Broadband Wireless World Forum
By John Latta
(March 2)

WAVE Issue #0112

With a glut of long haul bandwidth and a scarcity of first and last mile bandwidth there are enormous market forces to raise the delivered bandwidth. Wireless is the 4th leg of delivery with phone (copper), cable (fiber and coax) and satellite the other three. At the center of the trade events probing and exposing the wireless issues, technology, markets and companies is the Broadband Wireless World Forum. We went last year and was even more impressed this year. Tim Downs puts on a top conference. It was laden with new insights, controversy and surprises. If we have one disappointment it is that the trade floor was only open 2 out of the 3 days of the conference. Tim promises us this will be fixed next year. A measure of the rapid growth in interest is that this, the third event, has gone from 800 to 1600 and this year to 3400 attendees.

The news of the conference was the emergence of the fixed wireless companies using unlicensed spectrum for the delivery of symmetrical broadband. These companies have the potential to reshape the delivery of broadband - both the first and last mile. This report will discuss these companies, their presentations, and the WAVE-Report will provide an exclusive table of the players in this segment. We offer both our assessment of this market and our usual WAVE Comments at the end.

We begin with some market background collected from various talks at the conference.

There are 750,000 business buildings in the US and only 10,000 to 12,000 are lit with fiber.

In Europe with 850,000 buildings less than 2% are lit (<17,000).

There is increasing community resistance to tearing up streets for laying down fiber, especially in Europe, where some cities have a total ban.

Long haul bandwidth is plentiful and local bandwidth is scarce.

There is no question about the demand for broadband, be it consumer or small businesses, the critical market issue is delivering broadband.

Provisioning and installing business broadband is a lengthy process - OC-3 in the Ameritech region takes 9 - 12 months.

Local rates for business broadband delivery have been rising 7% to 8% annually.

The ILEC carrier has >90% local market share and represents a major impediment to the development to a competitive market.

Most ILECs are anchored in 60+ years of copper.

It costs approximately $250,000 to light a building but only $20,000 to bring fixed wireless to it.

Infrastructure investment in fiber is 80% labor and 20% technology while in wireless it is 20% labor and 80% technology.

The surprise of the conference was the emergence of broadband wireless operators using the unlicensed spectrum. There are two bands (ISM - 2.4 - 2.4835GHz) and U-NII (5.25 - 5.35GHz & 5.725 - 5.825GHz). Typically these carriers use ISM for delivery from the cell site to the home/office and U-NII from the landline termination to the cell sites. There remain two key issues using these frequencies: interference and LOS (line-of-site) propagation restrictions. These were only passingly discussed by the presenters. As one presenter said - "we have not discussed the "I" word."

As we assess these companies it is important to put into context the current VC and investment environment. Very difficult. Capital spending by the telecommunications carriers has dried up. The stock for telecommunications carriers has tanked and this is not a good time for a start-up. DCLECs (Data CLECs) and CLECs (Competitive Local Exchange Carriers), competing against the ILECs (Incumbent Local Exchange Carriers), are seen on a death march. Reeling from the crash of the dot.com companies VC's will invest only in firms that have a rapid path to profitability and with first class experienced management. Further, wireless broadband delivery using licensed spectrum is dominated by companies with deep pockets, in part because they have paid dearly for the spectrum. Being a start-up using unlicensed spectrum is like entering the Wild West.

AIR2LAN

Jai Bhagat, Chairman and CEO, gave a confident and self-assured presentation. He was CEO of SkyTel before it was bought by MCI. AIR2LAN operates in both the ISM and U-NII bands. They currently deliver 4.5Mb/s of bandwidth and will reach 32Mb/s in 6 months. Presently the company operates in Jackson, MS. November construction was started in Houston and New Orleans. Late December launched Houston operations, it recently launched operations in New Orleans.. The company plans on reaching 37 markets nationwide by 2003.

