Forum

Here is a Yahoo discussion group:

http://groups.yahoo.com/group/Radio_Mobile_Deluxe

I've used Ekahau ESS and Motorola (Wireless Valley) LANPlanner. There are limited versions of LANPlanner being offered by some other companies for a lower price.

You can expect differences in accuracy for software tools, and there is also a big difference in price.

I don't know the currently quoted price for LANPlanner. The last price I had for Ekahau ESS was about $3700, while LanPlanner was selling for several times that amount.

LANPlanner constructs a 3D building, so you get to plan and analyze signals through floors, while Ekahau ESS deals with a single floor at a time. If there is high enough floor attenuation, a floor at a time can still work. Of course, there is leakage at stairwells and elevator shafts that has to be accounted for.

Now that Motorola has bought Wireless Valley, they seem to be moving LANPlanner in a different direction. It still offers primary ray tracing, but not multiple ray tracing. There is research that concludes multiple ray tracing is more accurate, but requires much more CPU time to calculate.
The goal with the newer LANPlanner may be to improve the accuracy of predictions by using site measurements to adjust the attenuation values of each type of wall. Being able to display the errors between measured and predicted RSSI is very helpful, and there are tools in version 9.2.5 for that purpose.

If you use only predictions without any measurement data, the attenuation and reflection values assigned to each type of wall are more important, and in those circumstances, using multiple ray tracing may be more accurate, if you have a fast enough computer.

There are at least three other companies offering indoor WLAN propagation prediction software.

1. AirMagnet sells Planner, which as a standalone application is about $2,000. According to their sales department it is intended for one floor at a time, so it wouldn't automatically calculate interference from an AP on a floor above or below. The datasheet on the AirMagnet site does not mention the method they use for propagation prediction, and they say it is proprietary.

2. Actix sells Radioplan RPS, which has a version for around $10,000 that may be similar to the capabilities offered by LANPlanner for prediction. They have a student edition available for download.

3. EDX sells SignalPro, which probably costs about the same as Radioplan RPS or LANPlanner. There is a Microcell/Indoor module. You can also predict delay spread and view multipath, according to the datasheet, but I have not done any experiments or seen any information on whether this would help with future MIMO WLAN designs.

Based on their advertising, Radioplan and EDX primarily concentrate on outside predictions, but have indoor capabilities.

Don't expect to buy any propagation prediction software, take a few hours to learn it, use it in a large building based only on a floorplan with no measurements, and have a really accurate prediction. Wireless Valley used to say regarding LANPlanner predictions that with an accurate floorplan, they could get around 85% accuracy. If this means actually moving 15 of 100 access points and their Ethernet drops in a building, after installation, I would expect an unhappy client.

I have yet to get an accurate floorplan for any building, and there have often been walls added and removed without updating the only floorplan available for WLAN planning. Also, I have seen floors with as little as 12-15 dB attenuation, and upwards of 30 dB. This much difference would have a big impact on co-channel interference, but a floorplan can't provide this information.

Wall attenuation can vary from whatever is furnished as typical in the software package. While a difference of 1-2 dB isn't important for a single wall, predicted loss for each access point often includes several walls. If each AP's predicted coverage is off by one wall, that would not produce a good design.

Charles Preston
http://www.comsec-wireless.net

Regarding my last post, on prediction tools, I wrote:

"Wireless Valley used to say regarding LANPlanner predictions that with an accurate floorplan, they could get around 85% accuracy."

In case that's not clear, I meant that using floorplans alone, with no measurements, could produce about 85% accuracy. If you use extensive measurements and optimize the prediction, accuracy should be quite a bit better.

Charles Preston
http://www.infosec-wireless.net

Good stuff Charles.

I'll add to that, Motorola has the ability to import CAD drawings as well as doing floor plan traces, and I believe their accuracy predictions are based upon using CAD. Their sales pitch to me was that their software was so accurate that manual validation was not needed. However, like Charles, I've yet to really find that to be true. I always recommend manual validation if you plan to use the wireless network for mission critical apps.

With that in mind, both Ekahau and Motorola have the ability to also do manual site surveys, however I think both - particularly Motorola's manual surveying components are way behind their planning software. That is why I use the Helium Networks solution for manual validation.

Good stuff Charles.

I'll add to that, Motorola has the ability to import CAD drawings as well as doing floor plan traces, and I believe their accuracy predictions are based upon using CAD. Their sales pitch to me was that their software was so accurate that manual validation was not needed. However, like Charles, I've yet to really find that to be true. I always recommend manual validation if you plan to use the wireless network for mission critical apps.

With that in mind, both Ekahau and Motorola have the ability to also do manual site surveys, however I think both - particularly Motorola's manual surveying components are way behind their planning software. That is why I use the Helium Networks solution for manual validation.