Forum

Frans

Pse PM or e-mail me and let me know when you will be on Skype. Have a few things to chat about

http://news.yahoo.com/mars-bound-nasa-rover-detects-radiation-huge-solar-192605227.html

Dave

Remember that, unless properly shielded and oriented, that antenna cables become part of the antenna's ground-plane which can affect the (RF) pickup patterns. I see that [u]all[/u] the time.

In addition, inconsistent cable routing can really make one device behave very much different than another of the same exact model. That's not just RF cables either. The smaller, and more cramped the cable feed area, the stronger the negative influence.

Hi Wlanman

Good points. That may have been the reason for some of the smaller "objects" that were seen on that particular laptop at that particular time. The main interference areas seem like some form of broadband interference. It was only picked up due to the fact that a training class was in progress, otherwise it probably wouldn't have been noticed. This has been repeated on several different brands and at several different locations. Low level testing in the ISM band ( and adjacent bands ) by military/govermental units has been seen before. Rarely in the US, mostly overseas. Usually testing lasts a week or two. Most users just assume it's regular old interference.

Years ago, the UK was involved in a conflict with another country. Britain lacked adequate satellite communications capacity. Their opponent had a large carrier on a transponder. British units used a very low power spread spectrum signal to "hide" in the noise floor beneath their opponent's signal. They got a couple of months of free communications before their adversary discovered that something was under the noise floor, by chance.

The solar issue is still possible. We don't know enough about it. A lot of research is going on. Massive amounts of refraction can take place. Plots of high energy particle distribution ( these particles are known to corrupt or even destroy semiconductor memories...e.g. the very narrow gates in IGFETS....this happened in the 80s to a number of ATM cash machines on the west coast...they just failed, no masses of free cash ) show the particles reaching the earth on the sun facing side and then "bending" all the way around to the "dark side" of the earth at that point in time. RF signals become severely refracted.

NASA is concerned that as well as disruption to communications, the most important thing in the event of a Massive Solar Event of unprecidented energy would the damage to physical electronics which had not been hardened, such as with an EMP.

In many cases like this, you never really find out what was responsible, as it would involve some very sophisticated test gear.

It's an interesting one though.

Dave

Frans

Spoke to Fluke today. Patterns were not internally generated, so it's some form of broadband interference. If you're still there, drop me a PM and I'll be able to discuss it some more.

Dave

Hereby some stats on part of the sniffer traces. Almost all Acks are showed, so the number of data frames should be 2x the nr of Acks for a RA. Especially g/n misses are massive > 90% at some days.
Used the following filters:
wlan.ra == 58-94-6B-7E-E2-64 and wlan.fc.type_subtype==0x1d and wlan.fcs_good==1
wlan.addr == 58-94-6B-7E-E2-64 and wlan.fc.type_subtype==0x2c and wlan.fcs_good==1
wlan.addr == 58-94-6B-7E-E2-64 and wlan.fc.type_subtype==0x28 and wlan.fcs_good==1
wlan.addr == 58-94-6B-7E-E2-64 and (wlan.fc.type_subtype==0x28 || wlan.fc.type_subtype==0x2c) and wlan.fcs_good==1 and wlan.fc.retry==1

Do you get as many misses on the 2.4 GHz g/n ?

Date time type Acks Null QoS-Data Miss% Data retry %

27-jan 18:31 a 115 48 124 25 15
28-jan 10:35 a 215 70 264 22 1
28-jan 20:04 a 3871 3390 1321 39 8
29-jan 10:58 a 1807 1467 1076 30 18

26-jan 10:47 an 231 56 219 40 38
26-jan 11:02 an 228 90 222 32 28
26-jan 18:52 an 607 248 758 17 18
27-jan 10:42 an 578 89 165 78 16
27-jan 15:53 an 254 90 224 38 31
28-jan 10:25 an 148 45 229 7 36
28-jan 12:41 an 12082 6325 17959 0 12
28-jan 13:04 an 8461 3912 12892 1 16
28-jan 14:16 an 7808 2736 7303 36 3
28-jan 20:09 an 2571 2233 908 39 7
29-jan 0:13 an 10788 1895 17830 9 6
29-jan 10:39 an 614 556 229 36 17
29-jan 10:45 an 1213 936 1099 16 27
29-jan 10:50 an 1107 842 822 25 21
29-jan 11:03 an 1273 1003 969 23 20
29-jan 14:28 an 2813 1914 2065 29 23
29-jan 14:37 an 4809 1312 4700 37 14

25-jan 20:41 b 15663 0 27716 12 10
25-jan 20:44 b 542 42 997 4 3
29-jan 14:58 b 4734 95 8944 5 2

27-jan 18:25 g 139 10 236 12 1
28-jan 13:41 g 8707 755 12863 22 5
28-jan 13:49 g 5595 589 7763 25 3
29-jan 14:56 g 3853 526 6321 11 1

24-jan 20:01 gn 9972 24 804 96 49
25-jan 19:50 gn 2643 106 179 95 100
26-jan 10:21 gn 250 40 25 87 8
26-jan 18:52 gn 5729 35 280 97 71
26-jan 22:00 gn 2990 154 4047 30 20
27-jan 10:47 gn 490 32 67 90 2
27-jan 10:52 gn 400 32 53 89 11
27-jan 12:36 gn 180 29 22 86 37
27-jan 15:57 gn 237 38 217 46 26
27-jan 18:19 gn 139 68 124 31 30
28-jan 13:29 gn 7607 1094 3 93 2
28-jan 13:34 gn 2642 308 0 94 1
28-jan 13:56 gn 13646 998 32 96 5
28-jan 14:58 gn 6138 1211 6801 35 2
28-jan 18:22 gn 2978 373 3278 39 10
28-jan 18:36 gn 2696 1781 2138 27 24
28-jan 19:29 gn 2353 1540 1796 29 21
29-jan 0:07 gn 342 16 698 -4 34
29-jan 11:11 gn 678 265 855 17 14
29-jan 14:48 gn 1792 94 2050 40 13
31-jan 21:58 gn 363 115 283 45 23

didn't show up as nice as I had it lined up. Bottom line, is it normal to get up to 97% misses on g/n ?
Maybe I have been focussed more on auth, eap and voice rather than 11n data frames, I hope someone can shed a light on this.

Look on the net for the "MultiPing" application program. Download the trial version, and install it.

It will give you a visual display that will [u]show[/u] you when the problems really occur. Much handier and informative than looking at Ping log files.

I often use the licensed version, and it does a very good job.

Having confirmed with Fluke that the patterns are not "normal" for their equipment, under any mode, it is not solar interference. Sent the traces to a signals analyst, who concurred. It's some other form of interference.

Dave

A protocol analyzer, or “packet sniffer,” is a tool used to intercept traffic, store it, and present it in a decoded, human-readable state. Modern protocol analyzers like Wireshark can even spot rudimentary problems on their own and then perform statistical analyses with captured data.