AIR2LAN is on a fast track. It began operations in December 1999 and began the build out in March 2000. In May of 2000 it began commercial operations in Jackson, MS and raised $6m. Currently the company as 300 customers in Jackson, MS, there are an average of 15 users/account (4,500 users) and the typical contract for service is from 24 - 36 months. The distribution of bandwidths being acquired:

47% - 256Kb/s
26% - 512Kb/s
12% - 2Mb/s
9% - 128Kb/s
6% - 1Mb/s

Key attributes of AIR2LAN. It is serving a segment of the market that is underserved, its network deployment is efficient (don't have to light whole city like other wireless networks), it has efficient startup times (60-90 days after research is finished which selects markets by business density, provisioning time is typically 72 hours, and markets can be profitable in 16-20 months.

One of the reasons for the opportunity is that other technologies have been weak in implementation:

1. DSL has not been able to provide service in the market,

2. Cable modems are not in business markets; and

3. Existing broadband carriers are not providing a single solution which includes connectivity, scalability, provisioning, value added services, Web services, LAN design, mobility solutions but AIR2LAN provides this.

The company is currently seeking a $35m round and estimates that it can use these funds to become cash flow positive in its first 10 markets. Within a given market it estimates it can become cash flow positive in 22 months.

Kite Networks

Kite Networks presentation was given by Jerry Sullivan the President and CEO. Its network is based on only U-NII. The delivered bandwidth is 25Mb/s. Their first operational system is to be turned up on March 1, 2001 in Phoenix. By Q4 2001 they expect to roll out plans for regional coverage that can lead to national coverage.

Kite Networks targets markets based on profitability because its not the numbers that are signed up, its how profitable the customers are. They do this with pre market and demographic surveys. Telemarketing and e-mail services are also used.

Jerry stated, "People will pay for content, but if you don't have the network to support it, it won't get there. They will demonstrate, as their national deployment goes forward, that basic access has to be in place then their network will move into VoIP, content distribution, VPN and on into IP TV, distance learning/medicine."

Jerry feels that video will be a major use of these network connections. With high quality video delivery the interaction between individuals in remote locations will be radically transformed from the poor video conferencing that exists today.

Broadband2Wireless

The company will do a formal launch in April.

Thus, the presentation at BWWF was a preview. Paul Adams, the Cochairman, has an ambitious mission statement - for BB2W to be the premier, low cost provider of next generation broadband wireless Internet. This goes beyond the other companies in that he sees mobility included: fixed, semi-mobile and mobile, as part of this mission. He was also the speaker to make the most of provisioning. His OSS (Operational Support System) provides the ability of the prospective customer to do an online order (which is provisioning at the same time), to dynamically perform pre-qualification, i.e., determine if the customer can be served with the existing infrastructure, and from this output create the CPE (Customer Premise Equipment) and installation instructions. This package is then sent via FedEx to the customer and within a short time they can be connected to the network. The CPE goes within the customer's building. All of this accomplished within 24 hours. This has the prospect of a great consumer experience if BB2W can deliver.

Paul stated in his talk: "Right now they are operational in 12 cities nationwide, all tier one - high densities, high ROI. Another 7 cities to be completed by the end of this year. Service in 2 other countries this time next year. Focusing on underserved markets. Their semi mobile products will launch in the next 8 months once their geographic coverage expands."

BB2W claims today it delivers symmetric 3Mbs/s and will next year provide 54Mb/s.

Paul was both cocky and correct when he declared: we can do this without paying $1b for a spectrum license.

He also stated that the ROI per microcell, per hub site, per city, per region, payback per cell site is about 6 months or only about 2% penetration.

An important part of the implementation of BB2W's strategy is the role of GIS/mapping. Paul showed a color-coded map that has population density and the following conditions: cable modem access, DSL access, both cable modem and DSL and none. He claimed this information is updated daily. From this they plan the rollout of services in the following priorities:

Phase 1 - regions with No DSL & Cable Modems
Phase 2 - regions where customers can migrate from DSL
Phase 3 - regions where customers can migrate from Cable Modems
Phase 4 - regions where there is a strong interest in roaming and mobile applications.

The economics of broadband delivery are quite interesting:

Cable modem Cap Ex - $1000-1500/home passed,
DSL - $800,
Broadband wireless under 100 (about $20 per home passed in tier 1, $40 for tier 2 markets)

BB2W was the only company to present that stressed mobile support.

