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

[bcavros]

Great article great follow-up to your 2009 "Best Wifi you have never seen". Its sad to see that while consumer wired connection have increased by 3 orders of magnitude in bandwidth wireless technology has barely effectively increased 1 order of magnitude. The whole of wireless tech is so bogged down in meaningless marketing speech, I am glad you took the time to put together an article to quantify these points.

I would love to see part three based on home or small business environment. Yes, bring the Ruckus along for a mini-gun in a knife fight. Through in some wireless phones, garage door openers and poorly RF shielded case mods... Drop the APs price point down from US$700-100 to $150-250. Are there any consumer level unit with any antennae technology worthy of note?

 

[daglesj]

I've usually solved most wi-fi issues by slapping on a couple of 8-9dBi antennas. Total cost - $30.

 

[Cowgoesmoo]

[citation][nom]neverdyne[/nom]This article got me thinking if it would be overkill to use a Ruckus AP on a home. My wireless sucks really bad, my house is built completely of bricks and cement with very thick walls. My current router (Linksys WRT160N)cannot reach one of the upper floor rooms, and it's crucial that I get internet there. The cost of breaking down the wall and installing a wired connection would surpass the cost of a $500 AP.[/citation]

I have a Ruckus 2942 AP in my place (All Ruckus AP work stand-alone with no controller) all double brick walls etc. It is a damaged AP that I picked up from work. The POE chip on the board has a hole in it from a lightning strike but it works great on the power pack! Works heaps better than the ASUS box I used to have. In places where I had no signal windows now shows 3 bars.

 

[Cowgoesmoo]

[citation][nom]MountainFlip[/nom]D-link is my brand. Why weren't they mentioned at all?[/citation]

From my experience selling wireless LAN gear from several vendors when considering buy back deals none of the big players (Cisco/Ruckus/Enterasys/Xirius) consider DLink to actually be a competitor in the enterprise space and wont offer a trade it deal. As far as they are concerned it is home grade gear with a controller. From my experience fixing issues they aren't far wrong.

I can do a DLink site survey for you for any school anywhere in the world right now. "You will need at least one in every room."

 

[Cowgoesmoo]

[citation][nom]802uwaveguy[/nom]I find it interesting that this testing did not include a Xirrus AP. They are clearly a contender of not a leader in the 802.11n heavy AP market and when discussing 3x3:3 multiple radio products should always be considered. The Xirrus unit is costly but when conducting a performance test and bake-off, not considering Xirrus is a mistake. It seems that all reviews I read concerning .11n multiple radio AP test data interestingly don't include a Xirrus unit.[/citation]

I would also be really interested in a comparison. From what I can understand they are just a bunch of AP in a box each hooked up to separate sector antenna but it would still be interesting to see the results!

 

[daglesj]

This article brought to you by the Ruckus corporation.

 

[Guest]

Was this review paid for by Ruckus as Meraki claims?

 

[Guest]

I would have loved to see how the d-link, actiontech, Motorola, and 2wire gateways provided by DSL and cable companies fair in this sort of test. I work for a local ISP in customer tech support so i get a lot of calls about wireless issues... and at times its hard to give answers.

 

[cangelini]

[citation][nom]dbalad123[/nom]Was this review paid for by Ruckus as Meraki claims?[/citation]

We weren't paid a penny for this story by ANY company.

 

[williamvw]

[citation][nom]dbalad123[/nom]Was this review paid for by Ruckus as Meraki claims?[/citation]
Nobody likes to get shown up in a review of this sort. But I was there. I saw the data come in with my own two eyes. When Ruckus or Cisco fell behind, I said so. When *any* product in this roundup had issues, I said so. If somebody else wants to grab another collection of hardware like this and run these sorts of tests, go for it! And if Meraki or anybody else could deliver far better results through specific tweaking of their options, more power too them. I didn't show the data I did as a vendetta against Meraki. I've never so much as emailed with anyone from that company and have no opinion about the outfit or their products whatsoever. My purpose was to illustrate the concepts of contention, airtime fairness, and the other factors discussed throughout the article and show how they play out in real life. When it came to airtime fairness, the MR24's data amply illustrated what you *don't* want to see in your network. Could a different Meraki AP perform better? Or even a firmware update to this one? Very possibly! With infinite time and resources, we could find out. But this was what we could accomplish in the time we had.

 

[geekrebel]

@williamvw

What was Ruckus' involvement in designing & setting up the testing environment? Was the testing really done at their offices?

 

[geekrebel]

Also, what make was the "AP mounted to the ceiling"?

 

[cangelini]

[citation][nom]geekrebel[/nom]@williamvw What was Ruckus' involvement in designing & setting up the testing environment? Was the testing really done at their offices?[/citation]

Geek,

I can certainly appreciate your interest in transparency; please allow me to address your concerns as thoroughly as possible.

