3

I regularly get ping times (to google.com) on my netbook over wireless (802.11) of:

  1. 500-4500ms for about half the time
  2. 100-150ms for the rest of the time.

It seems to switch back and forth on the order of minutes.

Regression:

It doesn't seem to correlate with anything I can see:

  • location (I've seen this when using at least three different networks in two different states),
  • proximity to the AP,
  • other people moving around.

It's been that way almost from the get go. If it weren't for the fact that it does get good performance some of the time I'd write it off as bad antenna design. Any ideas?

p.s. How long can ping times get? I've seen 3-5 sec here and once witnessed (as a result of a borked router config) >90 sec.

BCS
  • 1,065
  • 2
  • 15
  • 24

3 Answers3

2

100-150ms is pretty poor for a best case scenario even over a WLAN assuming you have a decent wired network behind it. 500ms+ is definitely problematic.

Ping is a pretty blunt indicator to use to diagnose network health\performance but at least in this case it is telling you that something is wrong. Assuming for the moment that there isn't a problem outside of your WLAN there are a lot of things that can affect an 802.11 WLAN - many of which will cause ping times to drop. Here are a few to consider.

  1. Interference from other WLAN's - there are only a limited number of WLAN channels and the more active networks there are around you the higher the likelihood of interference. This is particularly true for 2.4Ghz (802.11b/g and most 802.11n variants) not such an issue for 802.11a and 5Ghz 802.11n.
  2. Traffic contention on your WLAN - the more active clients there are that are on the same WLAN as you are and connected to the same ASP the less bandwidth you get. As client numbers pass into the 10-20 per AP range for the aforementioned 2.4Ghz WiFi performance tails off fairly fast.
  3. Interference from non WiFi RF - things like Microwaves can stomp all over some 2.4Ghz WiFi channels. Microwaves can be fairly obvious but I've seen this happen with a lot of industrial\medical kit so if there are any big machines that draw a lot of power nearby then this can be an issue.
  4. Aggressive WLAN roaming settings - if you are close to more than one AP on the same WLAN your WLAN adapter may decide to periodically hop from one to the other if roaming aggressiveness is too high and the RF environment around you is variable (e.g. you are in a crowded area with lots of people moving around - humans are great absorbers of WiFi signals).
  5. Active QoS on the WLAN that is de-prioritising your traffic - this is not uncommon on Enterprise WLAN's that prioritize Voice over WLAN traffic for example. This would be quite unusual to be honest.
Helvick
  • 20,019
  • 4
  • 38
  • 55
  • Expanding on your #3: almost anything cordless: some phones, tv-transmitters, baby monitors, alarm systems, .. – Joris Sep 17 '10 at 06:13
  • While all of those are possible, I think I've seen my issue in a case where I could rule out every one of them (one at a time, not all together) at some point or another. I'd accept it was different cause at different points but reluctantly because the performance is much more consistent than the environment. – BCS Sep 17 '10 at 14:53
  • The issue could be upstream of your WLAN too of course - can you bypass the WLAN component of the link and if so what happens to your ping times over a suitable test period? Also I've seen poor Antenna design trigger item 4 - used be an issue on IBM T-30's. – Helvick Sep 17 '10 at 15:28
  • I've seen this from AP's admined by a dozen or so places on at least 3 different networks in 2 (non adjoining) states. – BCS Sep 17 '10 at 17:00
  • All with the same machine? If so there's only one thing consistent between all of them surely, the PC itself. I get <50ms pings to Google across a wide range of WLAN environments, and <=1ms on pings to other systems across a healthy WLAN. – Helvick Sep 17 '10 at 17:16
0

Along with what Helvick has identified as point 2. Any chance you are continuously running any application(s) which may be accessing and/or bursting network transfers (e.g. P2P-type application, cloud backup/synchronization, etc.)?

user48838
  • 7,431
  • 2
  • 18
  • 14
  • Not unless it's hiding it's traffic. – BCS Sep 17 '10 at 14:52
  • Please elaborate what you mean by "Not unless it's hiding it's traffic." – user48838 Sep 17 '10 at 19:45
  • The network activity graph shows nothing to indicate high bandwidth utilization. OTOH some kind of malware might be able to bypass some counter or something and hide it's traffic from the reporting tools. – BCS Oct 19 '10 at 15:30
  • UDP traffic may fit the bill in that it is heavily used by P2P "applications" and is also sometimes difficult to measure and control. – user48838 Oct 20 '10 at 05:00
0

I just had a similar problem. I was able to resolve the issue by changing the channel that the Access Point was broadcasting on.

SLY
  • 1,286
  • 1
  • 13
  • 28