Interesting article relating to high wireless N speeds

[00Roush]

I ran across this article yesterday and figured some of you guys might want to give it a read. Good to hear that it is possible to actually get good wireless speeds. http://www.tomshardware.com/reviews/beamforming-wifi-ruckus,2390.html

00Roush

 

[thetoad30]

I saw the same thing and actually was thinking about posting a request for review to see if these things really do work as stated... Then I saw the price. $1000 a unit. I mean, if wireless wasn't changing so fast (they haven't even gotten two-stream 300mpbs going yet and are already wanting to transition to multiple streams) I would actually pay for something like this to make sure my wireless worked as long as the reviews were good.

 

[00Roush]

Wow... that price is a bit steep! The one they used in the article seemed to be an enterprise level device. Have you checked to see if the consumer setups are much cheaper?

00Roush

 

[thiggins]

Thanks for the article link. I'm not familiar with William Van Winkle. But it looks like he tried to do a pretty thorough job in testing. A few points:

He used a vendor-supplied tool for most of his data-taking. I would never do that, since you don't know what sort of optimizations could be baked into it.
He included IxChariot TCP/IP based tests, but that's not how he did most of his testing. Of course, IxChariot can test using UDP or RDP or other protocols and also has test files that emulate video streaming.

Second point is that UDP isn't always used for video streaming. It depends on how the player connects to the source file. TCP/IP is actually more resistant to packet loss than UDP, as I showed in Video Streaming Need To Know: Part 1- Encoding, Bit Rates and Errors.

From the streaming tests that I have done, I've found that you're actually more likely to be using TCP/IP when playing a file from a NAS using a networked media player.

The AP he tested is an enterprise AP, and as he pointed out, not designed for media streaming per se. You would be more likely to use one (or a pair) of Ruckus' MediaFlex products (MediaFlex 7811 multimedia access point (AP) and the Ruckus MediaFlex 7111 multimedia Wi-Fi adapter).

From the way I see the industry going, it's going to take a dedicated product to handle good HD streaming over any distance. Consumer wireless manufacturers are optimizing products for low cost and high speed, often at the cost of range and throughput stability. And 5 GHz range remains poor compared to 2.4 GHz.

I actually wrote the Tom's Guide MIMO Router Face-Off (now attributed to "TG Publishing Team"). Ruckus was not at all happy about having their MF2900 and MF2501 compared to other MIMO products (and that they didn't do that well), since they said the pair was designed for video streaming, not for general purpose wireless networking. I haven't heard much of anything from them since then

The article is also correct that Ruckus didn't like life in the retail networking space. I think the MediaFlex 7111/7811 pair is primarily used by service providers now.

I don't think they are interested in retail business, but I'll ping Ruckus to see if they would be interested in providing the 5GHz MediaFlex pair for a video streaming test.

 

[00Roush]

Good point about using the vendor supplied tool. I should probably change my first post to "Good to hear that it might be possible to actually get good wireless speeds". It would be great if they would send you some gear to test.

00Roush

 

[thiggins]

I sent Ruckus an email yesterday and had a nice chat with their head of marketing. They are considering my request, but are concerned about my specific test methods. Right now, they're still thinking about my request.

Their Mediaflex products are specifically aimed at the IPTV market and are tuned for UDP performance. Read the last page of the review for some info on what I'm talking about. The author is a bit fuzzy on the setup, which sounded like he made some basic setup goofs. You never want APs, or APs and clients within inches of each other. I use a minimum separation of 10 feet when testing product.

 

[thetoad30]

I'm actually more interested in their beam forming technology... I'm not sure how to test that except to use their enterprise ($1000) AP.

I'd think that, regardless of how optimized the unit is for wireless or media streaming, that regular wireless transfers would still be leaps and bounds better with beam forming than any other AP that is "optimized" for WLAN access.

I could be wrong, but I guess I'm saying I'm more interested in the foundational technology than I am for any specific function of the AP. Does that make sense?

 

[thiggins]

I could be wrong, but I guess I'm saying I'm more interested in the foundational technology than I am for any specific function of the AP. Does that make sense?

In my experience, no. The technology stories and buzzwords that are used in product marketing are there to try to impress you and impart a "gee whiz aren't these guys high tech" feeling to consumers.

In the end, it's the performance delivered to the user that counts. Beam-forming, beam steering and focused antenna arrays aren't unique to Ruckus. Beam-forming is an optional part of 802.11n and doesn't require fancy circular antenna arrays. This article provides a pretty clear explanation.

 

[jdabbs]

Ralink's upcoming RT2883 supports beamforming, so it may be an option in some high-end consumer routers.

 

[thetoad30]

In my experience, no. The technology stories and buzzwords that are used in product marketing are there to try to impress you and impart a "gee whiz aren't these guys high tech" feeling to consumers.

In the end, it's the performance delivered to the user that counts. Beam-forming, beam steering and focused antenna arrays aren't unique to Ruckus. Beam-forming is an optional part of 802.11n and doesn't require fancy circular antenna arrays. This article provides a pretty clear explanation.

Ok, maybe I should rephrase... I'm interested in their implementation of beam forming since no other manufacturer appears to be implementing this technology. All that I see is the MIMO applications of 11n which is by default for multiple streams.

So I don't care if it was Ruckus or Cisco or No Name, I'm more interested in the technology and whether or not it is something that consumers need to push manufacturers into doing. Because unless we push, it's one more thing that is omitted so that future sales can be had by then implementing the technology after it's already too late (i.e. multiple streams (>2) in 11n to increase throughput before 11n is even ratified).