Among the various markets touted for WiMAX, the large enterprise is the least talked about. For early stage WiMAX, the most immediate markets are seen as last mile in underserved regions, public safety, plus backhaul for cellular operators (see separate item) and small to medium businesses (SMBs); while in the medium term all the attention is on mobile access applications. Large corporations are naturally cautious of new technologies and have complex and high performance requirements that first generation 802.16 may not meet, argue many. At least one major operator begs to differ, and is seeking to build a significant business around WiMAX for the enterprise ‘ this is AT&T, which in the run-up to its acquisition by Regional Bell operator SBC remains as enthusiastic as ever about broadband wireless, although its strategy has shifted somewhat to accommodate the priorities of its parent-to-be.
Early adopters of pre-certified WiMAX technologies have been heavily focused on the SMB sector. Until a low cost subscriber unit emerges, and standards drive overall network costs down, it is hard to create a business model for WiMAX as a consumer access technology outside of areas that lack any cable or DSL build-out. But large enterprises require capacity, support levels and guarantees of quality and security that may be beyond the scope of start-up providers, especially using first generation kit, or that eat heavily into profit margins. The SMB, which is typically outgrowing DSL but daunted by the cost of leased lines for its backhaul and access requirements, is a far more attractive market in the short term, especially as premium rates can be charged for services like VoIP and VPN. Companies like Towerstream, NextWeb and Airband in the US or Libera in the UK are examples of the importance of focusing early efforts on SMBs.
For larger operators, of course, the grander scale of the big enterprise requirements is more appealing than selling to and supporting large numbers of SMB clients, and the major providers already have the skills and support systems in place for big projects. But their target customers remain cautious about new technologies, particularly wireless ones, which always carry the association ‘ real or perceived ‘ with poor security and reliability. AT&T, in a recent briefing on its WiMAX strategy, came out with the most aggressive statements of any operator to date on the critical topic of security, claiming 802.16 lines are actually more secure than T1, and seeking to squash other perceptions of enterprise WiMAX that it believes are outdated or wrong.
This is important for the carrier, which generates a large proportion of its business from national and international enterprise customers, offering them a wide range of managed services based on its extensive and modern fiber network, from redundant back-up to backhaul to high speed access and VoIP. All of these, AT&T believes, could be enhanced in price/performance, in many scenarios, by including WiMAX in the technology mix. In particular ‘ and despite potential conflicts of interest with SBC ‘ it is interested in WiMAX as an alternative to T1 in the middle mile between its fiber and the access. Just like the cellcos (see separate item), large enterprises are trying to reduce their high outlay on leased telco lines for backhaul, back-up and dedicated links between sites, at the same time as their data bandwidth needs are exploding and they are getting serious about VoIP. Some will start to consider making a capital investment in wireless equipment themselves, but most will be more interested in a managed service from an operator that can access licensed spectrum for improved quality of service (especially when access services are included in the mix); can promise to upgrade its technology as WiMAX goes through its rapid development curve of the coming three years; and can integrate broadband wireless with other platforms such as fiber and even emerging options like free space optics and wireless fiber, to deliver the best cost/performance combination for different sites with different priorities.
This is what AT&T aims to offer, ahead of most of its major rivals, though other companies with significant international enterprise services business are taking a similar interest ‘ from European fiber provider Colt to incumbents like British Telecom. AT&T’s advantage while it remains independent from SBC is that it has few local copper lines and so does not have the conflict of interest that the Bell operators or a company like BT suffers by introducing wireless, which could cannibalize the highly profitable T1 business. Indeed, AT&T’s original fascination with WiMAX sprang mainly from its desire to cut the huge sums it spends itself each year on paying Bell operators for access to their last mile networks. This motive may be diluted by the SBC acquisition ‘ even though AT&T’s enterprise services business is likely to be run quite independently ‘ but it still holds true in non-SBC territories in the US and abroad. And SBC, which is heavily focused on increasing its large enterprise revenues, will appreciate that the sacrifice of some T1 business may be a necessary corollary of providing a more integrated and all-round service, based on a combination of technologies, to high value customers ‘ and a means of preventing these sites dumping T1 anyway and going to another provider altogether.
Sprint Nextel, although its public plans for its 2.5GHz WiMAX network are mainly focused on consumer access, will also see opportunities to bolster its already significant enterprise presence through broadband wireless, combined with its long lines. But for now, AT&T is blazing the trail in this sector and highlighting many important issues that need to be addressed before the enterprise is fully confident about WiMAX.
Most important of these is security. After the bad experiences that companies had with Wi-Fi, which only this year achieved corporate class security ‘ with the addition of critical authentication and encryption mechanisms like AES ‘ there is a general assumption that wireless networks are unsafe. On the contrary, says Sandy Brown, AT&T’s vice president for product management ‘ ‘the reality is that WiMAX wireless link encryption (AES-CCM), coupled with end-to-end network security techniques and Layer 2/3 encryption makes this connectivity actually more secure than traditional T1 and DSL circuits.’
WiMAX is more inherently secure, in the base standard, than Wi-Fi was and supports essential standards like AES from day one. Its main security issues revolve around the fact that AES and other key methods are optional, rather than mandatory, in the fixed version of the standard, although this will change with mobile 802.16e. However, this concern is not one for the large enterprise, which will have the knowhow, internally or through its services provider, to ensure that only equipment with the maximum security levels is employed. The main problem for enterprise Wi-Fi, now that AES-compliant access points are widely available, is ensuring that all the client devices are up to date in security terms. This is not an issue for WiMAX, since first stage subscriber equipment for business needs to be corporately purchased. By the time end users are bringing in their own WiMAX kit in laptops or phones, the Forum should have made AES and other key technologies mandatory.
