- 18 Hot IT Certifications for 2014
- CIOs Opting for IT Contractors Over Hiring Full-Time Staff
- 12 Best Free iOS 7 Holiday Shopping Apps
- For CMOs Big Data Can Lead to Big Profits
Network World - You're undoubtedly already thinking about that next wireless LAN upgrade, partially because BYOD and accelerating demand for mobility is straining your current installation, but also because the advent of gigabit-class wireless LANs is motivating an examination of what's working -- and what's not -- in your shop.
The only certainty in networking of any form, wired or wireless, is growth -- growth in demand, coverage and capacity requirements, applications, and essentially every IT element contributing to the success of the organization.
|The Wi-Fi-enabled world|
|Getting ready for gigabit Wi-Fi|
|Hotspot 2.0 promises to redefine the Wi-Fi experience|
|11ac will be faster, but how much faster really?|
|Don't sweat 802.11ac Wi-Fi -- because 802.11ad will knock your socks off|
So even as we look forward to the newest 802.11ac systems to help address our ever-expanding needs, now is an opportune time for tuning the existing enterprise WLAN. Getting this optimization right can make a huge difference between success -- and having to work nights and weekends.
As it turns out, none of what follows is difficult -- remember, we're looking for optimal, not perfect. And the declining costs and dramatically-improving price/performance of the required systems, coupled with a reduction in operating expense thanks to greater operations-staff productivity and an overall improvement in end-user productivity (the real bottom line in any network), can easily pay for any effort involved.
Given the degree of functionality defined in the 802.11 standard -- roughly 3,000 pages at present -- and the remarkable capabilities of chipsets based on the standard, it's easy to assume that most of the work required in optimizing a WLAN has already been done by the IEEE and chipset vendors. It's important, however, to consider two additional complicating factors:
First, radio communications of any form is a function of the laws of physics. Radio itself is a statistical medium -- not fundamentally unreliable, but indeed subject to significant swings in performance simply as a consequence of such elements as available spectrum (channel bandwidth), the type and orientation of antennas used on both ends of a given link, interference (unintentional and otherwise), client motion and a wide variety of other variables. It's easy to see how predicting WLAN performance can be somewhere between complex and impossible.
Second, specific WLAN system implementations can have an equally dramatic impact on performance. Most enterprise-class vendors have sophisticated and often-proprietary techniques for allocating traffic to channels, prioritizing traffic and otherwise optimizing their products. It's thus advisable to learn the details of these capabilities and apply any appropriate, even if specialized, optimizations.
While we tend to think of performance in terms of throughput alone, the shared nature of the WLAN medium places a much greater emphasis on capacity -- being able to serve a large, growing and diverse (in terms of both devices and applications) user base while maintaining the overall goal of boosting end-user productivity.
The basic philosophy of WLAN deployment has, however, historically been based on coverage. This is an artifact of the time when access points were expensive, applications were primarily vertical (think logistics, transportation, warehousing and data collection) and throughput requirements were low.
Today, however, the WLAN is the primary, default or even only access for many users. Couple that with increasing demand, time-bounded traffic (voice and streaming video), the need for pervasive coverage and the environmental and vendor-proprietary variables noted above, and performance optimization is often viewed as complex or even theoretical at best.
And it gets worse. Traditional performance evaluation usually takes the form of benchmarking, involving the analysis of throughput given a quantifiable synthetic workload. While this can and does work well on wire, the inherent variability of wireless can yield unreliable conclusions unless additional steps are taken during the benchmarking process -- such elements as multiple test runs and spectral analysis are vital, given the statistically varying medium that is radio.