Cirond Corp.s AirPatrol Mobile provides an attractive visual representation of known and rogue wireless networks and devices in and around an organizations airspace. Although the program is easy to use and relatively effective at pinpointing the location of access points and ad hoc networks, administrators looking for in-depth radio-frequency analysis tools or constant detection capabilities should look elsewhere. AirPatrol Mobile is similar to the shareware NetStumbler (available from www.netstumbler.com) in its ability to identify wireless access points, ad hoc networks, radio-signal strengths, security settings and hardware addresses. However, AirPatrol Mobile takes it to the next level of ease of use and visual impact, using multiple data points to triangulate the location of wireless network devices and approximate the location on a floor-plan diagram.Click here to read more about AirMagnets Laptop Trio. AirPatrol Mobile was released last month for $995. Cironds Web site (www.cirond.com) lists only two fully supported and tested client adapters (Netgear Inc.s WAG511 and D-Link Systems Inc.s DWL-AG650), but AirPatrol Mobile worked capably with every adapter eWEEK Labs tried in tests (although different radio components may slightly affect location specificity). We performed the bulk of our testing with Netgears WAB501 client adapter installed on a Dell Inc. Inspiron 600m laptop running Microsoft Corp.s Windows XP Professional. In tests, we imported blueprints showing two floors of our San Francisco office building into the application. (AirPatrol Mobile accepts .jpeg or .bmp files.) Although creating locations and defining floors was simple, administrators must make sure that floor plans are cropped precisely so that multiple floors line up correctly when laid atop one another. Wed like to see Cirond add the ability to allow for slight differences in the images. To calibrate the distance-to-pixel ratio in the plans, we defined the real distance between two locations on the diagram. We then selected the Maximum Accuracy Scanning Parameter, which automated 10 listening cycles at each sensor location. Administrators can customize settings to maximize for time, but this will result in fewer data points and less accurate measurements. We then began walking around the office, performing scans at regularly spaced intervals, for a total of 15 sensor sweeps. Our test network consisted of three "known" corporate access points and three "unknown," or rogue, access points. Each access point was detected during at least eight of the sensor sweeps. We also discovered numerous access points and ad hoc networks outside the building, some from more than a block away. AirPatrol Mobile cleanly displayed the predicted location of each device on the floor plans, with some basic hardware and RF information about each device found, plus a visual representation of which sensors detected the device. Administrators can use this information to tag known devices so that they are displayed in a different color than unknown devices. Wed like Cirond to beef up this capability further, offering a separate display icon for known trusted devices versus known neighbor devices. Our initial location results were less than impressive, particularly for devices located near the office perimeter. Cirond officials said that AirPatrol will display a device to within approximately 15 feet of its true position, but our initial survey results placed some of our access points as much as 50 feet outside the office walls. To compensate for RF vagaries such as multipath propagation, reflection and attenuation, AirPatrol Mobile offers several ways to improve location results. Administrators can characterize each floor with general layout templates, such as "offices only" or "cubicles and offices," to estimate attenuation due to walls and partitions. AirPatrol Mobile does not provide the in-depth attenuation characterizations weve seen with Trapeze Networks RingMaster survey and detection suite, but the AirPatrol attenuation features are easier to use than RingMasters. Click here to read eWEEK Labs review of Trapezes WLAN Mobility System. After scans are complete, AirPatrol Mobile requires that administrators characterize the Building Offseta decibel adjustment that accounts for attenuation factors and physical layout variations in a buildingusing simple tools that estimate average errors and confidence factor. Administrators can use similar tools to estimate the signal broadcast strength of each access point to improve location detection further. We found these attenuation offsets to be somewhat useful for improving location detection, although access points near the building perimeter were subject to greater error because we could not surround the device with sensor data points. However, the process could be quite tedious with many devices. AirPatrol Mobile lets managers fix the position of known devices to display their whereabouts accurately, but these manual corrections are not taken into account in the detection algorithms. Wed like to see Cirond improve this mechanism by extrapolating the data into location predictions for other devices as well. It took a little less than an hour to scan our two floors. Multiplied by the floors in a big building or across a campus, AirPatrl Mobiles manual scans can add up to many hours of work for results that are only snapshots in time. Cirond representatives said they are readying AirPatrol Enterprise, which uses a set of distributed sensors to gather data, for release next month. Technical Analyst Andrew Garcia can be reached at firstname.lastname@example.org. Check out eWEEK.coms Mobile & Wireless Center at http://wireless.eweek.com for the latest news, reviews and analysis.
That said, AirPatrol Mobile cannot perform the kind of in-depth RF analysis that AirMagnet Inc.s Laptop Trio does, and it cannot detect performance or interference issues.