Airband

Airband is initially focusing on U-NII. This requires both an outside antenna and internal CPE equipment. The delivered bandwidth is symmetrical at 512Kb/s, 1Mb/s and 2Mb/s. The company has an operational Dallas infrastructure with plans for Houston; Washington, DC and Chicago.

The company is providing Internet access solution (1.5-10Mbps) by:

targeting black holes (no fiber, DSL or cable),
mostly small businesses and corporations (200-500 employees) in Tier 1 cities using direct sales organization, telesales and beginning on channel development (with Cisco).

The company claims it has 125 years of wireless operational experience.

We have assembled a table with the details from the companies in this market space.

http://www.wave-report.com/other-html- files/BroadbandWirelessCarriers.pdf

Assessment

There is a broader context for these companies and the market dynamics. Consider the following. Noticeably absent from these presentations were claims on voice. We read this as a direct reflection of the unlicensed spectrum. It is not possible to provide quality competitive voice given the uncertainties of interference and propagation. More importantly, these companies are directly attacking the weaknesses of broadband cable and DSL. These legs for delivering broadband provide bandwidth less than 1Mb/s and the pattern of customer abuse, which rivals that of the airlines. The industry has another term - bad customer experience - but these are just softer words for the same thing. Using databases with assured delivery, self-install and the responsiveness of a hungry company the fixed wireless companies described here are promising a delivery and user experience that customers crave. At the same time the companies are being very selective in their targets - the classic market cherry picking. By going after the bandwidth starved, i.e., no DSL, no cable modem, and the upper scale demographics, the potential for consumer sign up and retention is great. This only further improves the prospects for ROI and positive cash flow.

In addition, with some of the companies focusing on 2nd and 3rd tier markets it is much less likely that competitors, at least those with unlicensed spectrum, will encroach on a well penetrated consumer base - the cherries have been picked. What then are the underlying dynamics? We see that the basis for value in these new companies is NOT the infrastructure they have built or the fees they paid to get it, i.e., free. Using unlicensed spectrum and relatively low cost equipment, compared to other technologies, it is not the network but the loyal customers that drive the value of the company. We expect that natural consolidation will take place in this market and the consolidators will be those who own the licensed spectrum. What they will need, when the economics are favorable for the technology they use, are customers. The most valuable customers will be those who are loyal, i.e., with low churn rate, and are tightly aggregated and thus best suited for wireless coverage. It is our prediction then that the value to unlicensed spectrum operators will be based on a cost/customer basis.

What then are the holes in this argument? First, one should never underestimate the FCC. It would only take one petition to the FCC that states that these uses for spectrum undermine the intent of payment for spectrum, which has economic value. It would certainly be a mess for the FCC if it went back and sought funds for an operating company in an unlicensed band. This did not happen to the broadcasters when they got free ATV spectrum for digital broadcasting. Second, if services are delivered, at substantially higher bandwidth using U-NII, it would be impossible to deliver the same bandwidth using the lower frequency and more restrictive MMDS.

More from BWWF Session:
Contrasting Non-Line of Sight Technology Approaches

It was refreshing to see a session address one of the core issues of Broadband Wireless - how to overcome the line of site restrictions, in part, due to foliage penetration at these frequencies. However, it was one of the surprises of the show. The entertainment came from Hatim Zagloul, the Chairman and CEO of Wi-LAN. He stated that his company would sue anyone that infringes on their OFDM patents. He cited a minor victory in the suit against Cisco when they did not show up in court at a recent hearing. Further, when Iospan was describing its technology he stated that the only reason they have not be sued is that they have no products shipping. The area of non-line of site propagation is an important topic and IP plays a critical role. In addition to the market developing over years it is clear the courts will so be engaged.

WorldCom Broadband Solutions

Its President and CEO, Robert Kaiser gave a keynote. WorldCom is one of the largest holders of licensed MMDS spectrum. Key points include.

Rural and suburban markets are starved for bandwidth.

DSL and cable are slow to market and technologically limited.

