I hope I have researched this issue long enough to justify a post here.
formatting link
wireless router has 4 ports for non-wireless 100Mbps connections. All pcs and laptops have WindowsXP w/SP2 and latest updates. I also have the latest firmware and drivers on all machines. I have a 10 Mbps cable internet connection 6that plugs into the router from the cable modem. I have one PC plugged into the router on one of it's 4 additional ports. I can speed test my connection speed at approx. 9.6 Mbps on anything plugged directly into the router. I have one PC with a USB wireless adapter and three laptops (one Dell, one compaq, and one HP/Compaq). All are WirelessG, all updated, ect. I have tested with all the wireless devices and can only get about 4 Mbps, regardless of the location (it is all in a 2 bedroom apartment). I always test with all machines turned off except the one I am trying.
This implies that the bottleneck lies with the router or wireless in general. Should I expect this much of a fall off in throughput? I use WPA TKIP encryption for wireless devices.
If you have seen this issue before with this router, please let me know. If it is to be expected, it would be dissapointing. I wonder if there is an issue with the router. It functions normally other than this issue,
Are any of your unspecified model PC's and laptops using an 802.11b only wireless device?
Since you didn't bother to explain how you're testing the speed, I'll assume that you're using your ISP's online speed test. That's limited by the speed of the cable modem. Looks like you're on Road Runner. It doesn't matter how fast your wired or wireless connection is running, the speed will be limited by the speed of the:
Cable ISP.
Cable modem.
Ethernet router section.
Wireless access point built into the router. Your wired connection test only includes the first three, so any slowdown in the wireless section will not appear.
The ethernet part of the router might limit the speed. See the chart at:
Ooops. The WGR614 isn't listed or reviewed. Oh well.
So, you get to do your own benchmarking. Go unto:
and download IPerf 1.7. On one of your unspecified wired PC's, run the IPerf as a server as: iperf -s On another PC or laptop, run it as a client with: iperf -c ip_address_of_server You should get the TCP speed. With a wired 100baseT-FDX connection, you should get about 80-90Mbits/sec TCP thruput.
Now, replace the wired connection from the client PC or laptop with a wireless connection and do the same test. If the client is fairly close and you're getting a good solid 54Mbit/sec connection, you should get about 25Mbits/sec thruput with one wireless link.
If you get the expected speeds, it's working as expected (duh...).
Now, take a walk with the laptop while running the test. As you get farther away, record the approximate distance, connection speed, and TCP thruput from IPerf. Extra credit for being organized and generating a neat graph. The farther away you go, the slower it will go.
At one point, you're going to hit 9Mbits/sec. That's the maximum speed of your cable connection. Any farther away, and your wireless connection is going to become a bottleneck for internet access.
That's nice. Any particular reason you don't want to disclose the makers and model numbers of your hardware? Perhaps you don't want me to suggest that you purchased garbage, the wrong hardware, don't have the latest firmware, or that the product might have a known problem?
That means you're testing over the internet, using some unspecified internet speed test, and possibly in the same or next bedroom (hard to tell from your description). Obviously, something is wrong because with a 54Mbit/sec connection in the same room, you should be getting the full 9Mbits/sec from Road Runner, or 25Mbits/sec using IPerf.
4Mbits/sec sounds like you're stuck at 802.11b speeds. That's about what you'll get with an 11Mbit/sec 802.11b connection on a bad day. Wanna disclose your hardware?
Encryption has very little effect on thruput.
It might be the router. I use quite a few WGR614v6 routers for non-critical home application. However, most of my customers are no DSL which doesn't do much faster than 2.5Mbits/sec. No clue how they work at faster speeds. My guess is that there's something odd with your unspecified collection of client radios and computahs.
I have a 10 Mbps cable connection. The cable modem is plugged into the WRG614v6 router. I turn off all computer equipment to avoid interference with the test and reset the router to factory defaults. I have a brand new HP/ Compaq NC6400. If I plug it into a port on the router, I can "speed test" to any of several online sites at 9.6 Mbps. (I live in Dallas and use the "dfw.speakeasy.net" speed test"). No problem with internet/ethernet. Enable wireless connection on the laptop and unplug from the router and "speed test" drops to 4.4 Mbps, whether I am a few feet from the AP or in the next room. The signal quality is "Excellent" at all times. This same test has the same results with my Dell Inspiron 600M w/internal wireless G, Compaq Evo with Wireless G card and Dell GX270 with Linksys compact USB Wireless G adapter.
