, Toshiba includes "disk slots" to add hard disk capacity, but I don't think that saves the product. For $1399, "technically advanced" households are likely to dedicate an older PC as the household server and add some disk drives to it, especially as USB 2.0 and IEEE-1394 (FireWire) make it more and more feasible to easily expand disk drive capacity using serial connections.

Bold Prediction Department: Lots of bad press, user angst, and probably lawsuits will result from failures of home server appliances and lost data unless the vendors make backups painless and nearly witless (using tape, online [a huge opportunity for someone, perhaps the vendor of the home server appliance to get online backup right... as in obscenely profitable], DVD-R, or whatever. More info at http://www.pcworld.com/news/article/0,aid,101965,00.asp. (Thanks to Good Morning, Silicon Valley (http://www.siliconvalley.com/mld/siliconvalley/business/columnists/gmsv/3493922.htm) for the pointer to this story.)

Steve Stroh

Copyright (c) 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

Proxim Is Buying Orinoco Wireless LAN Product Line From Agere

Finally, the agony this long suffering group has suffered is about to be ended. What is now the Orinoco line of wireless products began life at NCR as WaveLAN, one of the first Wireless LAN product offerings. In succession, NCR was bought by AT&T (and later spun back out), with the WaveLAN product line being transferred to Lucent. Lucent recently spun out its electronic components group as Agere, and the Orinoco product line went with Agere. Now the more visible parts of the Orinoco product line are being sold to Proxim, a company oriented entirely around Wireless LANs (and derivatives).

With this announcement, Agere refocuses on providing only components and subassemblies. "Agere will retain its 802.11/Wi-Fi chips, modules, and cards business, representing the majority of Agere's total Wi-Fi business." From this, it would appear that Orinoco cards and 802.11b modules will be made by Agere, but sold at retail by Proxim under the Orinoco brand along with the Orinoco Residential Gateways (RG-xxxx products), access points (enterprise), and all the outdoor products.

The joy is not all one-sided; Proxim is gaining one of the better-known brand names in the WLAN market, and a considerable market within enterprises. In very large enterprises, where good support is prized almost above all, the Orinoco product line has done well.

Busy times at Proxim, still busy assimilating the product lines and personnel of Western Multiplex, completing that merger just a few months ago.

Steve Stroh

Copyright (c) 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

XO Communications Enters Chapter 11

The Wall Street Journal reports that XO Communications (www.xo.com) has at last filed for Chapter 11 Bankruptcy. XO was the last of the big four Broadband Wireless Internet Access companies that made such a big spash in the late 1990's to enter bankruptcy. The others were Winstar, Advanced Radio Telecom (ART), and Teligent.

Steve Stroh

Copyright © 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

IEEE 802.11G Working Group May Propose 4 Non-overlapping Channels

On the IEEE's 802.11 page, Task Group G (http://grouper.ieee.org/groups/802/11/Reports/tgg_update.htm), it states: 4 Channels in the 2.4GHz Band - Calls will discuss the feasibility of adding additional channel specification that would allow the use of 4 non-overlapping channels in the 2.4 GHz band instead of the current 3. This would be a big win for all concerned. 802.11b's occupied channel is 22 MHz. The total width of the 2.4 GHz band is 83.5 MHz. Currently there are only three non-overlapping channels specified for 802.11b (Channels 1, 6, and 11). The channel width of 802.11g using OFDM 20 MHz, so it's possible to cram in an extra channel and still stay within the band.

Steve Stroh

Copyright © 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

Chipset For 80211b SD Cards

In a related development to [Pocket PC With Integrated 802.11b Debuts], on June 10, 2002, SyChip announced that they will soon offer a chipset that will make possible 802.11b Secure Digitital (SD) cards. To put this in perspective, many PDA vendors, including those using Palms, have standardized on SD as the expansion card slot of choice. This announcement presages the possibility of easily adding 802.11b to many PDAs. Thanks for the heads-up on this development to Glenn Fleishman (80211b.weblogger.com) and 802.11 Planet (www.80211-planet.com). More information on SyChip's announcement is at http://www.sychip.com/PUB/WLAN%20press%20release%20060902%20v1.pdf.

