MSM460s v. Aerohive: poor RSSI

[ndoudna]

I spent some quality time at a community college's campus center coffee shop,
with two laptops and a Wi-Spy to check out their wireless.  They're running an
experiment with Aerohive and HP.

The Aerohive worked a lot better, and I could use some help understanding why.

There were 2 AP locations, each of which had an Aerohive AP and an MSM460 placed
inches from each other.  There were two SSIDs, one for Aerohive and one for HP.
The Aerohive APs had 6 external antennas; the HP APs I think were MSM460s (so
3x3:3 APs).

My HP laptop's RSSIs were roughly -50dB for Aerohive versus -70dB for the HPs,
measured 25 feet from one pair and 10 feet from the other pair of APs.

I checked the spectrum with the Wi-Spy, and all the RF activity in 2.4GHz was in
channels 1 and 11, where the Aerohives were.  The HPs were all running in channel 6.

So, can't point to interference as a problem for the HPs, the spectrum was pretty
clear there.  But interference wouldn't explain poor RSSIs, would it?  Just
data rates?

Actually, for one HP AP, my laptop couldn't even connect to it in 5GHz.  I've
had problems before with an MSM460 seeming to have a very poor 5GHz range, where
an MSM430 in its place was able to reach the same client in the same location.

So, some questions:

Could external v. internal antennas on the Aerohive v. HP APs explain the
big difference in RSSIs?

Are there known problems with the MSM460s in 5GHz?

Does lower transmit power affect RSSI?  Does HP's auto transmit power adjustment
take into account everything in the air, or just HP APs that a controller sees?

Could the auto power settings on the HPs be causing the poorer RSSI?

Is there a way to tell from using a tool like inSSIDer if an AP is running
"pure" 802.11n?  (all my clients were 'n' capable).

Would interference affect just data rates, or RSSI values a client sees too?

Any other thoughts?

thanks,
noemi

[JesseR]

Lots of questions...  :)

As for HP internal vs. Areohive external antennas, I doubt there's much to read into that, especially since you were so close, physically, to those APs.  The gain of the antennas might be different between the two models, but, at such short range, would a couple dB even matter?  I doubt it...?

Interference can explain poor SNR values, but if you've checked that already, you might be able to rule that out.  However, realize that tools like InSSIDer from Wi-Spy *ONLY* show potential channel interference from other wireless devices.  In other words, it's not a complete Spectrum Analyzer as it won't show potential interferences from other electronic devices such as wireless phones, microwaves (uggg...), and things of that nature.

I am not personally aware of any issues with 5Ghz radios on the MSM 460s.  I manage many of those APs (422s, 430s, 460s, etc.) and don't recall anyhing specific here.

Lower transmit power can affect (by that, I mean it can EITHER improve or degrade depending on environment -- higher power is not necessarily BETTER) SNR values seen by the client, sure.  If you're 10-25 feet away from an AP, even a radio set with very low transmit power should still provide a decent SNR.

As for auto-power, I wish I could answer.  It's my opinion that HP has done a POOR job explaining and showing exactly how auto-power functions with real-world scenarios and environments so, I don't tend to use it. :-(

As for interference causing poor data rates/performance as well as poor signal strength, it's been awhile since I've gone through these actual tests, but I would think intereference (depending on how bad it is), can definitely cause poor data rates/performance and result in lower signal strengths as well.  Or, you might end up still seeing high signal strengths, regardless of interference, but a worse noise floor, resulting in a crappy SNR, even with high signal strength.  There's sooo many variable with wireless, so every situation can be different.

Also remember, when your laptop is reporting RSSI values, it's depending on your wireless card's driver, etc.  So, there's a chance your wireless card has some oddball incompatibility issues with certain brand APs depending on the chipset/drivers they use, or just report the values different depending on what they connect to, and as a result show RSSI values that are not quite accurrate.  I might be more interested in ACTUAL throughput and performance (iPerf is helpful!) than what your laptop shows for RSSI values.

Just my observations/input... hope it helped somewhat. 

J

[ndoudna]

>>

However, realize that tools like InSSIDer from Wi-Spy *ONLY* show potential channel interference from other wireless devices. In other words, it's not a complete Spectrum Analyzer as it won't show potential interferences from other electronic devices such as wireless phones, microwaves (uggg...), and things of that nature.

>>

Actually, the Wi-Spy device *is* a spectrum analyzer.  inSSIDer is a free tool made by the same company (MetaGeek) and it's just a signal analyzer -- but the Wi-Spy DBx plugs into my laptop via USB, and has its own antenna, and it reports RF activity in 2.4GHz and 5GHz (and ZigBee channels).   We also paid extra for Chanalyzer Pro software to run reports.

 http://www.metageek.net/products/wi-spy/

I used both inSSIDer and the WI-Spy, because inSSIDer has a prettier interface (and doesn't have the same bugs as Chanalyzer Pro with reporting bonded channels in 5GHz).  The Wi-Spy confirmed that the spectrum was pretty clean, and most of the RF activity was in channels 1 and 11, where the HP APs *weren't*.  Though this was near a coffee shop, channel 6 was pretty clean (guess they don't run a microwave).

