13

I have a router connected to a modem. My intention is to connect to this router from an office that's about 200 feet away and through 2 walls.

The router signal is not strong enough to get to the remote location, so I have put a range extender in between.

The thing that's extremely confusing is that when my computer (or any client) connects to the extender while sitting right next to it, it works perfectly (ping times of about 30ms).

However, when I move to the remote location, even though the wifi signal strength is either 3/4 or 4/4 bars, the packet loss is massive, and the latency varies wildly where for a stretch it will report pings of about 50ms, but then will drop to 800+ and about 10% are just timed out packets altogether.

So, my question is... if the signal strength is good, why would this distance create such horrible packet loss? And what could cause such interference? I'm in a residential neighborhood, but there is absolutely nothing, line-of-sight between the laptop and the signal repeater other than 2 walls (with whatever electrical wiring is in them).

0

3 Answers 3

15

However, when I move to the remote location, even though the wifi signal strength is either 3/4 or 4/4 bars, the packet loss is massive, and the latency varies wildly where for a stretch it will report pings of about 50ms, but then will drop to 800+ and about 10% are just timed out packets altogether.

if the signal strength is good, why would this distance create such horrible packet loss?

Signal strength is only one metric... also consider signal to noise ratio, which is often the problem for scenarios like this.

Wifi latency and packet loss are cousins of each other. 802.11 frames contain a sequence number that is ACK'd... if the sequence number isn't ACK'd (due to loss or a bit error in the original frame), then the sender attempts to retransmit the frame a certain number of times. These 802.11 retransmissions show up as increased latency or outright packet loss if the interference is bad enough.

I have literally seen 802.11g latency that is over 40 seconds (yes... seconds) when I'm only 50 feet from the LWAP. That particular environment had a lot of tools that also operated in the 2.4GHz bands, so obviously the potential for errors was quite high.

what could cause such interference? I'm in a residential neighborhood, but there is absolutely nothing, line-of-sight between the laptop and the signal repeater other than 2 walls (with whatever electrical wiring is in them).

Wifi operates in (mostly) open spectrum bands from the FCC... bluetooth, microwave ovens, phones, toy cars, we can only speculate about the source of the interference.

You could try using a directed antenna with a focused beam (i.e. a yagi, or a cantenna) on your stations... those might help if the interference is not in the direct path to your wifi source.

Finally, if you have a wireless sniffer or access to a linux system (suggestion: Backtrack Linux LiveCD) then you can diagnose your wifi problems with Wireshark / tshark. Cisco also has a good reference for Wireshark 802.11 display filters, which help filter out noise so you can focus on the problems at hand.

12

Wireless video cameras cause about the worse interference. You don't need other 802.11 devices or access points to cause trouble as microwave ovens, some DECT phones, bluetooth, etc. can all run into the 2.4GHz band. Line-of-sight isn't the only way you can "see" other devices competing for your channel. Wireless is a shared-medium, so even with other 802.11 devices that play by the rules -- microwave ovens do not -- that medium can get congested.

Signal-strength (RSSI) is not an indication of channel congestion. Only when you try to push something over it do you really find this out. Try using another channel. Grab the "Wifi Analyzer" for Android or equivalent for your device so you can see some (not all) of what's taking place in your environment.

EDIT: Look at Ekahau HeatMapper for a you-get-more-than-you-paid-for poor-man's site survey tool; it's free!

0
8

There are many causes for this type of behavior, and one of them is what is called the hidden node problem. This is a very common problem that is often overlooked (in my experience).

The use of a range extender can often create a situation where this problem occurs and can reduce the effectiveness of some of the solutions to the hidden node problem. In your case, the AP/router (or even other stations) may be able to still "hear" the station when it is close to the extender, but not when it is moved further away.

Range extenders are not generally a good solution in almost any situation. The main positive point they have going for them is that they are inexpensive.

I would look at other solutions, while it may cost more, it will be far more reliable. Here are a couple ideas to start (not knowing enough about your needs or the site to make any real recommendation):

  • A second AP with higher gain directional antenna(s) pointed at your target location.
  • A point to point wireless bridge solution. No matter which company you choose, I would not recommend a 2.4GHz solution.

Edit to add means to test: Gave this a bit more thought and how I would go about testing this is as follows:

  1. After hours or during a maintenance window, "disconnect" all other stations. You could do this by using a MAC allow list (allow only the extender and testing station) or changing the SSID and/or security. Make sure there isn't another SSID that they will connect to automatically.
  2. If you changed the SSID and/or seurity, reconfigure the extender.
  3. Run the tests close to the extender and further away as was done previously.

If this improves your perfornamce, then it is most likely a hidden node problem. Hidden node problems are most pronounced in a wireless environement with multiple stations connected, espeially when all of them are transmitting/receiving data. By removing the other stations from being an active part of the network, this should reduce the number of collisions that occur.

If this doesn't improve performance, then you will need to look at other sources of interference as has been mentioned in other answers.

4
  • I thought you were going to add a foil hat somewhere in the troubleshooting steps. ;-) Commented Jun 15, 2013 at 3:23
  • @generalnetworkerror, that is almost how I often feel about the "interference is the problem" answers. In the past several years I have done ALOT of wireless, and 90% of the people I come across that is their first and only answer. While I fully agree it is a major issue, in many cases it is simply not the problem, but some people fail to look beyond it. In this case, it sounds more like hidden node to me. Extenders are often not "intelligent" and they extend a signal pattern to be more oblong than circular which has the opposite effect of the hidden node fix of turning down AP power.
    – YLearn
    Commented Jun 15, 2013 at 19:09
  • For someone doing wireless for the first time, they don't necessarily understand the degree to which the interference problems exists. Some don't even understand it's a shared medium. The OP's statement of `I'm in a residential neighborhood, but there is absolutely nothing, line-of-sight between the laptop and the signal repeater other than 2 walls...' lead me to believe the interference possibilities were not well understood. Commented Jun 15, 2013 at 20:37
  • 2
    @generalnetworkerror, I am not arguing that interference isn't a problem, nor am I saying that many people don't understand how much interference is out there and how it can impact wireless. Those are good points. My point was that often people with knowledge often point to interference in almost any situation as the source of the problem and stop there. Wireless has many different kinds of problems facing it, and interference is only one of them.
    – YLearn
    Commented Jun 16, 2013 at 5:03

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.