Wi-Fi

Real-world tests of 802.11n technology: How does it rate?



Last week, I took a look at what various hardware providers are saying about the performance of their new 802.11n or 'Wireless N' product lines. The general message was that the next generation of wireless systems are faster and more powerful than ever before: up to five times faster along with up to twice the range. These high throughput devices are designed specifically with the needs of today's users in mind with high quality streaming media being consumed at an increasing rate.

Anybody who's used a wireless access-point or router will know that you never really get the types of range or coverage claimed on the box. Real world conditions tend to interfere and dramatically reduce the access point's coverage. Structural barriers like walls and doors will of course greatly reduce the range, while electronic interference coming from microwave ovens and mobile devices will cause throughput to suffer. I read a recent summary of the Farpoint Group's research on the causes and effects of interference; microwave ovens caused throughput to be reduced by up to 62%; Bluetooth headsets reduced throughput by 20%; and some wireless devices such as cordless phones reduced throughput by 100%. Now these examples are pretty extreme but they do highlight some of the many sources of interference in an average household or workplace.

So how will 802.11n devices really perform in the home? I've taken out my old Netgear DG834PN 'RangeMax' modem/router and replaced it with the DG834N from Netgear's latest 'RangeMax Next' line in hope of finding out whether or not 802.11n is worth all the fuss. While the DG834PN is capable of 108 Mbps if using the correct client adaptors, I have been using it in standard 802.11g mode (54 Mbps) as my wireless adaptors are 802.11g/n; not the Netgear 108 Mbps units.

While the external appearance of the two routers couldn't be more different (gone are the flashing blue lights), in terms of hardware, the two devices are exactly the same -- a reset button, ADSL connector, power socket, and a four-port hub. All very straight forward. In terms of configuration, there is little to do. Make sure everything's plugged in to the right place, switch it on, browse to the admin page, and start the setup wizard. The router attempts to detect the connection type and automatically configure itself. In my case this worked brilliantly, my only input being the ISP-provided username and password. The actual software of the router is identical to that of the previous model which is nice as I don't have to learn anything new! Of course there is an additional option in the wireless configuration menu: 270Mbps! It's worth noting that if you're using the wireless access point at full speed then there are only two options for encryption: None or WPA-PSK [TKIP] + WPA2-PSK [AES]. If you have any older wireless clients that only support WEP encryption then it's going to be time to do some upgrading -- this is really for the best as even home users should avoid using WEP; it's too easily broken.

That's the installation covered. Now for some simple tests. I must point out that these tests are not carried out with scientific precision; they are simple performance indicators in a real world environment to give me an idea of how the two technologies compare for day-to-day use.

The first test is of throughput. Grab a 1-GB file (in my case a .vob) and see how long it takes to transfer from a wireless machine to a computer connected to the wired LAN ports.

'Wireless G' transfer time: 9 minutes

'Wireless N' transfer time: 6 minutes

The second test was throughput again, but this time from one wireless client to another. The test is with the same 1GB file.

'Wireless G' transfer time: 14 minutes

'Wireless N' transfer time: 9 minutes

The third test was again of throughput, but this time rather than measuring the transfer times of a specific file, I wanted to see how streaming media performance varied. To do this I had one machine playing a 192-Kbps iTunes radio station and a second machine playing music from the shared iTunes library of the first machine. Both machines use wireless networking of 802.11g / 802.11n. The results of this test were interesting; I found that while using 802.11g both the first and second machines played music faultlessly, no jitter or re-buffering required. While using 802.11n, designed for HD streaming media with 'Steady-Stream' technology -- I actually had some problems. Even with the buffering set to maximum, the Internet radio station frequently cut off and decided to re-buffer. I would have put this down to random Internet congestion except streaming music from the shared iTunes library also suffered, constantly stopping to re-buffer.

My final test was range. Very simply, I took my notebook and moved down my garden until the signal became unusable. The DG834PN 'RangeMax' router had actually significantly increased my range over the unit I had previously installed (a standard Belkin 802.11g modem/access point); with the DG834PN I would get about 50 meters down the garden while maintaining a useable connection. I was pretty happy with this considering the router is actually in a cupboard under the stairs at the centre of the house. The 'RangeMax Next' again gave a considerable improvement. I found with the DG834N that I gained another 30-40 meters, allowing me to sit right at the bottom of the garden. I'm sure this will be appreciated in the summer, allowing me to disappear to a quiet spot!