Of course, 802.16 only specifies security for the physical and media access portions of the network and, as with any network, threats remain at the higher layers, including denial of service attacks and viruses. This is one of the key selling points on which a managed network provider like AT&T seeks to attract customers, promising not only an inherently secure network link based on WiMAX, but the end-to-end security services a corporate provider can deliver, including IPsec secure tunnels between the customer premises and AT&T network edge.
With that sensitive issue dealt with, AT&T turned to the reasons why enterprise customers need to be thinking beyond T1. It said the key demands of its corporate client base are higher bandwidth for less cost because of exploding data access and backhaul requirements. The main necessities are, it says:
‘ Cost effective access diversity for mission critical applications
‘ Reduction of costs without giving up bandwidth, or getting more bandwidth.
‘ Fast provisioned access for disaster recovery.
‘ Shorter installation times for new lines.
‘ Reliable, secure access.
‘ A wider choice of technologies, both to increase price competition and support a wider diversity of applications and configurations.
WiMAX can support all these requirements in certain scenarios, although, of course, its performance is limited compared to fiber or wireless fiber alternatives. However, AT&T feels that, even in a configuration that achieves 10Mbps over a range of five miles from customer site to AT&T network, WiMAX fulfils the needs of a large proportion of customers. Fiber to the premises may therefore be an expensive but unnecessary luxury for many sites ‘ a conclusion Colt has also come to, as it seeks to broaden its access base by using WiMAX for the middle and/or last mile, in effect stretching its fiber reach to more customers without further build-out.
According to AT&T, 50% of enterprise customer access circuits are less than three miles from the point of presence, and a huge 98% are 10Mbps or (more usually) far less. Therefore WiMAX would offer better performance than current circuits for less cost than T1 leasing.
For those requiring backhaul over longer distances or at very high data rates, AT&T is also looking to integrate various other options into its mixture with WiMAX and fiber. WiMAX can be stretched to 25 miles in point-to-point mode for backhaul, matching Type II DS1/DS3, and other broadband wireless technologies can go even further. For high data rates, broadband wireless networks in high frequencies such as 18GHz get up to around 200Mbps, and the millimeter wave spectrum supports so-called ‘wireless fiber’ links at near-fiber speeds, with companies like WiFiber promising 1Gbps rates moving up to multi-gigabit in the coming year.
Another technology attracting renewed interest as a complement to WiMAX and fiber is free space optics, which achieves up to 155Mbps though over relatively short distances of a few miles. AT&T was one of the first companies to carry out extensive R&D in free space optics, which uses light rather than radio waves, and particularly in combining this with microwave in a highly reliable dual system. Since both radio and optics suffer from different vulnerabilities a dual network can use the link that is most robust in a certain situation, and then switch if conditions such as weather change. The initiative in this area has been snatched by start-up LightPointe, but we can expect to see AT&T gradually incorporating a host of different approaches into its enterprise platform.
In terms of its real world progress with WiMAX so far, AT&T has been running two trials in Middletown, New Jersey and a larger one in Atlanta, Georgia, with a view to assessing real world performance and applications and evaluating enterprise reaction to the services. In New Jersey it is using an experimental license in 700MHz and 2.5GHz to deliver line of sight services to outdoor antennas; in Atlanta it has an experimental license in 3.5GHz and is testing non-line of sight configurations with roof-mounted and window mounted subscriber equipment. Here it has installed four base stations, all linked by wireless backhaul. Equipment from several vendors is being used and the services include voice over IP, private data networking and internet access. The company says that corporate customers are also interested in the potential of WiMAX to provide a back-up network for the wired system, and for creating temporary and ad hoc networks to support conferences or meetings, for instance.
“Our WiMAX trial in New Jersey is off to a great start,” said Hossein Eslambolchi, chief technology officer and CIO, in June. “Feedback is helping us increase the value of the technology for our enterprise customers.”
The next stage will be to test indoor equipment and then mobility. AT&T is a key part of the growing Intel R&D ecosystem forming around WiMAX, and has a joint development with the chip giant focused on smart antennas to improve rate, range and mobility. Sprint Nextel is also a partner in this program, along with various equipment vendors such as Alvarion, NextNet and Nortel.
AT&T’s strategy looks strong, even with the uncertainties of how it will be affected by the SBC takeover. The nature of its business means it does not need to be distracted by issues of mass consumer devices and roll-out, where the major question marks still hang over WiMAX and where mobility ‘ a year away ‘ is critical. Although there will be strong applications for mobile broadband in the enterprise, this sector has short term needs that are more urgent and that AT&T is well positioned to address. Just as the start-up WiMAX operators have found that focusing on the business customer’s backhaul, VoIP and access needs is more profitable in the short term than trying to address the consumer space too early, so AT&T is making the same assumption at the high end of the market. By reducing its own costs to provide enterprise services, and passing on some of those savings to the customer along with new performance levels and configuration flexibility, it stands to exploit its advanced fiber network more effectively and bring a rejuvenated corporate services business to its new parent.