View Full Version : co-channel interference-- limitations of the DRM system?
To illustrate what I ment there:
http://www.drmrx.org/forum/showpost.php?p=26536&postcount=619 and a few posts before that.
This is a recording of 1296kHz in AM - same equipment, same location, same time of day.
To listen, just remove the .txt ending of the filename.
And that's why I say DRM reduces the target area.
maxpower
13-06-2009, 12:00
To illustrate what I ment there:
...
And that's why I say DRM reduces the target area.
Hi!
No, that can't be said in general. In this case itsright. But this is because BBC decided to transmit in a higher bitrate for better soundquality instead of extending the range of the signal.
If they would use bitrates lower than 20 kbit (16 kBit for example) in Mode A instead of 23 or 26 kBit you would be able to listen without problems even at your place.
Stephan
I doubt that because according to my measurements in case of co-channel interference (which you also can hear in my example) anything less than 15dB ratio wanted/unwanted station will cause total audio failure whereas AM-co-channel with this same ratio is noticeable but will not harm the possibility to follow the contents.
Sound Example for this (http://privat.albicker.org/kurzwelle-drm/sound/gk_am_-15db.mp3)
maxpower
14-06-2009, 22:00
In Mode A with 14.76 kBit you need something about 10 dB of SNR for dropout free audio. At groundwave reception even 9 is more than enough for dropout free listening.
With the 12 kBit that Rai uses sometimes, even less is enough.
Co- or On Channel QRM should be included in the SNR mesurements in Dream. So SNR goes down if theres QRM, but if some XX SNR is required for correct decoding it shouldnt matter if theres QRM or not. If SNR is high enough for decoding, its simply enough for decoding. :)
Listen to HCJBs 15280 kHy programming. They use Mode A 14 kBit, too. For example: http://www.drmrx.org/forum/attachment.php?attachmentid=49917&d=1243205264
Stephan
Lab measurements show that a 64QAM signal needs around 16dB SNR, a 16QAM signal needs around 12dB SNR for a BER of 10-4
Measurements also show that even a single carrier around 5Hz off center frequency (which is common for "co-Channel" Situations) kills the signal when stronger than above values.
And if it's argued that DRM is good for ground wave you cannot stop skywave AM interferers from reducing your target area anyway.
These are simply facts.
Testsignal 1 ("16QAM" bzw. "B 4/16QAM"):
RobustnessMode: B
Kanal-Bandbreite: 10kHz
SDC QAM Mode: 4
MSC QAM Mode: 16
Interleaver: long
Hierarch. Mode: std
Coderate: 0/0.5/0.62
Testsignal 2 ("64QAM" bzw. "B 16/64QAM"):
RobustnessMode: B
Kanal-Bandbreite: 10kHz
SDC QAM Mode: 16
MSC QAM Mode: 64
Interleaver: long
Hierarch. Mode: std
Coderate: 0/0.5/0.60
btw.: Mode A or B doesn't matter in this respect as this only differs in the length of the guard interval (important for multipath skywave propagation)
maxpower
15-06-2009, 12:19
Lab measurements show that a 64QAM signal needs around 16dB SNR, a 16QAM signal needs around 12dB SNR for a BER of 10-2
I've been in Quito for the last weeks and got dropout free audio at groundwave reception from Pifo even under 10 dB avarage SNR without any problems. 16 QAM, Mode A, 14,76 kBit. The enineers at HCJB approved my observations. Try 12 or 14 kBit Mode A in your lab as well and find out. :cool:
Mode A or B doesn't matter (...) as this only differs in the length of the guard interval
At Mode A you've about 10% more carriers available that transport the data. So with the same audiobitrate you can achive much better error correction which stablizizes the signal. This allows audio beeing decoded even at lower SNR.
Btw: Did you test HCJBs transmissions on shortwave 15280 kHz or the RAI programming on 846kHz? The should work with SNR < 11 dB -using bitrates lower than 15 kBit - without any problems even at your place. Please take a look at the SNR over Carriers index tab.
Please try out the software spark as DRM Transmitter as well. If I remember right I could "receive" dropout free audio down to about 7 or 8 dB lowering the signal in "lab environment"
Measurements also show that even a single carrier around 5Hz off center frequency (which is common for "co-Channel" Situations) kills the signal when stronger than above values.
