One area of networking which has been growing at an ever increasing rate, and which looks poised to explode, is wireless connectivity. The ability to set up high-speed links without pulling cable, or laying fiber, results in great flexibility in deployment and rapid build-out rates.
Almost anyone in an office today is familiar with 802.11b, also known as “WiFi”. Transmitting at 2.4 GHz, it allows laptop users to connect to the local LAN from just about anywhere in the office. No longer are these users tied to their desks. WiFi has become so popular in office environments that most laptops now come with WiFi as an option; Intel have even built it into their Centrino mobile processors.
Interestingly, in developing nations, wireless connectivity is now the preferred way of building out both telephone and Internet services. By being able to cover relatively large geographical areas with only a few relay sites and no cables, capital costs are lowered. This means that companies can recoup their investments earlier, and are able to upgrade their infrastructure sooner.
Indeed, even in the “developed” nations, wireless telephony has resulting in a dramatic reduction in the number of “land lines” (read: copper pairs) being ordered. It is now quite common for someone to use a single cell phone not only for their mobile calls, but also their home and business. This, needless to say, is causing a great deal of concern to the telcos, since it’s much easier to change cell providers (known as “churn”) than land line suppliers.
And, additionally, the technical options have expanded considerably in only the last few years. As an example, for networking connectivity, while WiFi promises 11 Megabits per second (Mbps), it can only really deliver 3 Mbps on a good day, in a controlled environment. 802.11g, a successor to 802.11b/WiFi, on the other hand, promises up to 54Mbps in the lab, and can actually deliver up to 22 Mbps (depending on the encoding schema).
Newer solutions, based on proprietary technologies rather than standards designed by committee, can offer even better rates and flexibility. For example, using a technique known as “Orthogonal Frequency Division Multiplexing” (OFDM), there are products offering a true 50 Mbps over long ranges and Non-Line-Of-Site (NLOS). Similar results are promised from 802.16a, also known as WiMax, but only time will tell.
Similarly in the cellular telephony space, new standards like WCDMA and CDMA2000 allow faster data rates so providers can offer wide area, mobile data services in addition to Plain Old Telephone Service (POTS). But of course there’s the issue of capital expenditure to pay for the equipment which can support these new standards (both infrastructure and customer kit).
Particularly for providers in the developed world facing large amounts of competition, this can be hard to justify when they haven’t yet paid for the last generation upgrade. Ironically, providers in less developed areas have an easier time with such upgrade requirements because they can charge higher rates to recoup their investments sooner, and generally have less investment in existing infrastructure.
Now, to throw a little cold water on this warm and fuzzy view of technological wonderfulness, there are of course some issues.
First of all, security; my personal bugaboo. When you stop using a constrained conduit like a copper wire or a glass fiber, and start using the three-dimensional “Free Space” of an area, you immediately have to consider that you’re transmitting to everyone in that area, not just your intended receivers.
It is common knowledge that a huge number of all WiFi office base stations are leaking their signal far beyond their intended office boundaries, and far out into the street and other nearby buildings. Further, a recent survey showed that over 75% of all such transmission points didn’t have even the most basic security enabled (known as Wired Equivalent Privacy, or WEP).
Not to argue that WEP is going to solve the security issues of WiFi. It is also well known that WEP has a fundamental issue such that with a program like AirSnort a passive observer can decipher the key and gain access to a WEP protected network after less than 24 hours of observation. For those three out of four deployments which don’t implement WEP, compromise times are measured in seconds. Thanks for playing.
The solution is to provide an additional level of security, such as only running secure protocols, like SSL and SSH, over what must be considered an untrusted transport layer (even within an office). Failing that, additionally encrypt everything with a trusted tunneling protocol, such as IPSec3DES. How much is this done? Almost never. implementing
The second major issue with wireless technologies is interference. This is particularly a concern in the unlicensed spectrum bands, where everyone’s allowed to play (read: radiate), but with no recourse if someone else causes problems. It’s a free-for-all, literally.
This isn’t generally an issue within an office, but when transmitting between buildings major problems can occur. And, worst of all, the problems generally take place after the initial deployment — you’re running along, happy as can be, and suddenly your Wide Area Network (WAN) stops working. And there’s nothing you can do about it.
Not to disparage wireless solutions; to the contrary — they can be wonderful if designed, implemented, and monitored correctly. But the risks and potential issues must be understood and managed. Wireless technologies almost requires a degree in physics to fully appreciate. Unfortunately, generally speaking, those implementing wireless solutions don’t truly understand the environment in which they’re working. Invariably, this spells trouble.
Published in the Victoria Business Examiner.