I either have a faulty AP in the WGR614v6 or that model cannot handle more than 4.5Mbps an the wireless side. This is the information I was seeking. I will likely only know by replacing the router/AP with one that is known to function at 10Mbps or more. Though I hold an MCSE:Security, CCNA. CCA, and COMPTia Security+ and am pretty familiar with general wireless networking, I have never witnessed this particular issue before. I was hoping to hear, "Oh yeah. that model of router blows over 4 Mbps" or something like that.
I do appreciate your efforts though, as well as further ideas.
Actually, I was hopeing you would supply some of the missing detail. More simply:
What problem are you trying to solve?
What do you have to work with? (hardware, software, versions, numbers).
What have you done so far and what happened?
I'll be pedantic and point out that you also didn't bother to mention the maker and model number of the cable modem. If you really have all that MS and Novell wallpaper, you should know better.
The defaults are not always the optimum configuration for wireless. For example, some older routers slow down with the wireless essentially scanning between 802.11b, 802.11g, Turbot-G, etc. However, the difference is too drastic in your case, so this is unlikely to cause the speed problem.
Resetting the WGR614v6 also eliminates having your router stuck in the
802.11b only mode, which was my best guess. However, you might want to check the settings to make sure you have actually reset the router and that it's NOT stuck in the 802.11b only mode.
Runing what operating system? Updated with the latest MS or Linux updates? Once again you're leaving out important information. What wireless device? Drivers from HP or from the manufacturer? If XP, are you using Wireless Zero Config or the manufacturers driver?
Agreed. You're getting the advertised cable modem speeds.
I understood that in my last reply. I offered a way to reduce the number of possible culprits by using IPerf. That would eliminate the cable modem and the ethernet router part of the puzzle, leaving only the access point section and the ethernet switch. The test will also show error statistics as will your unspecified operating system diagnostics.
Back to being pedantic. How many feet are a few feet? What signal strength (in dbm) is "excellent"? At what SNR (signal to noise ratio) or noise level? Numbers, not generalizations.
Again, I note the partial lack of model numbers and operating system versions. However, as 3 seperate computers and wireless devices apparently exhibit the same behavior, it's mostly likely a problem with either the wireless section or the RF environment.
It's also possible that you live in an RF polluted environment.
However, the mostly likely problem is something wrong with the wireless section of the router. I can't tell from here. If the laptops all function normally through a different wireless router (i.e. coffee shop hot spot), then by process of elimination, it's the WGR614v6.
Well, the easiest way to troubleshoot anything is by replacement. If you can borrow a different model router and try the same test, the conclusions will be definative.
Your unwillingness or inability to supply numbers and details is not very impressive.
I have seen this problem before, but not with this particular model. The problem turned out to be a nearby municipal wireless system that was trashing the channel. Changing channels solved the problem. In most cases, it is just plain bad firmware which was solved by an update. If replacing the router with a different model does NOT solve the problem, methinks you should consider doing a site survey with a spectrum analyzer.
I can usually look at layer 2 (MAC layer) statistics for the wireless device and determine what is causing problems. Same with sniffing the traffic and analyzing the retries, retransmissions, collisions, and corrupted packets using Ethereal.
Well, why didn't you just ask that question? Unfortunately, the router performance comparison at:
does not offer this specific model. That's why I wanted you to actually measure the thruput, without using an online speed test. There are also some router and client combinations that simply hate each other. Eventually, these also get fixed with firmware updates.
I checked various reviews, but there's a problem. The different hardware mutations of the WGR614 all have different chipsets and therefore different performance levels. Here's one that might be a v6 running at 22Mbits/sec wireless throughput.
No furthur ideas. Do the Iperf speed test and see if the wireless is really the problem. If you think that your firmware update caused a problem, try flashing the firmware again followed by a hard reset.
Incidentally, I do NOT have an MCSE, CCNA, CCNE, or other certifications. So far, I haven't needed them and haven't had the time to study for them.
I just setup a WGR614v7. Not the same as v6 but still should show similar performance. Connected to a Comcast cable. Speed test (nitro.ucsc.edu) showed 14Mbits/sec download and 1.5Mbits/sec upload with ethernet connection. When I switched to wireless, it was approximately the same. A fast IPerf test showed about 22Mbits/sec with a 54Mbit/sec wireless connection in both directions. The only gotcha was that it wouldn't even get a DHCP IP address from Comcast until I updated the firmware.
Cabling-Design.com Forums website is not affiliated with any of the manufacturers or service providers discussed here.
All logos and trade names are the property of their respective owners.