More responses to your answer later (and thanks for that!!).

thanks

noemi

[JesseR]

Oh, cool.  I am just used to people using InSSIDer only, I should have asked.  I have the USB wi-spy dongle thingy too, but mine is only for 2.4Ghz spectrum and is about 5 years old now.  Out of curiosity, do you think it's worth replacing and getting the newer 2.4/5Ghz model?   

J

[wirelessjohn]

I have seen massive coverage issues with the 460's but i dont think HP want to admit to them. I have replaced the 460 with another AP and the coverage is so much better, I have found 5Ghz range to be usless if your more than 10 meters away from the AP with line of sight. So much for an enterprise AP

Check out this review from Toms Hardware is shows the coverage issue in the 2.4Ghz range.

http://www.tomshardware.com/reviews/wi-fi-performance,2985-5.html

[JesseR]

Wow, that is great read.   Your link starts on page 5, so anyone reading that, make sure to read the previous pages for information about the tests/setup, etc.

Thanks for sharing that...  quite valuable reading.  Overall it seems HP did pretty well in some areas, not so good in other areas. 

[ndoudna]

>>

Oh, cool. I am just used to people using InSSIDer only, I should have asked. I have the USB wi-spy dongle thingy too, but mine is only for 2.4Ghz spectrum and is about 5 years old now. Out of curiosity, do you think it's worth replacing and getting the newer 2.4/5Ghz model?

>>

Well, turns out, there's rarely much going on in 5GHz, so I never use it except to quickly check to see if there's anything there, and there never is.  The waterfall view in 2.4GHz is far more entertaining.  That said, it seems irresponsible to call yourself a site surveyor without 5GHz!

>>

Also remember, when your laptop is reporting RSSI values, it's depending on your wireless card's driver, etc. So, there's a chance your wireless card has some oddball incompatibility issues with certain brand APs depending on the chipset/drivers they use, or just report the values different depending on what they connect to, and as a result show RSSI values that are not quite accurrate.

>>

That'd be a shame considering I was using an HP laptop.   Actually, I had two laptops, one a small Netbook with a cheaper chipset that always shows poorer RSSIs than the HP.  But comparing RSSIs from one laptop to the next is worthless -- it's the RSSI from the laptop to the APs that counts.   I'm aware that RSSI is an somewhat arbitrary value that the chipset designers choose, but presumably RSSI on a given NIC is calculated the same to two different APs.

More importantly, people using iPhones look at "bars" (a cruder representation of RSSI) -- and the Aerohive showed 4 "bars" while the HP MSM460 showed 2.  Every client I tried showed better signal "from" the Aerohive than from the HP.

I've heard that the external antennas shouldn't make that much difference, but...shouldn't they?  Don't know.

>>

 I might be more interested in ACTUAL throughput and performance (iPerf is helpful!) than what your laptop shows for RSSI values.

>>

Sure, iPerf was the first thing I tried (I brought two laptops for a reason!) -- but this installation disabled client-to-client communication in both the Aerohive and HP WLANs.  I tried using SpeakEasy Speedtest, but the results were all over the place and couldn't be tied to one WLAN or another, or one band or another within a WLAN.  There must be a consistent testing point in the Internet for performance testing; SpeedTest wasn't it.

>>

I have seen massive coverage issues with the 460's but i dont think HP want to admit to them. I have replaced the 460 with another AP and the coverage is so much better, I have found 5Ghz range to be usless if your more than 10 meters away from the AP with line of sight. So much for an enterprise AP

>>

Ah-HAHHH!!!  That TOTALLY confirms what I saw at this coffee shop, and some testing I did at home too.  Even my Apple Airport -- and an MSM430 -- could reach another room in 5GHz that the MSM460 couldn't !!    What's UP with that?!  A SERIOUS liability!

Can you tell us more about the coverage issues you've had with the MSM460s?  Just in 5GHz or in 2.4GHz too?  (thanks so much for that validation!)  What AP did you replace them with?

Looks like I'm due for another field trip to that coffee shop, this time with Ekahau in hand too.  (Ran out of time for that last time.)

As another note, walking around a coffee shop with a laptop with an antenna attached that says "Wi-Spy" draws a lot of suspicious looks -- !

thanks,

noemi

[ndoudna]

I love Tom's Hardware!  (well, content, not format, those ads are annoying and the "next page" thing is a pain.)

Here's the top-level link:

Why Your Wi-Fi Sucks And How It Can Be Helped, Part 2

http://www.tomshardware.com/reviews/wi-fi-performance,2985.html

This page shows screenshots from Chanalyzer Pro, that I used for the testing I mentioned:

What Interference Looks Like:

http://www.tomshardware.com/reviews/wi-fi-performance,2985-4.html

thanks,

noemi