Overall, I have to say that I'm happy with the performance increases. The throughput has increased by around 30% which can only be a good thing. It means copying large files over the wireless network is a little less painful than before! Increased range was not particularly important to me as the previous 'RangeMax' unit performed exceptionally well; however, the added coverage is a bonus, and I certainly can't complain about it. The only disappointment comes with the loss of quality and consistency while using streaming media services. It seems particularly strange seeing as Netgear are pushing their 'Steady-Stream' technology: "…a technology called 'Steady-Stream', which ensures your wireless network offers a steady constant connection, with sufficient bandwidth available to stream, download, browse the Internet, and transfer files all at the same time. With RangeMax™ NEXT Wireless N technology you should not experience any more frustrating interruptions or freezes, just a constant connection that enables you and your household to use the Internet and wireless network as you choose, whenever you choose." Hopefully, the next firmware update will resolve this issue…

14 comments
darob1990
darob1990

Interesting, here it is 2011 and I found your article after discovering my new D-Link dual band N router streams media better while broadcasting in g mode no matter how I tweak it. Perhaps there's a reason the n protocol hasn't been finalized, it's still not there yet, 4 yrs later.

philipsmith23-subscriptions
philipsmith23-subscriptions

Installed as soon as it came out, DG834N & the correct client adaptors on 3 laptops, 2 Desktops plus 2 x-Box. Performance is superb, would recommend to everyone.

tyhoward
tyhoward

Glad to hear that your only problem was with streaming data on your tests. Unfortunately I am involved with a disaster concerning ?N? technology routers for a customer here. So far I have tried two WRT300N Linksys routers and a NETGEAR WNR834N router with no success. Requirements were metal building to metal building 200 feet apart. Don?t need more than 11mbps solid and stable connection for the second building. Second location has 300n PCI cards with external antenna mounted on the desks. Tech support at Linksys recommended the 300N as the optimal solution, and assured us that it would work within those parameters. First router failed consistently and completely stopped transmitting wirelessly at random intervals. Testing with two different laptops (G cards including Linksys brands) within 15 to 50 feet could not retain a solid connection (inside the first building or within 25 feet radius of building). 4.5 hours and 3 service calls to Linksys recommended that perhaps the building was the culprit (no mention of replacing the router at this point). I sent the router back and paid for a second identical unit to avoid the 7-15 day delay in delivery on the replacement! Second WRT300N unit worked fine for 16 hours ? no dropping out completely. The problem I had was that I could only connect to it within 80 ?120 feet from the router itself. I was assured by Linksys that placing the router was the key, and that it should work. It did not. I replaced the Linksys with a Netgear WNR834N, still retaining the Linksys PCI 300N cards in the second location, figuring that we could get them working with no problem. Since the N draft units are supposed to be so much better on range, I have been unable to find any external antennas, which can be connected to the routers (unlike those range extenders that are available for the ?G? environment). I was a little concerned that the Netgear unit has already had four firmware upgrades in less than one year, but updated to the latest before beginning with that one. It constantly disturbs me that I can see another network within 425 feet at both of these locations. That is that I can see it from sitting inside building number one with the G card on the notebooks, and I can see it while sitting at each of the computers inside building number two ? with the PCI N card computers and two different laptops. I cannot connect to it but it sees it at about 15% strength consistently from within each of the buildings. It is running a Linksys 54G router, which similar to yours was sitting haphazardly on a shelf in the middle of that building (yes, I did go to the other business to ask to see how they were setup and what they were using). It is a standard out of box router, just plugged into their standard DSL modem. With onsite and technical support calls, we now have about 9 hours total investment, with no good solution available. I will be taking out the N series PCI cards and the N series routers, and replacing them with G series units until the promised increased range becomes a reality in my world! Thanks.