Single carrier? Without any modulation? Wow, what kind of TX has that been? A SparkTX? :rolleyes:
A standard TX - with carrier only - somwhere near the DRM Center frequency wouldn't rot the complete signal in normal operation in real world!
The signal you measured should
a) have been very noisy or
b) very - veeeerrryyy - strong?!
:confused:
Interesting thing: Upper side Co-Channel QRM (for example 5 kHz up) seems to be more a problem then lower side. Thats maybe because of the pilot carriers that shouldn't be disturbed.
Dreams mesurements of the SNR are not very accurate at some situations. They are always estimations. But in gerneral they work out quite well.
Btw: How would the BBC @ 1296khz perform using digital Modulation at 16 QAM? Should work out very well at your location, even with some QRM, doesn't? :D At that operation Modes DRM can achieve at least the same coverage than AM (AM quality that listeners a willing to accept for listening). And if you take the lower output power (1296 is operated in DRM with 35 kW only) in consideration as well...
Greetings,
Stephan
Single carrier? Without any modulation? Wow, what kind of TX has that been? A SparkTX? :rolleyes:
Just a Rohde and Schwarz signal generator (take old SMG or newer SMV, results are the same) in CW mode
My good old RE signal generator cannot tune in 1Hz-steps.
But, off course, you can also use a DT230 test equipment for that purpose ...
Upper side Co-Channel QRM (for example 5 kHz up)
I spoke of 5Hz, not 5kHz, that's about 10% of inter-carrier-spacing of the OFDM signal.
take the lower output power (1296 is operated in DRM with 35 kW only)
That makes me wonder as a standard AM-TX of 500kW can operate in 150-200kW in DRM - or is that British TX a real post war oldtimer?
A standard TX - with carrier only - somwhere near the DRM Center frequency wouldn't rot the complete signal in normal operation in real world!
As I told you.
Carrier within clearly less than 5Hz off center may be as much as 15dB stronger than the DRM power, but with 5Hz and more offset it has to be reduced to -15dB compared to the DRM-signal not to harm decoding.
And for these cases it doesn't matter if you have carrier, AM, multitone or whatever interferer signal, even a second DRM-station as co-channel interferer, the result is always the same: you need - roughly - 15dB of signal-to-interferer ratio (I have to check for exact results with my report, but that doesn't really matter since natural fluctuations of field strength are far worse than that.
maxpower
15-06-2009, 16:33
I spoke of 5Hz, not 5kHz, that's about 10% of inter-carrier-spacing of the OFDM signal.
I know. :-)
I wrote about the typicle interferer at shortwave. Its mostly 5 khz down, up (and makes QRM with one of the sidebands...) or at the center frequency.
That makes me wonder as a standard AM-TX of 500kW can operate in 150-200kW in DRM - or is that British TX a real post war oldtimer?
I don't know, but it's probably a linearized MW Txer which is capable up to broadcast with about 10% AM Power (50 kW max) or it's simply not working at maximum power for bringing down power bills/tube cost.
As I told you.
Carrier within clearly less than 5Hz off center may be as much as 15dB stronger than the DRM power, but with 5Hz and more offset it has to be reduced to -15dB compared to the DRM-signal not to harm decoding.
And for these cases it doesn't matter if you have carrier, AM, multitone or whatever interferer signal, even a second DRM-station as co-channel interferer, the result is always the same: you need - roughly - 15dB of signal-to-interferer ratio (I have to check for exact results with my report, but that doesn't really matter since natural fluctuations of field strength are far worse than that.
Maybe you can post a picture of SNR over Carriers from this case?!
Maybe I understand wrong what you try to tell me (still jet-lagging). One carrier without modulation more than 5 Hz away from the center ruins the complete DRM Signal?
Stephan
I don't know, but it's probably a linearized MW Txer which is capable up to broadcast with about 10% AM Power (50 kW max) or it's simply not working at maximum power for bringing down power bills/tube cost.
The transmitter is only a few years old it is a Nautel NA200 capable of 200 kW AM or 80 kW DRM, so if its being run at 35 kW that is reduced power.
