I'm thinking of upgrading my wireless network equipment to "11n". I gather that there are "Draft-N 2.0" equipment (routers and adapters) now, but I was wondering if it'd be better to wait until the "official" adoption takes place and for "official" 11n-compliant equipment is available?
If the answer is that the "Draft-N 2.0" equipment is ok, any recommendations for both a router and desktop adapters?
Well, what problem are you trying to solve by replacing whatever you are currently using? Are you looking for: More range? More speed? More reliability? Fewer errors? Less susceptibility to reflections? Less interference? You can actually get improvements in all of these, but only a few at a time. There's no free lunch with RF. Note that you'll probably need to also replace your client radios (the ones in your computahs) to be
802.11 Draft-N 2.0 release whatever compliant. That's often difficult, expensive, or impossible. Change everything?
Does your client radio have the ability to take advantage of spatial multiplexing MIMO? If it's a single stream MIMO, as found in almost all laptop radios with a single antenna, multiple streams are not going to happen. Therefore, no speed or reflection immunity improvements. Note that when I say single antenna, I don't mean the two antenna diversity antenna system found in most laptops, which is really a single receiver, that switches between two possible antennas. For spatial diversity MIMO to work, you need one receiver per antenna.
I can't wait to see the porcupine antenna mess at the client end. More likely, it will be an internal PCB (printed circuit board) antenna, with no projecting monopoles. Also, external antennas are going to be difficult, weird looking, pricey and/or impossible.
In my never humble opinion, I would wait. My guess(tm) is that it will take a few months after the official announcement or premature leak for the manufacturers to deliver badly tested firmware and products. The hype, feature bloat, and proprietary enhancements will surely follow. Certainly some litigation such as the CSIRO mess. In most cases, methinks the major improvements offered by MIMO will be in the area of speed and reflection tolerance. I can't wait to see the numbers for Linksys and Buffalo with their 3x speed or 5x range hype. There will be benchmark tests, magazine articles, and the usual premature conclusions. In other words, a gigantic muddle that far exceeds any of the current 802.11 pre-N confusion. Fasten your seat belt please.
Everyone lies, but that's ok because nobody listens.
How about this one:
Dlink Xtreme N 450 which still hasn't appeared 9 months after the product release. Claims 450Mbits/sec thruput. It's kinda hard to read between the lines, but my guess is they bond the 2.4 and 5.7GHz bands together resulting in effectively double the thruput.
Marginally related drivel:
"Beamforming: The Best WiFi You?ve Never Seen."
134Mbits/sec on 5.6GHz band. About half that on 2.4Ghz.
Thanks for your thoughtful and reasoned response. Some (maybe a lot) was "over my head" (I'm technical, but wireless/radio tech is not my specialty... for that, I'm mainly a "user").
To answer the question you asked, I'm mostly looking towards/hoping for faster speeds on wireless.
When we built our current house, awhile ago, I had pre-wired some of it with CAT5, but over time, I'm finding that there were places that I wished that I had wired-to.
So, I've been slowly dabbling (not in a huge way) with adding wireless to the desktops in my house.
Members of my family have laptops, but wireless speed with those is not a primary issue (to me :)).
I currently have a couple of 11G routers and/or 11g routers setup as APs (for coverage), and a small NAS that I use to share files within the house and as a personal FTP server, but I've been seeing ads for "Draft N 2.0" routers with prices that are pretty cheap, so was thinking that maybe it's time for "the next step".
Anyway, as I kind of suspected, it sounds like from what you said that it's probably a little early. Even though, professionally, I guess that I'd be considered "high tech", I tend to be less-than-bleeding-edge, and more "cheap tech" than "high tech" :), so I'll probably just wait a bit to see how all of this "N" stuff pans out.
Following this NG, among other things is helpful with respect to that.
Well, speed usually means you have to have high signal strength and reasonable signal to noise ratio. That usually means line of sight and no interference from the neighbors. That's the real problem with getting the dramatic increases in speeds that 802.11 Pre-N promises. Unless you live in a tunnel, homes are rarely arranged for optimum wireless coverage. Building materials are absorbers, reflectors, or both. The neighbors streaming video link is spewing interference continuously. Meanwhile, the local government has installed a municipal wireless access point nearby. Meanwhile, someone has a leaky microwave oven that belches kilowatts for interference and effectively shuts down Wi-Fi in the area. Meanwhile, your neighbor has an ancient 802.11b wireless access point, that monopolizes huge amounts of air time, thus reducing your thruput dramatically. Your nightmares may vary. Needless to say, the aforementioned RF hell is not particularly ideal 802.11 Pre-N. Under such circumstances, you'll be lucky if Wi-Fi even works, never mind getting a speed improvement.
I've worked with various architects on such things. My rule of thumb is that there should be a network outlet on BOTH sides of a doorway. It's really no problem running CAT5 along the baseboards to other parts of the room, but crossing doorways is a big problem. Nobody like to see CAT5 going over a door frame. That means at least 2 outlets per room. Personally, I prefer to run conduit, which allows for fiber optics, RG-6/u for coax, alarm cable, thermostat wiring, intercom, low voltage lighting control, solar control, etc. The problem is that conduit is considerably more expensive than just hanging the wires inside the wall. It requires planning, pull boxes, decorative cover plates, etc. It's rarely done except on new construction. Most of the time, I can convince the architect to install conduit between floors and to the roof or attic, which are a big help.
The idea arrangement would be to have an access point in each room, where the rooms have aluminum foil backed insulation in the walls to block most of the RF. The access points will need to be on different channels, but there are really only 3 non-overlapping channels available (1, 6, and 11). There's always a chance of creating your own interference if you have more than three AP's.
Much better is to switch to 5.7Ghz. There are more channels that can be bonded together to get more speed. If you look carefully at the specs for the newer dual band wireless routers, the 5.7GHz speeds are usually about twice that of the 2.4GHz section. Interference is much less on 5.6GHz. Most recent laptops come with dual band client radios. Otherwise, they can be retrofitted with newer wireless cards.
Well, your quest for more speed might be better addressed by tinkering with better antennas on the access points. Low speed can be due to many things, but usually it's lack of a sufficiently strong signal. Because of spatial multiplexing, even a lousy signal can have improved speeds if there are reflective paths available. However, I would guess(tm) that you can get a bigger improvement by doing whatever it takes to improve the signal strength. Of course, if you have an interference problem from the neighbors or municipal wireless, no amount of technology is going to help other than moving to 5.7GHz.
I would suggest you find out why your speeds are inadequate first, and then consider the solutions available. Benchmarking thruput would be the first step. See Iperf and Jperf. (I prefer Jperf)
Setup a wire connected (not wireless) fast machine somewhere on your network. Run the iperf or jperf server on this machine. Walk around the house testing thruput using iperf or jperf on a laptop. See where you're at now. If you're at the low end of the wireless speeds, such as 0.5 to 6Mbits/sec (IP thruput is about half the connection speed), then no amount of new acronyms is going to help. If you're at the high end, such as 25Mbits/sec, 802.11 Pre-N spatial multiplexing will help you go faster.
Lots of good stuff and major doses of reality can be found on SmallNetBuilder.com.