9 Replies Latest reply: Jun 22, 2012 12:27 PM by dashlapak
dashlapak Level 1 Level 1 (0 points)

Just typed up a long and thorough description of the configuration and performance of my AirPort network, which the forum software proceeded to throw away (except for the title) when I navigated away to edit my product list. Lovely.

 

Nutshell...

 

My FiOS Quantum 75/35 service is giving me about 80Mbps straight out of the Ethernet port, as measured via Speedtest.net.

 

By and large, my 2008 model MBP is getting me around a quarter of that when connected wirelessly via the 5Ghz network, more or less regardless of where I go around the house.

 

My iOS devices are giving me about a third to a quarter of what the laptop is providing–six to eight percent of at-the-port throughput, in othe words—with the iPad 2 on the 5Ghz network, and the iPhone 4S on the 2.4 (it can't ever seem to find the five).

 

It's a fairly big, fairly brick-y house, 170 years old. Stringing Ethernet throughout is not an option, as much as I'd love to. In addition to the dual-channel base station, my network has three Airport Express 802.11n's, all set up to "Extend a wireless network."

 

I know that my house is not nirvana for wireless, and that I'm not going to get test bench throughput from my Airport devices in any event. But a factor of 15 degradation with my iOS devices seems kinda severe. I can live with factor of four with the laptop, I guess, but even that seems like kinda lousy performance.

 

Thoughts? Ideas? Pointers? Any and all would be appreciated.


iPhone 4S, iOS 5.0.1
  • Tesserax Level 8 Level 8 (49,260 points)

    A few thoughts for you:

    • To help set expectation of what your 802.11n AirPort Extreme Base Station's (AEBSn) WAN-to-LAN throughput is, I would recommend taking a look at the following Router Charts at SmallNetBuiders website. You will find, even the 1st generation AEBSn, has enough throughput via this interface for your 80 MBps download Internet connection.
    • Similarly, you can use this same site for the wireless data throughput performance for the same base stations.
    • When extending with additional 802.11n AirPorts, remember that this works in a sort of wheel fashion with the extended base station at the hub and each of the extending stations on various spokes of that wheel. Those extending stations within a 25+dB SNR range of the extended station will perform better than those outside of that range. Please check out the following AirPort User Tip for help in the best base station placement for maximum performance.
    • Finally, dense building materials do make a difference in the propagation of Wi-FI RF radio waves. Higher frequencies have a more difficult time passing through this type of material. Also, if you have nearby Wi-Fis or other RF-producing electronics (portable phones, baby monitors, microwave ovens, etc.) these can introduce addional RF noise to decrease the extended networks signal quality.
  • dashlapak Level 1 Level 1 (0 points)

    Ok, so my first Extreme is good, but the second one is below 20dB S/N (for the 5Ghz network), and by the time I get downstairs, it's down to 14dB. The S/N on the 2.4 network is consistently better, but throughput is likewise consistently better on the 5.  Care more about the latter.

     

    I can't really relocated the router. Would adding additional Expresses to the network help boost the signal? Or am I just stuck with a lousy network topology in a lousy house for WiFi?

  • Tesserax Level 8 Level 8 (49,260 points)

    I can't really relocated the router. Would adding additional Expresses to the network help boost the signal? Or am I just stuck with a lousy network topology in a lousy house for WiFi?

    Possibly, but unlikely. Another option would be to explore the use of Powerline adapters. These adapters take advantage of your home's electrical circuit to create a pseudo Ethernet one. You would place one of these at base station location, and then, connect the corresponding base station, via Ethernet, to the adapter. In this case, all of your AirPort would be connected by "Ethernet" instead of wireless and you could configure the whole setup as a roaming network.

  • dashlapak Level 1 Level 1 (0 points)

    Does the "roaming" set-up work with Express devices as well as full-size base stations? The KB article specifies that the "extended" stations are connected via their WAN ports, and Expresses don't have those, right? Just Ethernet LAN?

     

    But, the basic drill would be, plug the FiOS router into the primary base station, the base station into the powerline adapter, and the adapter into the wall? Then, wherever I want to boost throughput, plug in another powerline adapter and connect that via Ethernet to another Airport device? With the whole network configured per the KB article you referenced?

     

    By the way, I very, very much appreciate the time you're taking to help me out with this.