Steve Stroh

Copyright © 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

Pocket PC With Integrated 802.11b Debuts

The WSJ also reports that Toshiba has announced the Pocket PC e740, claiming it to be "The Industry's First PDA with Integrated Wi-Fi". It's impressive that in addition to the built-in 802.11b, the e740 also comes with slots for Secure Digital (SD) and Compact Flash (CF) expansion modules. This is the first of what will undoubtedly be many, many announcements of PDAs with integral 802.11b capability. I think PDA's will quickly become the majority of the equipment that is used at Public Wireless Access Points, accelerating the demand for PWAP coverage (see my weblog for 6/14/2002). Detailed information on the e740 is at http://cdgenp01.csd.toshiba.com/content/product/pdf_files/detailed_specs/pocketpc_to_e740.pdf.

Steve Stroh

Copyright © 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

WaveRider - OK, I'm Impressed - June 6, 2002 (Posting to the ISP-Wireless Mailing List)

I thought this would be of interest to readers. The title comes from a June 6, 2002 posting to the isp-wireless mailing list (http://www.isp-wireless.com) by Marlon Schafer of Odessa Office Equipment (http://www.ooe.com). Marlon is a Wireless ISP in Odessa, WA that is (deservedly) held in very high regard by the inmates... er, "audience" of the isp-wireless mailing list. Rather than co-opt Marlon's words, I'll refer readers to the archive version of Marlon's post at http://isp-lists.isp-planet.com/isp-wireless/0206/msg00863.html. Marlon's posting is well worth reading to get a sense of just how different the performance of the WaveRider equipment is from Marlon's well-developed 2.4 GHz Wireless Internet Access network in Odessa, WA.

WaveRider Communications (http://www.waverider.com) is a Canadian Broadband Wireless Internet Access equipment manufacturer. WaveRider's Last Mile Solution (LMS) product line operates in the 902-928 MHz ISM band. This is significant because 902-928 MHz has far superior foliage penetration characteristics than equipment that operates in the 2.4 GHz band. The relative lack of foliage penetration is a major factor in Wireless ISPs not being able to serve every potential customer who desires Wireless Internet Access service. To date, the 902-928 MHz band has been largely ignored by the vast majority of Wireless ISP equipment vendors (two important exceptions are Wi-LAN - www.wi-lan.com and the late, much-lamented Metricom who used the 902-928 MHz band extensively for user-to-poletop communications.) Largely ignored, because 902-928 MHz offers a smaller amount of spectrum in comparison to 2.4 GHz (26 MHz versus 83 MHz) and is 902-928 MHz is, for the most part, only available in North America (the 2.4 GHz band is nearly universal). I've long thought that with proper equipment, 902-928 MHz could easily be a "Wireless ISP Secret Weapon" in areas where Wireless Internet Access service is desired (customers waiting...) but 2.4 GHz equipment just won't penetrate. While there are many other vendors of 902-928 MHz data communications equipment (largely used for SCADA and other low-bandwidth telemetry applications), such equipment is largely inappropriate for Wireless Internet Access use. WaveRider's LMS product line is relatively new and hasn't been very extensively reviewed. As I state in my posting, I had some previous experience with 902-928 MHz equipment so had some basis for comparison.

On June 7, 2002, I followed up Marlon's posting with my own posting to isp-wireless:
Well, Marlon beat me to the keyboard, but he was about 200 miles closer to his than I was to mine at the end of the day.

I had to give Marlon a bit of hell because for his initial tests he had like 2/3 of the antenna taped up against the steel mast, and I knew on sight that was severely compromising anything resembling a normal radiation pattern (if not actually causing SWR and foldback of transmit power). Something that I didn't reveal to Marlon at the time was that I'd had some experience with 900 MHz gear, and I was fairly sure that what he was describing was not normal behavior for a 900 MHz system, which in general penetrates trees, houses, etc. pretty well.

For the tests we ran together, if we had mounted the base antenna up and in the clear better (something resembling a permanent installation) and away from the shading effect of the water tank, I think we would have had total saturation coverage at both ends of town where we lost signal.