This piece was conceptualized many, many months ago during a conversation with the folks at Ruckus, who’ve been exceptionally helpful in our efforts to ramp up network testing at Tom’s Hardware. Space was rented, testing was conducted, and it was determined that our results were not consistent enough to provide any useful insight. The whole project was scrapped.

Months passed, and we ended up on another call to talk methodology. That was when the idea for this piece came into focus. Now, perhaps you know something that I don’t (at least, that’s what you seem to be suggesting), but the plan was that empty office space would be rented for the purpose of this piece alone, and then set up with all of the hardware you saw evaluated in the piece. [Quick edit: I wasn't on-site for the testing; our writer William was]

Tom’s Hardware/Bestofmedia received zero money for this story; to the contrary, it *cost* us thousands of dollars to put together.

To answer your question about the ceiling-mounted AP, it was an Aruba 125.

If you have any other questions, please feel free to ask. While we did have help putting this piece together (its data would have been impossible to procure otherwise), all of our methodology is well-documented, and if you’d like to see the raw data, we can make that available to you.

Best,
Chris

 

[Silex99]

Not a single mention of how different wireless encryption methods do to performance. I have been using WEP for years and recently switched to WPA. My wireless sucks now. I was hoping for some insight on this from such a detailed 2 part article.

 

[destinova]

Since we're on the topic of full disclosure:

What did Ruckus charge for the use of their equipment?

Has TH received or will TH receive any free or drastically discounted Ruckus gear for future use?

Why was the 7343 used instead of the 7962?

 

[orange222]

It's a nice review but from the picture which is called "Download TCP performance 10' at LoS, One 5GHz client, result in Mb/s" you missed to show the HP 460 AP.

 

[geekrebel]

@cangelini, there was a comment earlier that stated that Ruckus' office space was used, I merely wanted to know if that is true.

I don't think (nor suggest) that Ruckus paid you, but if they were involved in the actual setup of the testing, it does pose some hard questions.

Were Ruckus on site during the tests? Is there any chance that they could have interfered?

If they were on-site, or even if they knew when & where the tests were taking place, they _could_ have done signal jamming (I'm not saying they did or suggesting they did, simply that it's a probability that would make a "blind" study impossible)

(My disclaimer: I've used Meraki in various deployments with great results. Although biased, I may well change loyalty if I ever see Meraki performing as poorly as in your tests)

 

[fayskittles]

I would have liked to see Netgear in the round up. I love my Netgear WNDR3700, It even has a cool feature called "Wireless Isolation." When enabled you can't see any other device on the WiFi with you. I would like to see some more real world test like you have done on more routers.

 

[kgilchrist]

Good article!

I live in a 25 story building and there are 20+ networks that my Ipad can see.
Who knows how non-SSID announced networks, baby cams and other polluters there are.

My next wireless purchase will draw on this. I rely on powerline as much as possible and only do web browsing on web devices and don't even bother trying to stream video at this point.

I often hear from neighbors how their Cable TV co provided router doesn't work well for them.
Easy to see why.

 

[Guest]

Do you have an Intel chipset for your Wi-Fi? If so, all bets are off for Ruckus. Tech support told me "there is a driver issue with Intel". I wonder if they specifically chose the broadcom chipset for those Dell laptops.

 

[Guest]

with an ever increasing need for bandwidth within the 2.4ghz range, surely a smaller cell size which will allow for a higher density of AP deployments is a good thing?

 

[Dude4Linux]

It's been two years since you posted this excellent article. A lot has changed in those years. We now have the 802.11-ac products appearing in the marketplace. I'd like to see a followup showing results for the latest products from the top vendors plus some new contenders such as Ubiquiti. Additional tests that I'd like to see are heat charts for other devices i.e. iPhone, Android phone, and tablets iPad, Galaxy Nexus, etc. Pick a representative device from each category and compare the heat charts for each access point. Also I'd like to see tests which explore the effects of polarization. Pick a client device with a known antenna orientation and then generate heat charts with both horizontal and vertical orientation. Lastly I'd like to see a comparison between ceiling mounts and wall mounts for each access point to show how they handle less than optimal positioning. Hint: use a robot when generating the heat charts so the path for each test can be repeated uniformly.

 

[Dude4Linux]

It's been over two years since you posted this excellent article. A lot has changed in those two years including the introduction of 802.11ac products into the market place and the emergence of new vendors such as Ubiquiti. I'd like to see a repeat of the testing using the latest b/g/n hardware compared to the -ac access points. I'd also like to see additional testing using clients other than the notebook, i.e. iPhones, Android phones, iPads and Android tablets, etc. Pick a representative from each class and run the heat maps test to show if any problem areas appear. Secondly, I like to see some comparison of polarization effects. Pick a device with a known antenna orientation and then repeat the heat map tests with both horizontal and vertical orientation of the device. Lastly, I would like to see a comparison between ceiling and wall mounting each access point to see which access points can handle sub-optimal positioning. It would be great to have some actual test data from a trusted, un-biased source.