Looking at the Nuts and Bolts: The Issue of Carrier-Grade Infrastructure Equipment

The information in this section is perhaps more pertinent to operators of 802.16a equipment
than to those using 802.16 millimeter microwave gear, but in a sense it is relevant to all broadband
operators inasmuch as it touches upon an issue that is central to the positioning of
competitive access providers utilizing new networking technologies vis-à-vis the telco
incumbents.
Telecommunications industry professionals like to distinguish between equipment that is
known as carrier grade or carrier class and that which is enterprise grade. The distinction is
asmuch about the attitude of the network operator as it is about design details of the equipment
itself.
Equipment intended to be carrier grade is designed to meet the requirements of what
are known as common carriers, a term applied to providers of traditional telecommunications
services. Common carriers tend to want heavily standards-based, highly redundant, overengineered
equipment that is based on highly reliable computing platforms and utilizes closed
architectures and operating systems and traditional programming languages. Such equipment
must be tested and approved by industry certification bodies and must be backward compatible
with legacy infrastructure. Because of the stringent performance requirements and limited
size of the marketplace, carrier-grade equipment has been expensive to produce, and although
its manufacturing has not been a monopoly business, it has been confined, for the most part,
to a few large companies offering fairly similar products and, incidentally, not inclined to compete
aggressively on price.
The requirements imposed on the manufacturers reflect the basic business philosophies
of traditional telcos, which stress excellent service within a monopoly marketplace. Flexibility,
modularity, low pricing, and easy programmability—or even ease of operation—have not been stressed because such attributes were not necessary to succeed in a public utility business,
such as telephony throughout most of its history.
Enterprise grade, as telecommunications professionals see the issue, is a relaxed standard
applying to equipment serving large private networks within corporations and government
agencies. Such equipment is built to a price and sacrifices some reliability to meet a given price
point. Because it is sold not to a few quasi-monopolies but to a myriad of different types of
organizations, it is subject to considerable variation in design and has attracted numerous
manufacturers both large and small.
Enterprise-grade equipment has been characterized by falling prices, a rapid succession
of generations and standards amendments, ever-expanding feature sets and capabilities, and
the use of open standards, interfaces, operating systems, and programming languages. Such
equipment is increasingly self-configuring and self-provisioning.
To date, the majority of broadband wireless metropolitan public networks set up have
used enterprise-grade rather than carrier-grade equipment; that is, they have tended to rely on
wireless local area network (WLAN) equipment. Prior to the confirmation of the 802.16 standards,
relatively little equipment that could remotely qualify for the carrier-grade designation
existed, and what little existed was costly in comparison to enterprise-grade equipment, most
of which was based on the older 802.11 standard for wireless Ethernet.
I have already touched upon the limitations of the 802.11 standard in respect to public
services in the discussion of wireless broadband standards in Chapter 1. Inherent in the
standard itself are capacity, range, and quality of service limitations that discourage using
such equipment in public service networks. That such equipment is relatively less robust,
redundant, and reliable than equipment designed for the carrier market also gives one pause
regarding pervasive public access deployments. Finally, the fact that almost all public networks
of any size attempting to operate with private LAN equipment have failed suggests that
the strategy of employing it is fundamentally wrong. Perhaps the latest generation of modified
802.11 equipment might fare somewhat better in the marketplace, but this has yet to be
demonstrated.
This should, it seems, put the matter to rest. But, as it happens, the issue is not so simple.
Clearly the intent of this book is to promote the new 802.16 standards and to warn against
attempts to substitute wireless LAN equipment in an application for which it was never
designed. At the same time, I should state that WLAN equipment can have a place in a network
that is 802.16 based.
Already, 802.11 equipment is finding a prominent place in a certain rather specialized type
of public network familiarly known as a hotspot. Hotspots may be considered to be the Internet
equivalent of a pay phone—a short-range access point situated in a public space that would
enable an individual with an 802.11-capable phone or computing device to access the Internet
after an online credit transaction. Some tens of thousands of these hotspots are already scattered
across the globe, and their number is increasing daily. Within the overall category of
broadband public networks, they represent the greatest success story thus far.
The installation of hotspots represents a legitimate strategy for the broadband wireless
metro operator, though it should be considered as merely one service offering. Network operators
may also choose to provide backhaul services to other hotspot operators, and in this
market they may be able to compete effectively with the digital subscriber line (DSL) and T1
services favored by most hotspot operators today. Also, 802.11 airlinks may be useful for in-building extensions of the public network as in
multitenant units (MTUs) and business parks, and their installation may be offered as a valueadded
service by 802.16 network operators.
What I do not advise is an attempt to deploy 802.11 WLAN equipment as the core infrastructure
in a wireless broadband metropolitan network. Such a move is a false economy and
is almost certain to result in a degradation in service that will ultimately cause the network to
fail as a commercial entity. As indicated previously, a few products in the marketplace are
based in part on the 802.11 standard but are provided with ancillary protocols and specialized
interfaces that do represent legitimate attempts to modify the technology for use in the metro;
however, such products almost invariably cost substantially higher than WLAN equipment,
often double or triple the cost. Therefore, a decision to utilize equipment designed from
the ground up for use in the metro rather than modified office equipment appears to make
more sense.