The frequencies that Wi-Fi uses are unlicensed and end up being shared by a number of non-Wi-Fi products (such as microwave ovens and portable phones). Even the 5.8-GHz region interrupts the spectrum allocated to 802.11a and, while it's usually more quiet than the 2.4-GHz region, is subject to broadband radio frequency noise that compromises signal quality.

---

How we did it
Archive of Network World tests
Subscribe to the Network Product Test Results newsletter

---

We recently tested two new products: the AirMagnet Spectrum Analyzer (AMSA), which can help network engineers scan the airwaves to find interfering noise sources; and Network Chemistry's BlueScanner, which can detect and articulate the rapidly growing number of Bluetooth-enabled devices on a network. Because these products have different goals, we gave them separate scores.

Spectrum analysis

The AMSA is a CardBus notebook adapter that includes its own internal antenna and has an additional external antenna attachment. The internal antenna wasn't nearly as good as either the supplied external antenna or a third-party directional antenna.

The card acts as an analog-to-digital converter, which the Spectrum Analyzer software then checks. The card and software track the 2.4-GHz to 2.5-GHz band (802.11b/g) and the two portions of the 5-GHz band (802.11a) that are reserved for unlicensed, low-power Wi-Fi use. Installing the card was simple, taking less than 5 minutes.

With notebook in tow, you can walk around to track down noise sources, perform Wi-Fi air quality before an installation or reposition Wi-Fi gear. Because a wealth of information can be displayed on the notebook, the display should have a wider geometry than the typical 1,024- by 768-pixel resolution found on a notebook. We initially tried a notebook using this "low" resolution and found that multiple display readout windows used up screen space quickly.

The software presents a blank area that can be filled with different displays, and at the bottom of the displays is a statistics box where interference sources are listed. This is the key area to watch, as the listing of sources that pop up in the box are a near real-time indication of the noise source (and, likely, its type).

Once the AMSA finds an interfering listing, you can hunt it down methodically using a display that finds the device through antenna-positioning manipulation. This is where a directional antenna becomes useful to find sources quickly. Some devices can be found readily (such as a 5.8-GHz FM wireless phone), while others are sometimes more difficult (such as a leaky microwave oven). Ease or difficulty of finding sources depends on the transmission features: For example, FM phones send periodic pulses, even when not in use; a microwave oven might be on for 3 minutes, then off for hours. Hunting down a source requires a rapid response, and the operational radius and sensitivity of the AMSA make this a monitoring rather than an alarm tool.

In testing, we wished that a loud noise occurred when an offending device was found in the AMSA's operational perimeter. And remote sensors with the AMSA's analog-to-digital conversion might allow a premises to be monitored over a larger area. This is why sensitive and highly directional antennas are a must-have for using AMSA in reactive situations.

We could add differing display plots as discrete windows and tile them within the AMSA display. We could save each setup to be recalled later, but were frustrated with having to exit and reload the program to do this, rather than simply pick the profile of plots we wanted from a menu of the ones we had built. Building the items to put into the plot display is very easy, although if we chose something that wasn't valid (for example, asking AMSA to display the same spectral sweep through two checkbox choices), we didn't know this was an error until we had saved the profile, exited AMSA, reloaded, and found an error message.

The upside is that AMSA tried to load a default plot, similar to the one we'd erroneously designed. Chosen in this way, the default plots were only occasionally correct, based on the goals of the error plots we designed. We'd rather see a method that parses the choices before saving the plot, saving numerous steps and head scratching.

We were thrilled to be able to track all IEEE 802.11b/g/a frequencies, not just the ones that are "legal" in the United States. Because this is a passive device, choosing all frequencies is good; some organizations are unaware that "illegal" channels are in use, because other products don't monitor them.

Whitepapers

• Secure Wireless Printing Options
• Network-Wide Change Management Visibility with Route Analytics
• Getting in Compliance With Government Data Regulations By Leveraging Online Security Technology
• How to Offer the Strongest SSL Encryption
• Maximizing Site Visitor Trust Using Extended Validation SSL
• The Latest Advancements in SSL Technology

View more NETWORK MANAGEMENT whitepapers

Webcasts

• Managing the End User Experience-A Real World Example of Success
• Branch Office Trends and Best Practices
• Correlating granular network and application visibility
• Migrating to an MPLS-Based Network
• Round Table Webcast with Forrester Research
• Best Practices for Deploying WAN Optimization for Disaster Recovery

View more NETWORK MANAGEMENT webcasts

Special Reports

• WAN Optimization: The Ultimate No Brainer
• Ehternet Services: WAN options mature
• Keeping Spam at Bay
• Network World Executive Guide: The Virtualization Equation
• Application Performance Management Executive Guide
• Executive Guide: Building a Service Oriented Architecture

View more NETWORK MANAGEMENT special reports