Carrier WiFi

WLAN: The Four S's

802.11 wireless LAN is being touted as the "next big thing" in wireless, allowing companies to reach employees, remote locations, vendors, or even just conference rooms at speeds that put a typical T1 connection to shame. Furthermore, WLAN technology enables an enterprise to do all this without spending hard dollars on wiring, cabling, installation, and construction. But there are also hidden costs to consider, and wireless LANs do have two big "gotchas": security and quality of service (QOS). From the security perspective, there are folks (literally) driving around searching for unsecured wireless networks. The practice is known as "War Driving," or "stumbling." War Driving – which takes its name from the 80s hacker classic "WarGames," which invented the now-popular term "war dialing" for the technique of mass-prefix dialing – is a way for hackers to gain easy access to your corporate system. Some security firms do this as well, as a way to inform companies of security risks on their networks. Clearly, the goal of any firm implementing a wireless LAN is to make it just as secure as any wired LAN. This sounds simple, but thanks to the inherent weaknesses of the Wired Equivalent Privacy (WEP) protocol, which encrypts data as it travels over the airwaves, the security issue becomes dizzyingly complex. For state-of-the-art WLAN security, an administrator must consider such issues as Temporal Key Integrity Protocol (TKIP), Message Integrity Checking (TKIP-MIC), Mutual Authentication, Policy-based Key Rotation using RADIUS technology, and more. A solidly protected WLAN will be integrated into an IPSec-based VPN, and will have continually updated security policies. Now it’s not sounding so easy, is it? In reality, these are all issues a business should address when implementing any kind of technology. In this case, however, the rub is the implementation and management. Administrators must clearly define who needs access, whether it’s concurrent, individual working habits, QOS issues, etc. Site surveys are critical to identify radio coverage, signal strength, rate of packet loss, etc. Then, of course, you need to actually design the network and define the security and QOS policies. But let’s not forget our friend the OSS (operational support system). While planning and surveying for a wireless LAN, a truly prepared organization is doing the exact same thing to your OSS. Knowing the impact to your systems will be the key to overall success. How will you know if you’ve been hacked? Your OSS. How will you know if your users are receiving the QOS levels set? Your OSS. How will you know if you need more bandwidth to a certain site? Your OSS. And so on. If time were abundant, capital dollars were flowing freely, and you had ample resources to throw at it, you could examine all the functions of your systems to better grasp the impact of a wireless LAN on the whole. The reality of today’s business environment, though, dictates a more frugal approach. So, prioritize. Focus on the Four S's: Security, Service quality, Support, Scalability. Security: Tie your fault-monitoring system into your wireless routers. Establish mechanisms to track IDS in a consolidated manner through your OSS. Take security management as a mandate, not a nice-to-have. Enforce it, enforce it, and enforce it. Service quality: QOS drives policies, which drives capacity planning, which drives traffic engineering, which drives strong inventory management, which drives strong provisioning processes. Link them together and establish systems and procedures, and you can finally climb this mountain. Support: Some users are dubious at first, but then come to expect the speed and ease of a wireless LAN. When a problem arises, however, frustration will rapidly shoot through the roof. Tightly integrated systems will reduce anxiety, increase responsiveness, and allow for more successful remote implementations. Scalability: Plan the inventory, bandwidth, policies, and applications to scale. Know that if it’s easier for users to get to their applications, they’re going to want them more often and faster. — Nicole Gillen is a Client Partner at Greenwich Technology Partners. She may be reached at: [email protected].

jacksullivan66 12/4/2012 | 11:58:39 PM
re: WLAN: The Four S's Saw an article in a competing online rag (http://www.commsdesign.com/new... about how differing camps within the 802.11e TG are unwilling to give in, and as a result, have created a "stalemate" - and that two competing QoS technical solution will emerge as a result!

Hey unstrung gang - what's the word as you hear it? Dying to know. If this is true, and holds, the .11 mess will get even messier.

Anybody up for a CE / home networking / WPAN solution based on one, common UWB standard instead? I'll bet you the CE and home networking crowd are wondering about exactly this question?

IPobserver 12/4/2012 | 11:58:33 PM
re: WLAN: The Four S's Not sure about .11e, but the UWBers are also in conflict.

Some folks involved in that process are now saying that two, or even three, UWB standards good emerge... much like a,b or g.
jacksullivan66 12/4/2012 | 11:58:31 PM
re: WLAN: The Four S's Now come on, Gabriel. Don't leave me hanging like that.... What have you heard, specifically? Let's have it... ;)

I track that 15.3 Alt PHY group fairly closely, and know there are several camps pushing seperate technical visions. But everything I've heard (including public qoutes from both sides) suggests that one standard will emerge - despite the divisions. It appears all parties are well aware of what's at stake.

We can only hope they don't make the same mistakes that are beginning to plague the .11 camp (.11g is actually slower than .11b in heterogenous environments?, and .11e has turned the idea of one common QoS standard into a maybe)... These two decisions alone could begin to derail the momentum of 802.11 - especially for VoWLAN, multi-media, and enterprise applications... If I were an OEM, I'd definitely be getting frustrated...

IPobserver 12/4/2012 | 11:58:30 PM
re: WLAN: The Four S's Well to be clear, I'm not saying its going to split, but I have on-the-record quotes from a leading developer that this is what they're now expecting.

I need to write this up properly, but it shouldn't be too tricky to work out who's saying this... just think who's claiming working product and therefore can't afford to wait around for the standard.

On a general point, all these standards folk are so good at politicking, they should be running for president.

Sign In