Like Marlon, I was very impressed with the coverage (once we finally cottoned on to which end was front of the CPE panel antenna). TAKE THE HINT, WAVERIDER - PUT A LABEL ON THE FRONT! One thing I observed was that when we lost signal, it took a few seconds to recover, like the radio wasn't going to let the ping go through until it was "sure" that the signal had been re-established. In that time we had moved 10 feet or so. In a real-world test, the radio isn't designed to be mobile, so I don't count that "slow acquisition time" against it. In a real environment, you'd move the antenna around the room in various places to find the best signal, and then LEAVE it there.

A minor nit... the CPE panel antenna is very light in comparison to the relatively stiff, heavy coax cable. I can easily imagine the antenna getting pulled out of a window sill by the weight of the coax. I didn't get to see the entire system as it came in the box (Marlon pretty well stripped it down to essentials to take it on the road), but if there isn't a mounting frame that gives a bit of weight and stability to the panel antenna, there should be.

Another minor nit - the RSSI utility (constantly writing out the RSSI readings when in a telnet session) worked great, but the telnet session timed out way too quickly for our use in road testing... of course, it's designed as a quick test for optimizing the location of the antenna within the range of the length of coax (which is about 6' or so). So, I don't count that against them either, it just would have been nice to see RSSI readings continuously just like we were seeing the pings.

One last thing that I clarified with Mitch while on the phone. The cavity filter that WaveRider offers may not be necessary unless you're going to deploy in close proximity to paging transmitters AND you're going to run enough base stations that you'll need to use 5 MHz channels near the edges of the band. For Marlon's test we set it on 915 MHz and it worked fine... but there weren't any known paging transmitters nearby.

All in all, nice job, WaveRider. Please tell Bruce Sinclair that THIS was the product that I had REALLY wanted to see and hear about when WaveRider had their big public debut that I attended in Vancouver, BC in 1997 or 1998.

This final line of my posting warrants some additional comment. WaveRider first unveiled the idea for the LMS product line at a press conference in Vancouver, BC in 1997 or 1998 which I attended (it was sometime shortly after the debut of my Wireless Data Developments column in Boardwatch Magazine.) The nearly-ready-for-sale product that WaveRider announced at the time was a 2.4 GHz version of LMS. At that press conference, WaveRider spoke of the LMS product line eventually expanding into other license-exempt bands such as 902-928 MHz. I felt at the time, and (I think) stated this to WaveRider personnel that were present, that WaveRider's 2.4 product was entering a very crowded market for 2.4 GHz equipment, and that the announcement of a 902-928 MHz version of LMS, even though still in development, was more significant in the long term. In the years since, as far as I can tell, that early assessment has been borne out. WaveRider only began to gain notable success in the marketplace with the emergence of the 902-928 MHz LMS.

Steve Stroh

Copyright (c) 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

802.11 Planet article - Can Public Hotspots Make Money

Link To Article

Reporting from the 802.11 Planet conference in Philadelphia, Brian Morrissey writes:

Begin excerpt
At the 802.11 Planet Conference here, the message to the public hotspot market is: Forget John Q. Public, unless he's carrying a briefcase.

Analysts said the market for public hotspots, despite all the hype, doesn't appear all that large, except for niche providers for business travelers. According to Cahners InStat analyst Gemma Paolo, today there are just 41,000 hotspot subscribers. Even with that figure growing to 40,000 by 2006, the economics of the industry are difficult.

"For the pure plays, there's probably room only for a handful to survive," said Yankee Group analyst Adam Zawel. "We've been playing with the numbers, and then you've got all these players dipping in, and it just gets really thin."
End exerpt

I just enjoy the heck out of these types of stories proclaiming the inevitable death of the hotspot markets. I feel very strongly that the Public Wireless Access Points (PWAPs) will suceed and become totally ubiquitous in nearly every public area, and that, collectively many fortunes will be made. (That's not to say that every PWAP venture will succeed - far from it. There will be lots of casualties as there is a "dot com feeding frenzy" feel to this trend as way too much clueless, desperate-to-cash-in-on-the-next-big-trend venture investments are being pumped in.) I feel that every analysis of the PWAP trend that I have seen to date is very fundamentally flawed because they do not take into account the following factors:

• The cost of backhaul bandwidth to each every individual PWAP is assumed to be equivalent to the cost of a dedicated T-1 line. That assumption is fundamentally flawed. The primary requirement for a PWAP is that it offers some reasonable amount of bandwidth wirelessly; IE, a PWAP just turns on their wireless-equipped laptop, logs in / pays / automatically authenticates, and begins their Internet activity. That backhaul could easily be bonded dialup (roughly equivalent to 128 Kbps), ISDN (128 Kbps), DSL (up to 1.5 Mbps or faster), and most likely... wireless! (The last is especially true if the wireless backhaul is provided by a Wireless ISP with an existing network; the additional bandwidth demands of supplying a PWAP are a negligable addition to their existing customer base's bandwidth requirements.) All of these transports would provide more-than-adequate bandwidth to a PWAP at an order of magnitude lower cost than a dedicated T-1 line.
  
• The average PWAP customer uses a laptop to access the Internet. In 2002 and before, true. In 2003 and beyond, false. Current models of Palm's PDAs have a Secure Digital (SD) expansion slot, and SD 802.11b cards will soon appear. Once PDAs begin to ship with embedded "no pain" 802.11b (no significant change in form factor or battery life), I feel that PDA users will rapidly become the largest group of users for PWAP services. (This assumes that pricing and other business issues begin to approach reasonable - pricing for purely commercial PWAP service is way too high at the moment.)
  
• It's assumed that PWAPs are limited to "premises coverage". Bzzt. Bad assumption, extrapolated from typical 802.11b access point coverage which are specifically designed to limit coverage area - inefficient antennas, low transmit power levels, etc. 802.11b technology is easily capable of providing "much larger than premise" coverage. If such "wider coverage area" 802.11b access points are implemented, the pool of potential PWAP customers goes way up.
  
• We're now seeing the emergence of hot zones - coverage of outdoor areas such as public parks. We'll rapidly see the emergence of Hot Zones in places such as campgrounds, truck stops, marinas, private airports, gas stations, nearly anwhere someone would want to access the Internet (enabled in such places by alternative backhaul transport described above).

My standard disclaimer (modified): I consider this a significant development in the Broadband Wireless Internet Access industry. I've written extensively about this development, in significant depth in my newsletter FOCUS On Broadband Wireless Internet Access.

Steve Stroh

Copyright 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

FCC NPRM 02-180 To Commercialize Use of 71-76 GHZ, 81-86 GHz and 92-95 GHz Bands

The FCC has initiated Notice of Proposed Rulemaking (NPRM) FCC 02-180 to establish rules leading to the commercial use of the 71-76 GHz, 81-86 GHz and 92-95 GHz bands. The FCC's terminology for this spectrum is "W Band". What is unique about the W Band is that to date there has not been any commercial use (other than an experiment by Loea Communications) - an example of the fabled, long-thought-to-be-extinct "Virgin Spectrum". Among other approaches, the FCC will consider the possibility of making some portion or all of these new bands license-exempt.

I consider this a significant development in the Broadband Wireless Internet Access industry. I will be analyzing this development in significant depth in a feature article in my newsletter Focus On Broadband Wireless Internet Access.

Some pointers:

• Official FCC Web page (this story will eventually scroll off), see links to separate Commissioner's statements:http://www.fcc.gov
  
• Official FCC Announcement (PDF) - FCC Initiates Proceeding to Promote Commercial Development of Spectrum in the 71-76 GHZ, 81-86 GHz and 92-95 GHz Bands - http://hraunfoss.fcc.gov/edocs_public/attachmatch/DOC-223386A1.pdf
  
• Good dc.internet.com story by Roy Mark - FCC Launches Effort to Free Spectrum - http://dc.internet.com/news/article.php/1365371
  

Steve Stroh

Copyright (c) 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.

Administrivia - Blog Updates Via email

One of the more interesting capabilities of Radio Userland was a configuration line for setting up automatic checking of an email account for blog updates. That would be incredibly handy when attending conferences, travel, etc. It was a bit maddening to get it configured just right, but it's working now. I expect to make heavy use of this feature.

Steve Stroh

Copyright (c) 2002-2004 by Steve Stroh. This article originally appeared on my original Broadband Wireless Internet Access Weblog hosted on Radio Userland.