Fixed Wireless Equipment

Redline Dares To Cross NLOS Limits

Our quest for inexpensive, small, but reliable non-line of sight wireless equipment continues. This time, we're off to field tests in Canada by way of Southern California.

by Gerry Blackwell [February 26, 2002]

The recent Broadband Wireless World Forum in Anaheim, Calif. was a kind of coming out party for Redline Communications, Inc., an interesting new player in the wireless network equipment market.

Mike Moldoveanu, the company's current chief technology officer, and partner Radu Selea, vice president of software engineering, founded the Toronto-based firm in 1999.

At the show, Redline introduced what hopes to be it's breakthrough product, a non-line-of-sight (NLOS) U-NII-band point-to-point/point-to-multipoint Access Node 50 (AN-50).

Redline does not pretend this product offers the NLOS market a Holy Grail combination—cheap, small, and stable equipment. At least, not yet anyway. It's priced at $3,500 and fits a 19-inch rack—a design feature requested by service provider and corporate customers, the company says.

What the technology does promise is tremendous performance. In line-of-sight mode, it boasts raw data rates up to 72 Mbps with base-station-to-subscriber ranges up to 30 miles. In NLOS mode, the company says it can deliver net data rates of at least 16 Mbps over distances of at least a mile.

And Redline has already begun the process of reducing the key components of its technology to an application specific integrated circuit (ASIC), which will bring the total price of customer premises equipment down below $500.

The ASIC won't be ready until the end of the year, though, says vice president of marketing Keith Doucet.

Capitalizing on NLOS
Redline won't say exactly how much capital it has raised, but it has some substantial backers. The largest investor is Telemedia Corp., an investment holding company with an asset base in excess of $500 million.

At the show, Doucet handed visitors to the company's booth a competitive analysis matrix. He suggested they go around to other vendors and ask them to fill in their own system specifications to see how they stacked up against Redline.

Doucet wasn't sure himself what they'd find. In summarizing the results, he basically sums up the product's claimed advantages.

"We're offering much better throughput, we're standards compatible and we're completely non-line-of-sight," Doucet says.

"Apparently a lot of our competitors were saying, 'Where did these guys come from?' Customers were coming back to us and saying, 'If it's true that this is what you're building, then you guys are hot!'"

It's a big if. The first thing to note about Redline is that it's not building much of anything yet, although Doucet says the company has now selected a manufacturing outsourcer and placed orders. It expects to begin shipping in volume by May 2002.

In the meantime, even test equipment is in short supply. "Our problem right now, because we're a start-up, is that we have more customers than equipment," Doucet says. (The reverse, it occurs to us, might be a more of a problem.)

As for whether the technology can do what Redline says it can, that is indeed the $64,000 question.

Trial run
The company announced recently it had successfully completed "field trials" of the AN-50. But they were not full-scale customer field trials—those are just now getting underway. They were more field demonstrations, conducted for prospective service provider customers and the analyst firm The Strategis Group Inc.

The demonstrations appear to have been impressive as far as they went. Doucet set up one antenna at the company's offices in Markham, just north of Toronto. Placed on the roof of the two-story building, it was about 25-feet above ground.

The other antenna he set up on a tripod in front of his home about a mile away in Richmond Hill, another Toronto bedroom community. This one was only about seven feet above ground, he says.

The Markham building is surrounded by others of the same height or higher, effectively blocking line of sight. And there is also a long line of rolling hills between the two test sites.

"The bottom line is, we could not see any part of the other antenna," Doucet relates. "But we turned it on and instantly had 16 Mbps of throughput—that's net data rate. It climbed up to 18 Mbps before we finished."

Of course, a single test between two sites, however hostile the environment, does not prove NLOS performance across the enormous range of environments service providers are likely to encounter.

Still, Dudley Freeman, president and founder of Uniigo Communications, a Needham, Massachusetts-based wireless ISP, was impressed enough with a similar demonstration that his company is now field testing the technology.

Uniigo in fact is one of four service providers currently running field trials of the AN-50 or about to begin. The others, all near Toronto, are:

• ICS (Integrated Cable Systems), a division of Oshawa-based Cygnal Technologies Corp.
• Blink Communications, a division of Oakville Hydro Corporation in Oakville
• Alliance Communication Cables Inc. in Mississauga.

Uniigo's Freeman also agreed to wax enthusiastic about the AN-50 in a Redline press release.

"The product is extraordinary," he is quoted as saying. "We have been looking at several technologies in this market space for some time, and what we have witnessed at Redline is a system that is completely in a space of its own."

What is Redline doing right to enable the kind of performance it can apparently deliver? There are three key bits of intellectual property that really make the technology sing, Doucet says.

The most important is the work the company did with Orthogonal Frequency Division Multiplexing (OFDM), developing the OFDM "data engine" incorporated in the AN-50 product. OFDM helps minimize multipath problems, a particularly vexing issue with NLOS networks.

Redline's implementation of OFDM, developed at the company's two R&D centers in Romania, the once isolationist eastern-European nation, involved "approaching it as a mathematical problem," Doucet tries to explain. More important than how they did it is the effect.

"OFDM," Doucet says, "has been considered a very delicate technology in the past. It needed expensive components to work properly. Cisco, for example, had introduced a product but [the technology] really stumped them."

"We approached this as a waveform mathematical problem and figured out how to make [OFDM] work with everyday components. As a result we have a very cost-effective approach to the problem."

Step down to ramp-up
The other two key technological advantages Redline claims for the product come from the company's implementations of Acknowledge ReQuest (ARQ) and adaptive modulation. Both make AN-50-based networks more robust than competitors', the company says.

ARQ is a network-layer bit error-correction technology that is now a proposed standard of the IEEE 802.16 committee, in which Redline is very active.

"If data bits don't get through because of some obscure propagation problem [our technology, using ARQ,] can correct the problem before the network has to deal with it," Doucet explains. "This is transparent to the IP network world."

Adaptive modulation is a more familiar technology. A network that implements adaptive modulation can change modulation schemes to suit propagation conditions.

The more sophisticated the modulation scheme, the more data a given chunk of bandwidth in a network can carry over a given distance. But in a wireless network, the higher-capacity the modulation scheme, the more sensitive the connection is to propagation problems.

The Redline product can dynamically step down from the highest-capacity scheme—64QAM (64-state quadrature amplitude modulation)—through 16QAM, quadrature phase shift keying (QPSK) and biphase shift keying (BPSK).

With each step down, data throughput is reduced, but the connection is at least maintained.

According to Doucet, in most other implementations of adaptive modulation, network administrators must make a decision based on observed propagation conditions about which scheme to use for each subscriber, and then set it more or less permanently.

The Redline product changes modulation schemes automatically in response to prevailing, sometimes temporary, conditions—trees getting wet, for example. So the subscriber can enjoy optimum performance most of the time, yet still maintain a connection when conditions deteriorate.

Most other implementations of adaptive modulation also only work in the downstream channel, Doucet points out, while the Redline implementation works in the upstream path too.

Redline developed all components of the AN-50 product, holds patents on some, and will also manufacture all. "This way we can balance and fine tune each element to work the way it's supposed to," Doucet says.

While it may not be the NLOS Holy Grail, Redline believes there is a healthy market among service providers targeting business customers, especially in areas not under served by DSL and other wire line broadband access services.

And by the end of the year, this company may have something a lot closer to fulfilling its quest. Definitely worth watching.

—End