trfh
trfh

Your test results are interesting. I haven't studied the technical changes between g and n, however as I understand it you have tested a "mixed" network, i.e. n one end and g the other. Presumably the n device has been constrained to operating in g mode to maintain compatibility here? So - being devil's advocate here - the manufacturer might say this isn't a fair test of their n kit. In particular the ability to stream sound/video seems to have suffered - and your n/g network is now worse than the older g/g network. When you get around to it - a test of an n/n link would be interesting. Richard

jim
jim

It would be interesting to have a comparable story where you replace your current system with the Linksys, D-Link, Belkin, etc. products including their compatible wifi client cards.

rvolkman
rvolkman

The problem with 802.11n in the 2.4GHz band is going to be channel space. The problem we see now in shared office spaces is channel interference and 802.11n will only make this worse, it would appear, due to the wider channels and the increased coverage of MIMO transmission. The 5GHz band would be better, but then range issues start to come back again. What's really needed is a wider range of 2.4GHz channels. 802.11n may work well in an isolated residence where it will be great for wireless multi-media, TV, etc. RV

Justin Fielding
Justin Fielding

Not quite correct, the 'n' and 'g' figures are two different routers: With the 'n' standard router I have used 'n' compatible client adapters (both my iMac and Macbook have the Airport Extreme upgrade) and with the 'g' standard router I used both 'n' adaptors in my Macs and 'g' PCMCIA adapters while running from a Linux LiveCD. It didn't seem to make any difference whether I used the 'n' or 'g' client adapters with the 'g' standard router. I guess this means wireless 'n' client adaptors are 100% backwards compatible (at least the ones I have used so far seem to be).

Justin Fielding
Justin Fielding

Send me a bunch of free hardware and I'll be happy to post test results. Without that I'm afraid I can only report on what I have. :)

billy.blak
billy.blak

i replaced my dlink g with there n router and cards and the improvement was awesome.i can now sit in the front basement with the router in the back 3rd floor and get excellent reception.havent checked distance from house yet.they were slick,said the n desktop card was backward compatible(not)had to buy the matching router.all for good made me improve security.router works with g cards.book states do not guarantee product to be fully functional in mix mode

Neon Samurai
Neon Samurai

Granted, you can't just run cables through a historical landmark class building and if that is the case, wifi is your best option. It sounds like the first person is running imobile desktops off wifi cards. Wifi between buildings was a nice thing to try and much less expensive then a microwave dish link or rented security pair ISDN line. ah.. guess the investments made though. No expensing all those pci boards now that they are there.

Neon Samurai
Neon Samurai

You may want to confirm that your router's manufacturer can provide a firmware upgrade for 802.11n when the ISO finally decides on a standard. Back in the modem days, a few people got stung by choosing the wrong 56k competing pre-standard then couldn't upgrade the modem when a 56k standard was made official. Just a heads-up incase it effects your hardware. Since your running a paired NIC and router, it's probably moot.

Justin Fielding
Justin Fielding

I agree, although I have both wired and wireless connections available in my office I tent to stick to wired unless I'm moving between meeting rooms etc. It's so much faster and more reliable. At home I almost exclusively use wireless as I'm generally not transferring huge files internally and the convenience can't be beaten.

Neon Samurai
Neon Samurai

I think the original question specified a business in two seporate buildings using desktop workstations (wifi cards are pci not pmcia). For mobiles; absalutely, it hurts to go back to a wired connection after having wireless freedom to lounge in your easy chair. As a business, I hope they had more than convenience as justification for running a desktop office on wifi but it's neither my business or client contract to question.

jdclyde
jdclyde

Wired vs cables. People forget that when you are running wireless, you have another program running on your system, taking up more resources, to control the card and connection. You also have more overhead on a wireless signal, so when you drop from 10/100 to a G card, you really have lost more than half of your bandwidth. Then you look at what your doing. If your accessing the internet and only have a 5 meg connection, does it matter if your running at anything more than 5 megs? Convenience does go a long ways, because of that. I LIKE to relax in my easy chair with my laptop. I don't LIKE to have to deal with anything more than my mouse cable. (the arms are flat wood, and perfect for a mouse!)