Fiber Optics has prohibitive costs for many seeking bandwidth.

MMDS is at 2.5 to 2.7MHz and has 200MHz of bandwidth.

The MMDS spectrum is under consideration as a 3G spectrum and MMDS could get displaced. However, WorldCom feels that it will not get disrupted in this service.

WorldCom has service operational in Memphis and technology trials in Baton Rouge and Jackson, MS.

WorldCom expects to serve 30 cities by the end of 2001. The planned service includes:

Up/down
384Kb/s/384Kb/s
768Kb/s/512Kb/s
1Mb/s/512Kb/s

With the following features

Hosting storage of 10MB
Multiple e-mail accounts
2 DNS names
20 static IP addresses
Premises installation

Radiant Networks

Nadeem Siddiqui, its Chief Operating Office, spoke of the value of Mesh Technology in improving service in wireless broadband. The key to the mesh topology is that radios not only receive but pass on information. The connection between the radios is based on narrow beams for point-to-point links. These links are controlled over 306 degrees. The node and the base station are the same equipment. The benefits are:

95% of the customers can be reached in a given area when there is only 1% penetration on the installation.

Customer data rates can reach 25Mb/s, symmetric, and can support up to 100% of the customers all with 4Mb/s symmetric.

It is claimed that the network cost for a 25Mb/s symmetric connection is $1,000.

ART

ART has another disruptive approach for delivering broadband. Its BroadStream service operates in Dallas/Ft. Worth, Houston, Los Angeles, Phoenix, San Diego, San Jose, Seattle, Portland and Washington, DC. The service provides wireless 100Mb/s bandwidth for $1000 per month. It uses LMDS at 39GHz.

At the conference ART announced that it has an agreement in which it will provide fixed broadband service to Cable & Wireless for 10 years.

IEEE 802.16 Standards

Roger Marks of NIST, along with others on the 802.16 group, gave an excellent overview of the standards setting process and its status. This is a fast track.

802.16 Task Group 1
Standard Air Interface for Fixed Broadband Wireless Systems
Both PHY and MAC layers
Final draft due to the IEEE Standards Board - 3 August 2001

802.16a - Task Group 3
MAC for licensed bands from 2 - 11GHz
PHY proposals due January 2001
MAC enhancements due in consolidated addenda to 802.16 MAC due May 2001

802.16a - Task Group 4
MAC and PHY for Unlicensed
Draft Standard - May 2001

WAVE Comments

The LMDS and MMDS technologies are more mature that those in the unlicensed bands. However, with the 3rd generation of these technologies - many thought that this could be the year of broadband and others said it is too early. The technology has and continues to be too expensive for the operators to have a viable business case. Many commented that the situation is different in Europe where broadband wireless has shown a greater presence. However, in Europe the cost of an E1 is 3X that of a T1 in the US. Thus, LMDS (largely for businesses) and MMDS (focused mostly on the consumer) have yet to get widely deployed in the US. WorldCom spoke of its intent to roll out by the end of 2001 and Sprint, which has delayed its purchase of equipment numerous times did not speak at BWWF.

While the license owners dawdle, the unlicensed pioneers forge ahead. In many respects they have little to loose. With no prohibitive license costs from auctions, they are free to carve out markets, rush to install and rapidly build customer bases. Over all of this was the black cloud of the stock markets. The telecommunications sector is caught in a dilemma. It lives off CAP EX to build infrastructure but with the markets seeing this sector as a black sheep it is hard to get funding for build outs. All the equipment vendors we spoke to were singing the blues on the lack of orders. This is a situation that we believe will not exist beyond 18 to 36 months. Quite simply the demand for broadband is too great, fixed wireless has a compelling case and the ROI favors wireless. We believe that the pause will be a good one for the industry. It allows time for the equipment to continue to mature and the prices to reach economically favorable conditions. Further, the potential for symmetric 3Mb/s service wireless will blow away cable modems and DSL. One of the reasons is video and what we believe will be a major market driver - casual video. Broadband technologies available today are anchored in the past legacy of installed infrastructure and services. With that maybe fixed wireless can bring the prospect for the end of broadband customer abuse.