The Problem to Be Solved: Horrendously slow Wi-Fi in a large apartment building.
The Cast of Characters:
- Gigabit Wired LAN
- 6th Gen Apple AirPort Extreme
- 50 Mbps symmetric fiber connection (shout out to Ellum Net, the best ISP. EVER.)
- Two MacBook Pros
- iPad Air 2
- iPhone 6+
- Nest Thermostat
- Belkin WeMo Switches
- Two Apple TVs and Rokus
- A Very Frustrated Nerd
The Backstory: I moved into a new building in March 2015. It was an awesome building. Over 100 years old. Ford Model T's were once built where I now sleep! I was able to ditch Time Warner Cable and go with an indie ISP. Life was good.
My household has always been streaming-only when it comes to TV; everything I watch comes in via Netflix, Hulu, etc. As I settled in for my first night in the new place after a day of moving and unpacking life went from awesome to abysmal - the streaming quality was terrible! My "Friends" we're nothing but big pixelated blocks. I didn't understand how this could be with my big fast internet connection. Being the honorable nerd that I am, I set out to figure out what was going on. I was using the AirPort Extreme setup with both 2.4 GHz and 5 GHz networks both named Kennedy and a 2.4 GHz guest network named Clinton. Yes, I have a fascination with former Presidents.
The Process: Let the testing being! I first started with a good old fashioned speed test on iPhone 6+ using the Ookla speedtest app. Enter the problem: I was shocked when I saw I was getting only around 2 Mbps. I was certain my internet connection was better than that! When I wired up one of the MacBook Pros directy to the AirPort Extreme, I got around 90 Mbps both directions. Something wasn't right! Tested the MacBook on Wi-Fi and got the same 90 Mbps result.
After using "the Googles", my working theory was that because I live in a densely packed area of town in a large loft building the Wi-Fi frequencies were overloaded. Now, how to test this theory? Was it possible to test? Remembering back to wireless frequencies and that lower frequencies penetrate walls better, I wanted to split my 5 GHz network from my 2.4 GHz network and test them separately. Would the 5 GHz band be less crowded and stay "inside" my loft?
The Solution: Once I seprated Kennedy on the 2.4 GHz frequency and Johnson on the 5 GHz frequency (see what I did there? VP) I tested iPhone, iPad, and Mac on both bands. Sure enough, around 2 Mbps on 2.4 GHz and 50+ Mbps on 5 GHz.
One Last Burning Question: The only question left was why did the iOS devices not prefer the 5 GHz network over the 2.4 GHz network when they were identically named? Without an Apple Engineer, I can only speculate, but it appears that they prefer networks based only on signal strength. This is my working assumption as most places in my loft the Wi-Fi "seashell" signal indicator shows only 2 of the 3 bars on Johnson (5 GHz) while showing full strength on Kennedy (2.4 GHz). All devices that support 5 GHz are now configured to only connect to Johnson, and I use Kennedy for devices that require 2.4 GHz - Nest, Wemos, etc. Problem solved - speed is king!