James
capable of 200 kW AM or 80 kW DRM
That corresponds exactly to a crest factor of 10dB - what I supposed above
I wrote about the typicle interferer at shortwave. Its mostly 5 khz down, up (and makes QRM with one of the sidebands...)
But that's "adjacent channel interference"
Maybe I understand wrong what you try to tell me (still jet-lagging). One carrier without modulation more than 5 Hz away from the center ruins the complete DRM Signal?
Maybe you have a look at this (http://privat.albicker.org/kurzwelle-drm/empfang.html#2)
maxpower
15-06-2009, 19:25
But that's "adjacent channel interference"
Depends on definition. I would presume that every signal, thats direct on the same frequency-range is On Channel.
If you've got an AM TX transmitting 5 khz up from the center of the DRM TX, it's lower side band (and a little bit of the carrier) is "on channel interferrence, because its interfering the upper half of the 10 kHz DRM Signal. Same thing at the other side.
With my PC/Soundcard SDR I have very few problems with adjectant channel interference, because the "digital filter" work very nice.
It's the AM TXers that transmit much wider spectrum than 10 kHz that interfere into DRM (or AM) transmissions and cause Problems. (Look at 6075 that interferes sometimes with 6085 kHz because of the use of more than 10 kHz Spectrum by DW Relais-Partners). But anyway.
Maybe you have a look at this (http://privat.albicker.org/kurzwelle-drm/empfang.html#2)
Thanks for that Information and pictures! I'll take a closer look at 1296 this evening. (Hmm, its, only working at the morning this summer season...)
It's maybe interesting for you that Transradio has build a Transmitter for AM & DRM thats able to achive up to 73% of the AM Power (with up to 125% AM Modulation).
So the Transradio 100 kW AM Transmitter can transmit as much as 73 kW DRM Power. http://www.transradio.de/DRM-AM-TRANSRADIO.pdf (Page 9). I think thats quite an impressive development and you loose someting about 2 dB Power switching to DRM.
The transmitter is only a few years old it is a Nautel NA200 capable of 200 kW AM or 80 kW DRM, so if its being run at 35 kW that is reduced power.
Thanks, James (upps, sorry, said Jeff :eek: ) for the information and pictures provided. I hoped that you or one of your collegues would add some info here. :) Solid state Modules?
Maybe it's possible to make some tests with lower bitrates and higher protection in Mode A @ 1296??? :D
Stephan
It's maybe interesting for you that Transradio has build a Transmitter for AM & DRM thats able to achive up to 73% of the AM Power (with up to 125% AM Modulation).
The latter is a feature for the US market where 125% of (positive) AM modulation is known (and also regulated by the FCC) for a long time. It only means that the transmitter has to have some more linear headroom (so that it is able to run 2.25x instead of 2.0x the carrier voltage as a linear peak voltage).
Before FCC rules went into effect even 140 or more % of modulation were common.
And by extending linear response of the amplifier/modulator even further you also can get 73% in DRM mode.
Dissipation loss is not the big problem since peak power is reached only for a small fraction of time.
The latter is a feature for the US market where 125% of (positive) AM modulation is known (and also regulated by the FCC) for a long time. It only means that the transmitter has to have some more linear headroom (so that it is able to run 2.25x instead of 2.0x the carrier voltage as a linear peak voltage).
Before FCC rules went into effect even 140 or more % of modulation were common.
And by extending linear response of the amplifier/modulator even further you also can get 73% in DRM mode.
Dissipation loss is not the big problem since peak power is reached only for a small fraction of time.
Actually what Stephan refers to is not just the overmodulation capability, the main improvement is the implementation of DPE algorithms in the DMOD3 which reduce the crest factor of the COFDM signal by about 3 dB, I think in tests developers from Transradio achieved more than 80% of the AM power, still within the specification (especially concerning MER and out of band emissions).
@Stephan:The Nautel NA 200 at Orford Ness is similar to the transmitters described in this Brochure from Nautel (http://www.nautel.com/Resources/Docs/Brochures%20&%20Spec%20Sheets/Analog/AM/na200_b_w_ss_2.0.pdf)
Simone
the main improvement is the implementation of DPE algorithms in the DMOD3 which reduce the crest factor of the COFDM signal by about 3 dB
Just calculate 125% AM modulation and a crest factor of 7 instead of conventional 10dB and you get ... roundabout 72% ... ah, yes.