  • Tesserax Level 8 Level 8 (49,260 points)

    Does the "roaming" set-up work with Express devices as well as full-size base stations? The KB article specifies that the "extended" stations are connected via their WAN ports, and Expresses don't have those, right? Just Ethernet LAN?

    Roaming not only works with the Express base stations but also with routers or wireless access points from other manufacturers as well. Since only the "main" router will be configured as a router (NAT & DHCP enabled), all other base stations will be configured as bridges. In this mode the Ethernet port(s) are just that, Ethernet ports. They are not performing as either WAN or LAN ports.

     

    But, the basic drill would be, plug the FiOS router into the primary base station, the base station into the powerline adapter, and the adapter into the wall? Then, wherever I want to boost throughput, plug in another powerline adapter and connect that via Ethernet to another Airport device? With the whole network configured per the KB article you referenced?

     

    That is correct.

  • dashlapak Level 1 Level 1 (0 points)

    Thanks. I've ordered a couple of d-link units and will experiment with this tomorrow evening. If it works, I'll get at least a couple more for my other troublesome rooms.

     

    Really, really, grateful fr your help.

  • dashlapak Level 1 Level 1 (0 points)

    So, a couple of things.

     

    First, in preparing to test out a powerline network, I took down all of my AirPort Express units. Before doing anything further, my throughput went up by factors of two or more in most locations around the house. Since AirPlay, Bluetooth, and Apple TV can handle all our media streaming needs, I will no longer be using the APEs. So, addition—really, multiplication—by subtraction. Progress!

     

    I've been utterly frustrated, however, in trying to establish the powerline network. I bought two D-Link 500MBps adapters. Plugged oneinto the wall by my router, then plugged my AirPort Extreme base station into it. Plugged the other into the wall in my dining room, where I wanted to test its performance. Went through the configuration per the manual, and followed Apple's steps for setting up a roaming network. No freakin' joy; I can't get a valid IP address out of the "remote" adapter (Yes,  I confirmed that the base station was sending a non-bogus one to the "base" adapter).

     

    Called D-Link tech support. Turns out, the "only problem" with the product is that they won't work unless they're plugged into the same electrical circuit.

     

    Gosh, was that the "only problem" with the play, Mrs Lincoln? I mean, my house is over 3,000 sq ft and must have 15 separate circuits; what decent-size house doesn't? What are the odds that two arbitrary rooms will share the same breaker? Not great, at least in my 19th century Victorian.

     

    I had read that there could be difficulty if the units were plugged into different power *phases* (120 vs 240). I checked my breaker box, and all seems to be right on that score; both circuits appear to be on the 120 phase, flowing from the main power distribution panel.

     

    What really frosted my nads was that the adapter features an LED that is alleged to indicate that the device is connected to the network and communicating with its comrade(s). Turns out, not so much. Even though both of mine are shining a lovely, steady green—indicating not just that they are in contact, but are communicating at 80+ MBps—they're really not connected at all. In fact, according to D-Link tech support, all the "PowerLine Network LED" is telling me is that the unit is receiving power. Which I thought was the job of the cleverly-named "Power LED." Silly me.

     

    Are there any powerline products out there that don't suffer from this insignificant, tennsy-weensy, little flaw?

  • Tesserax Level 8 Level 8 (49,260 points)

    Are there any powerline products out there that don't suffer from this insignificant, tennsy-weensy, little flaw?

    First, I'm sorry that you are having such a difficult go at this. Unfortunately, all Powerline adapters work on the principle that they be on the same electrical circuit for continuity. I'm not aware of how they would work otherwise. It would, of course, help consumers if more Powerline vendors did explicitly call this out in their advertising. I apologize for not pointing this out as well.

  • dashlapak Level 1 Level 1 (0 points)

    Thanks. I guess that means the technology really isn't useful to do what I'm trying to do, which is improve the performance of a wireless network around my house. The chances of my home office, at the back of the second floor, and my dining room in the middle of the first, being on the same circuit are virtually nil. Same for the office and my bedroom, my living room, and pretty much everyplace else where I was having problems with throughput. Too bad; this seemed like a really powerful (pardon the pun) solution.

     

    And I don't don't know many houses that can run on 15 or 20 or 30 amps, anyway. Doesn't that limit Powerline's usefulness overall?

     

    I mean, does your house run on a single circuit?

     

    Oh, well. At least I sped things up a reasonable amount by pulling out those darn APEs, which Iwouldn't have done if I hadn't experimented with this.

     

    Thank you again.