Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[beanieboy182]

With the standard Netgear firmware you have to enable debug mode before using telnet. With the DGTeam firmware there is an option to enable debug mode permanently. Have you got that option checked?

yes its on just after i uploaded it i made sure that option was ticked

i can telnet into the busybox at anytime from a cmd prompt with admin password ( admin password)

works like a charm

but using those scripts its the login1.txt and login2.txt that seem to be messing me about

i think someone said they should have +1 or +2 after them ??
how do i find that out ?

[rhohne]

The +1/+2 refers to the size of the file. They should not have DOS line terminations (2 chars) but UNIX style terminations (1 char)

If username is admin then size of file should be 6 bytes, likewise for password it should be 9 bytes. Have you tried restoring login1.txt and login2.txt

Code: [Select]

dir log*.txt
 Directory of C:\ADSL_stats\Apps
18/05/2012  20:34                 6 login1.txt
18/05/2012  20:35                 9 login2.txt
19/02/2012  01:29                 4 login3.txt
               3 File(s)             19 bytes


[beanieboy182]

here are the current stats finally

[burakkucat]

So looking at your QLN graph and attempting to obtain the data directly from my laptop's screen, you have peaks at the following approximate tone numbers --

Code: [Select]

130, 160, 173, 202, 210, 218, 250, 262, 291, 304, 360, 408, 432, 465, 480 & 501

Of those peaks, three correspond to the valleys in your Bits and SNR graphs. They are at tones 291, 360 and 465.

Performing the calculations --

Code: [Select]

291 x 4.3125 kHz = 1255 kHz
360 x 4.3125 kHz = 1553 kHz
465 x 4.3125 kHz = 2005 kHz


[beanieboy182]

sounding more and more like a noob every post
so now i know that what do i need to do next ?

[beanieboy182]

they are radio broadcasts i think on AM MW
but the higher up tones are the ones that are suspected to be correctable ?

[burakkucat]

You might like to bookmark this page, then check to see if those frequencies are approximately correct for a transmitter near you.

Please be aware that my calculations are only approximate and you should allow for a maximum error of about +/- 4 kHz. 

[beanieboy182]

well I'm in bristol uk

so i can see on the map that there is 3 radio stations very close by that are in that list
but someone said back in this post that the higher up ones are not going to be radio stations as they are 2 high??

[asbokid]

(click for full size)

Perhaps I can appeal again for collaboration on a project to programmatically analyse QLN data for RF ingress?

The general problem is one of noise peak or spike detection.  There is a good survey of peak detection algorithms for mass spectrometry analysis here [1]. The survey was authored by Chao Yang, et al from Hong Kong University of Science & Tech.  All of the algorithms they surveyed are open source. And although intended for proteome research, they should be re-usable for our project.

In that survey, the 'winning' peak detection algorithm, in terms of its sensitivity vs its false discovery rate, is an algorithm based on Continuous Wavelet Transforms (cwt).     The algorithm was developed by a team at Northwestern University, led by Pan Du. [2]   It is written for 'R'. [3]

The algorithm integrates a smoothing filter (to remove background noise), and baseline correction.

Further work would involve mapping the frequencies of the detected QLN noise peaks to the broadcast channels of AM radio. In Europe, AM (MW) radio band is assigned 9kHz channel spacing.  Whereas ADSL/VDSL2 uses spacing of 4.3125kHz.   So the channels don't immediately coincide.  Some 'jiggery pokery' / tolerance / fuzzy logic is involved.

Does that appeal to anyone here?!

cheers, a

[1] http://www.biomedcentral.com/1471-2105/10/4
[2] http://www.bioconductor.org/packages/release/bioc/html/MassSpecWavelet.html
[3] http://www.r-project.org/

[asbokid]

As far as I can deduce, it is that time of the year when No-Feathers McGraw takes Mrs Eagle on their annual holiday . . .

Poor old Paul. According to Judith Chalmers, Hawaii can be very boring at this time of year. Hope Baldy_Bird has packed some graph print-outs to enjoy on the beach!

cheers, a

[beanieboy182]

so the 2 higher up spikes you have put ?'s next to how would i go about working out what is causing them :S

[beanieboy182]

so i have to look for a frequency of 1768 kHz ?
and 2000 khz ??
and work out what is making that noise on my line?
please explain how this is possible or what i could try :S

sorry to ask for so much help
i really appreciate it and will use my knowledge in the future

[asbokid]

so i have to look for a frequency of 1768 kHz ?
and 2000 khz ??

$1m question!

Those frequencies were Burakkucat's estimates from your QLN graph, so maybe examine the raw tonemap data for QLN to find the exact tone indices for those noise spikes.

and work out what is making that noise on my line?
please explain how this is possible or what i could try :S

Perhaps a wideband radio scanner with a directional antenna tuned to those frequencies could identify the origin of the noise? Is it AM modulation? It might not be a 'broadcast' at all, just some noisy appliance in your neighbourhood.  Or maybe try a shortwave (~160m band) radio receiver?  The cheapo shortwave receivers sold by Lidl/Aldi unfortunately are no good for this. [1]

Or if you want to study it in depth, perhaps a software-defined radio (SDR), professional [2] or otherwise (with RF transverter) [3] ?

Have fun 

cheers, a

[1] http://forums.digitalspy.co.uk/showthread.php?t=1672360
[2] http://www.radioworld.co.uk/catalog/wr-g33ddc_excalibur_pro_winradio_receiver-p-8188.html
[3] http://dangerousprototypes.com/2012/03/20/introducing-rtl-sdr-a-20-sdr/

[beanieboy182]

just checked stats again and here are the results
looks like the pesky 460 x 4.3125 so around 1983 kHz
im gonna have to search for a radio that receives am up to that :S
because i don't have 2 grand to spend on the machine i need to trying to find that noise haha

[Bald_Eagle1]

As far as I can deduce, it is that time of the year when No-Feathers McGraw takes Mrs Eagle on their annual holiday . . .

Indeedy.

Back in Blighty now & it looks like I brought some sunshine back from Cyprus for the Olympics too 

Poor old Paul. According to Judith Chalmers, Hawaii can be very boring at this time of year. Hope Baldy_Bird has packed some graph print-outs to enjoy on the beach!

Haha. Far too busy enjoying the all-inclusive facilities at the swim-up pool bar & the evening through to early morning entertainment in the nearby Ayia Napa to even let the thought of a graph or a script cross my mind.



@ beanieboy182,

Glad you eventually got the script(s) working O.K.