wireless isps ponder: to wimax or not?

BEYOND ALL PREDICTIONS AND ODDS, broadband wireless Internet service providers (WISPs) are not only surviving but doing well. Although the vast majority is small, with 1,000 to 2,000 subscribers, collectively they represent a significant chunk of the 30 million broadband subscribers in the United States and Canada.

Now the industry is faced with an interesting new technology, WiMAX, which is posing some challenging questions for these smaller service providers. WiMAX, the popular name for a technology that is based on the 802.16e standard now being finalized by the IEEE, promises to be highly standardized, something the industry has wanted for some time.

However, equipment is not yet available, and the first products, point-to-point only, are not likely to appear until fall of 2005. In the meantime, some vendors are offering interim proprietary products they say can be upgraded to true WiMAX with as little as a software change.

The question the industry is pondering is whether to wait for the “real” WiMAX or deploy one of the proprietary technologies, which promise to deliver improved performance.

Wireless service providers have been operating fairly reliably using equipment that has evolved from the 802.11 Wi-Fi technologies, a kind of Wi-Fi on steroids. They enjoy ranges up to 10 miles in favorable conditions and less than three miles in densely urban settings. Capacity is usually from 3mbps to 10mbps.

Some providers have focused on small business, and those are the ones likely to be enjoying the best financial circumstances, having the ability to charge upwards of $100 to $200 per month for service, compared to the $40 or less that is a typical consumer rate. Whatever the market, many of these companies have endured, and some have enjoyed the good fortune of being acquired by larger service providers able to exploit economies of scale. For a few, acquisition is the business plan.

WiMAX has been designed to deliver ranges up to 30 miles in unlicensed spectrum with throughput in the range of 45mbps — in other words, a significant improvement from today. Because the 802.16e standard only recently has been solid enough to begin product designs, chip manufacturers probably will not deliver product until the middle of 2005. That would mean a very quick turnaround to have finished products in the market before the end of the year, but that’s what some wireless vendors say they will deliver. Many believe that 2006 is a more likely delivery schedule.

WiMAX equipment is likely to be expensive. As much as $1,000 has been quoted for customer premises equipment, a price that compares unfavorably to the $200 or less typically paid for more traditional equipment. Also, service providers must choose between several alternatives: More-powerful versions of current technology, proprietary takes on WiMAX or real WiMAX.

The issue is whether there is any benefit for wireless ISPs in being truly WiMAX. Charles Wu, president of CWLab, which provides product testing, performance analysis and engineering for wireless vendors and service providers, and former owner of a successful WISP in downtown Chicago, is not impressed with the advantages of waiting for the standard. He asks, somewhat facetiously, “Do we buy the $800 WiMAX CPE that will hurt our interests, because then we will have WiMAX for real, or do we just deploy and get customers, so we have all the customers in the area?” He adds, “So what if it is not Wi-Fi or not WiMAX, but is a proprietary system that gives us everything that WiMAX does without WiMAX.”

One reason to wait is performance. “We’ve heard that there is increased security and embedded quality of service,” says Travis Johnson, president of Microserv Computer Technologies, a wireless ISP serving a broad area in Idaho.

NextWeb, a large wireless ISP in Northern California, is already planning to offer VoIP using “preWiMAX” equipment by Axxcelera Broadband Wireless. “It is basically long-reach unlicensed broadband equipment,” says Eric Warren, director of marketing and business development, NextWeb, which just joined the WiMAX Forum.

What would be important with WiMAX would be “if it can do that quality-of-service prioritization that people are talking about. That would be good for offering voice and other usage-based services,” says Warren. “And I can see the eventual evolution of the industry, like Wi-Fi. Wireless was around a long time before it became this brand and hype machine.”

The purported 45mbps throughput of WiMAX is an attractive carrot. WISPs worry that, with cable upgrading to 3mbps downstream, they will end up looking slow.

“We found that, regardless of their plan, 1 meg, 3, 5 or 7, the average use was the same amount of bandwidth. On download it took 1,000 customers to use 10 megs,” says CWLab’s Wu, whose customers were mostly businesses. “The first week it is all downloading movies. Then, after that, it’s back to e-mail and CNN.” One advantage of higher bandwidth is, “it does get them off the highway faster.”

Replies Tom DeReggi, president, RapidDSL, a wireless and DSL service provider based out of Germantown, Md., “Maybe that’s right. Today we do not have to offer 3 megs, but next year or the year after, yes.” It is important to offer more and more capabilities in order to migrate customers to advanced products, he adds.

Adds David Post, network administrator, Frognet, a wireless ISP in Ohio. “When people decide between us and cable or the phone company, the difference is that we provide static IP addresses.”

Another potential opportunity with WiMAX is the ability to offer a service that resembles mobile but with more bandwidth. However, “what does it mean to go from fixed to being able to do mobile?” asks Rebecca Shipman, wireless access strategist, Bell Canada. “There are so many more things to think about than the whole fixed-wireless assessment. So you have to weigh out all the risk factors in that. So I would go proprietary today and keep investing. There are just a lot of variables to assess with WiMAX.” She emphasized that Bell Canada is still forming its broadband wireless strategies.

Shipman believes incumbents will test the waters of this new kind of mobile, possibly using an advanced proprietary broadband fixed wireless, such as one offered by Alvarion Ltd. However, such testing would be expensive for a small wireless ISP.

