Slashdot Mirror
When 54 Mbps isn't 54 Mbps: 802.11g's Real Speed
eggboard writes "Matthew Gast, author of 802.11 Wireless Networks, filed this article for O'Reilly Networks explaining exactly how fast 802.11g really is: that is, what's the actual data payload and real throughput, not the rated maximum speed. His conclusion? In mixed 802.11b/g networks, which will be common for years to come, g is only 1.6 to 2.4 times faster than b, not 5 times faster as it is in its g-only mode. This article has real math based on the specs, rather than armchair speculation."
Gigabit ethernet is supposed to be 100 times faster than good ol' 10BaseT. It is, at the root layer. Most devices can't push that much data through the pipe, and with wireless, there is MUCH more error correction that needs to be done in communicating back and forth. Wired networks (normally) don't have the kind of interference that 2.4 GHz-band devices now suffer from.
When I said "armchair speculation," I was referring to the mass of articles that come out that talk about Wi-Fi speeds without actually looking at how the technology works.
Matthew has now provided a baseline. Someone could now perform real-world benchmarks against these theoretical maximums which are built into the standard.
Matthew's numbers provide optimal performance guidelines for network planning. Real performance will, of course, be even lower.
Freelance tech journalist for the Economist, MIT Technology Review, Macworld, and others
Oh, so I only get a 60% faster connection? Given that soon enough the price differential between B & G will be gone, I still think G is the superior choice. When the wireless cards are only $15 to $20, I think that pure G networks will be much more common. And then you will get much higher throughputs.
Maybe they should go after Dannon yogurt for decreasing the size of their container to 6oz from 8oz, but keeping the price constant. Then at least they would be reporting on something I could care about.
No, I don't trust in god. He'll have to pay up front, like everybody else.
However, when you connect a 10baseT NIC to a 100baseT switch, you don't slow down the rest of the connections to the switch, which can still operate at 100baseT. The situation with wireless, a shared medium, is more analogous to connecting a 10 baseT NIC to a 100/10 baseT auto sensing hub--when you hook up that 10baseT card, it slows down the rest of the hub to 10 baseT.
"Furthermore, the model ignores the sophistication in the TCP acknowledgement model. To avoid constraining throughput, TCP uses "sliding windows" and allows multiple outstanding frames to be transmitted before acknowledgement. In practice, TCP acknowledgements can apply to multiple segments, so this model overstates the impact of higher-layer protocol acknowledgements."
This reduces the "TCP" he uses to a stop-and-wait protocol.
You didn't even look at the article, did you? There was no testing. The author didn't model TCP windowing at all, and he even failed to take delayed ACKs into account.