What exactly is wrong with this picture?
Visiting the computer repair shop around the corner (yes, I’ve got one of those within walking distance of the house), my eye was drawn to the wireless access point devices on sale there.
“300 Mbps Wireless N Access Point” is what it says right on the front of the box.
But then I pick it up and look at the panel showing the Specifications.
“One 10/100M Ethernet Port”.
Now I’m thinking, “How can it give you 300 Mb/s when its hardwired connection to the rest of the network (presumably a hardwired router or perhaps even a cable/DSL modem) is only 100 mb/s?”
I could see this being possible if the access point is itself connected to a router over wireless N that was delivering 300 mb/s, but that’s not the use case I see most people in (sure, I’ve got a bridge or two around my property, but I’m an engineer!).
Don’t get me wrong, I really like the manufacturer whose product is featured in the pictures above. They make very affordable yet high quality equipment. Unlike a lot of vendors, particularly well-known ones, they deliver on their promises.
So what’s the deal? The answer from the wireless device industry is: channel bonding, the “magic” behind high speed wireless N.
To get additional bandwith over a wireless link the 802.11n spec allows a client to connect to an access point using two separate channels simultaneously. If each of those channels is providing a theoretical 150 mb/s (the maximum for a single wireless N channel), combining them together gives you 300 mb/s.
But a 100 mb/s wired connection to the access point is going to limit the rate at which that data is going to travel across to the rest of your network, or the Internet. A 1000 mb/s, or 1 gb/s, interface would make a lot more sense under the circumstances.
Of course real life speeds on the wireless part of most home networks never get near 100 mb/s, let alone 150 mb/s or 300 mb/s. A variety of factors, including having to run in mixed mode so 802.11g and even b clients can use wireless, as well as channel congestion caused by neigboring access points (a growing problem made worse by those who use channel bonding to occupy double the amount of radio spectrum), really make those kinds of speeds unlikely.
The whole problem with wireless networking is that the demand for the radio spectrum needed is seriously outstripped by the amount of spectrum available.
Until the authorities who control that spectrum make more available, or we see some fundamental improvements in the efficiency of its use, a wireless spec that provides for 300 mb/s connections is just an empty promise.