Positive experience replacing one router with another, with heatmaps

[fityorreeps]

We have a mixed environment at a very small school. Walls are cinderblock, outlets are scarce (and old), and ceilings are funky, which has discouraged us from installing a WAP in each classroom. We use a single wireless router to serve a handful of wireless clients. It's always been marginal in the farther classrooms, but it wasn't used that much, so we endured. The 2.4 GHz band is pretty clear at channel 11, and the 5 GHz band is clear everywhere.

Starting this year, we will have more wireless clients. After discovering that our router would intermittently fail (it would stop broadcasting on one band entirely until it was reset) we decided that at the very least we would start by replacing the router. We didn't expect range to increase much, but the old router had to be replaced regardless.

We replaced the Linksys WRT600N with an ASUS RT-N66U. Here are before-and-afters in both bands make with Ekahau HeatMapper. The descriptions are the filenames, so I hope those are visible. Don't laugh at my data points; I was learning how to use the program. The improvement is dramatic, and is borne out not only by my amateur benchmark but in real use as well. It remains to be seen if we will have to install some WAPs, but we're very happy with the improvement already.

 

[stevech]

note that you're looking at the to-client signal strength.
The weakest link is typically the from-client strenth at the AP/router.g

 

[fityorreeps]

note that you're looking at the to-client signal strength.
The weakest link is typically the from-client strength at the AP/router.g

That's true, and it's why I made sure to mention that the improvement "is borne out not only by my amateur benchmark but in real use as well".

Out of curiosity, is the from-client strength reflected in the number of "bars" of signal in the client OS? If not, is it possible that with an especially powerful router a client might have "five bars" of signal and yet a completely unusable connection? I've never experienced that phenomenon, which is why I'm curious if the OS somehow measures the from-client strength.

 

[stevech]

When I speak of client-strength, I mean the received signal strength at the WiFi router/AP. Some product display a list of associated clients and the signal strength or data rate for each. Data rate implies signal strength. The to-router signal often is weaker than the reverse, and per most 802.11 modes, the weakest link (to/from) is the overall speed constraint.

There's no standard for converting signal strength in dBm (engineering units) to % or to bars. The dBm scale is log; % isn't! and bars is a marketing choice. Apple learned this lesson when AT&T got caught giving Apple fake formulas for bars vs. dBm to make the iPhones look better than they were, for cellular.

 

[thiggins]

The proof is in the use. And if you got improved coverage that's what counts.

Out of curiosity, is the from-client strength reflected in the number of "bars" of signal in the client OS? If not, is it possible that with an especially powerful router a client might have "five bars" of signal and yet a completely unusable connection? I've never experienced that phenomenon, which is why I'm curious if the OS somehow measures the from-client strength.

Yes, those bars are related (somewhat indirectly as Stevech points out) to the received signal strength at the client.

It is possible to see "bars" and not have a usable connection with a connection showing 1-2 bars. But not with 4-5 bars, in my experience.