WISPs are being subjected to “scare tactics,” Microserv’s Johnson adds, “that if we don’t have WiMAX, we will not have security.” He discounts that speculation, but says his customers have expectations that Microserv is providing the best technology. “So clients don’t lose faith in you being on the leading edge,” WISPs should consider some kind of upgraded service, using standard WiMAX or a proprietary system.

Claims they are “WiMAX-ready” put out by various vendors is not helping the decision-making, WISPs say. “My big peeve is with the WiMAX Forum,” says Wu. “They mentioned that Trango is supporting WiMAX, and they are not supporting WiMAX.” Trango has announced a proprietary product that delivers many features promised for WiMAX.

One reason cited to wait for WiMAX, at least a short while, is that WISPs can use the point-to-point equipment, which is expected first, for backhaul. “What we want to use WiMAX for is its backbone to our current wireless network,” says Johnson, “but also as a replacement for point-to-point links and to replace T1 to commercial customers. We have line of sight to most of them, and it would cut our costs tremendously.”

“I would have gone to you and said, ‘You have a system now. Why not use it?’ For point-to-point, Trango announced a product for $2,500 [for access points] that will do 45 megs. It supports header tags, for just $2,500,” counters Wu.

Wu says Trango will offer a proprietary 802.11a product “with a WiMAX feature” set next year that will have CPE priced at $300 to $400. “That’s at 45 megs with quality of service,” says Wu. “That goes back to my original argument as an operator. With $800 CPE I save money, but with $400 CPE I save even more. I like standards, but I also like my pocketbook.”

One tangible advantage of WiMAX hopefully will be “plug-and-play CPE,” says Warren. “Today you have to buy CPE from the same vendor. Hopefully we can match service equipment with access points from any vendor.”

And then there’s the product availability issue of WiMAX products. Most believe that true WiMAX products will not be available for a year, and they will be only point-to-point. “And none [of the vendors] has committed to point-to-multipoint.” says Johnson, who has looked at equipment by Proxim Corp. “It could be four years.”

“It’s hard to see how these vendors are going to deliver something soon when the silicon is just getting designed now,” says NextWeb’s Warren. “People haven’t spec-ed out systems based on silicon because silicon is not even ready, and that’s how the whole industry works. When people have products and want us to try them, we will be glad to.”

Although WiMAX products have been quoted in the $1,000 range just for CPE, costs will come down. Does that encourage service providers to wait for standard equipment? “Yes, they will not be $1,000 forever,” says Johnson, who does not intend to use the equipment with customers right away. “But, if I have a chance to pick up 1,000 customers now with a proprietary system, maybe I will do that rather than wait two years.”

The price of CPE is critical to small WISPs. Most report that, when they are able to drop installation to about $100, roughly on par with DSL, they see an immediate uptick in subscribers. This means they are leasing rather than selling the CPE. The more expensive the CPE, the more time to get an ROI.

NextWeb is planning to spend more for WiMAX equipment from its vendor, Axxcelera. “We’re not saying that WiMAX is going to significantly change our business,” says Warren. “In fact, it probably is going to be more expensive in the beginning.” NextWeb will make the expenditure “because we expect to see benefits from the technology and are cautiously optimistic about those benefits being the same as they are being touted now.”

One engine that could drive better pricing is volume. Although ILECs in North America may not do large deployments of WiMAX, “BT is going live in Northern Ireland. There are big players in Asia and northern Korea that are going WiMAX,” says Shipman. “For BT, there is a lot of land mass that they can’t cover with DSL. In Korea, there are lots of places that they can’t get into with wired solutions.”

Nevertheless, says DeReggi, “I would like to have the cash today, and a lot can happen in two years.”

Adds Shipman, “WiMAX may not happen in two years. It may not happen, period.”

There are also fears that proprietary equipment may not migrate as smoothly as promised.

“I had a session with a manufacturer in broadband,” says Wu. “If you look at the current chipset in the evolutionary programs and ask, ‘Is it worth it to go ahead?’, consider that there is no software-defined radio out there — whether frequency-hopping, 264-carrier OFDM (orthogonal frequency-division multiplexing, a key WiMAX technology) or other OFDM — it’s not in software.” Therefore, an upgrade to full WiMAX could mean much more than simply uploading new software and rebooting.

The uncertainties create tough choices for operators. “So should ISPs dig in their heels and resist, or invest in the current product that is out today knowing that it may not be compatible?” asks Johnson.

There are numerous issues to be weighed regarding whether being nonstandard has disadvantages or advantages for competition. “Do I want to put in a product that is standard, so the ILEC can come in and take over? If I have proprietary solutions, then they can’t,” says DeReggi. But he answers his own question, saying, “On the other hand if you have a proprietary device, what if [the ILEC] comes in with a standard device that blows yours out of the water?”

Also, some worry that nonstandard equipment might limit their opportunities to get acquired by a large player.

Wu says, “If I have 5,000 customers, and a little proprietary solution running and happy customers, do you [the large acquirer] care what equipment I use or do you care about my customer? So, if you think the ILECs will not buy you out unless you have standard equipment, that’s not so.”

Post says, “They just look at customers. However, if the equipment is compatible, that could put you at the top of the list. But equipment will not make or break the decision unless the numbers are so huge that they would break the deal.”

Share this article: Email, Slashdot, Digg, Del.icio.us, Yahoo!MyWeb, Windows Live Favorites, Furl
Add this article feed to: RSS, My Yahoo, Newsgator, Bloglines