So it's a combination of both effects.:)
For comparison: Without overmod cap. it would be 56%, without overmod and DMOD3 the standard value is around 40%.
maxpower
16-06-2009, 11:53
For comparison: Without overmod cap. it would be 56%, without overmod and DMOD3 the standard value is around 40%.
Thats still 16% better than standard. Not bad I think.
I took a look at your screenshots at your homepage.
http://privat.albicker.org/kurzwelle-drm/bilder/gk-snrf_5hz-0db.gif
I'm sorry, but I have to say that your Signal generator seems to have a very bad phase noise figure! :eek: I hope you haven't paid to much for it?!
I think probably ony BSKSA Riad is able to get a more bad AM signal on the air than you did in your "carrier only" tests. :D
I tried to reproduce your results, but I had no chance in "real world" DRM transmissions.
But here's a screenshot from RAI 846 kHz, decoding at 10/11 dB average SNR with single carrier (with a bad, but lower phase noise than your signalgenerator) interference. (rai1.gif, rai2.gif). As you can see, there's only a very small loss of some few DRM carriers arround carrier 43.
And as I use a I/Q SDR I was able to create a situation where I've a modulated AM Interferer that is as strong as a DRM station by eliminating the "Q" Line in.
As you can see there's surely a huge effect at SNR because of the AM modulated signal on almost "the same" frequency but signal is decoded without any problems anyway. BBC/DW uses something about 18 kBit Mode A for its transmissions. This Mode is very robust for such kind of QRM.
Now you may say: But my signal was even 15 dB stronger than the DRM station on the frequency. You related on the 1296 BBC frequency which is targeted for the north of france, Belgium, Nederlands and the northwest of germany.
As the BBC broadcasts this QRG with 35 kW 15 dB more power equivalents almost 1000 kW of AM power that has to be transmitted on same frequency. Which QTH do you think of broadcasting at 1296 with 1000 kW somewhere nearby?
...
maxpower
16-06-2009, 11:58
So I made my conclusion on your comments:
1. Your "lab mesurements" that shall show that a DRM signal is completly killed by an single carriere more than 5 Hz away...
...show only that your signal generator is not very good in phase noise
2. You build your example on "only 15 dB stronger signals"...
... which would postulate a 1000 kW AM Transmitter working at the same frequency (which would make it nonsense to broadcast on the same frequency - DRM or AM).
Normally the AM bands are coordinated -at least a little bit - and on SW if you have that kind of interferer you change frequency.
3. You try to compare DRM and analog transmission performance interfered by a noisy carrier only AM signal...
...which does not occure in real world. Try a ususal carrier at some strenght 2 or 3 khz away which does take place often on MW.
You'll get strong sinewave interference on AM (which listeners wouldn't accept) and can listen without problems on DRM Modes when Station broadcasts in proper Modes and Bitrates.
4. You try to compare AM transmissions with DRM Broadcasts transmitted in higher bitrate modes...
... So you compare apples with oranges. It's the broadcasters choice which audioquality he wants to transmit. And if he chooses to lower the range and higher the bitrate, thats his choice.
Your comparison would be the same like you compared a 10 kW AM Stereo TX with a 100 kW AM Mono TX and would say: AM Stereo lowers the coverage area. :rolleyes:
I think, DRM-OM, you're probably more a "non-DRM-OM". :mad: Thats my (last) two cents on this discussion.
73, Stephan
Btw: Thanks Simone for the information. I was able to take some closer looks into a Harris DX 100 and its modules some weeks ago. Modern Solid state AM/DRM TXers are quite impressive needing only that small footprints.
I think probably ony BSKSA Riad is able to get a more bad AM signal on the air than you did in your "carrier only" tests.
(...)
I think, DRM-OM, you're probably more a "non-DRM-OM". :mad: Thats my (last) two cents on this discussion.
Obviously you are not able to do quantitative measurements yourself.
In any case, I think you should reconsider your style of discussion.
Back to the facts:
The following is a screen shot of a pure DRM signal
Next is a screenshot of
DRM signal
carrier at +1kHz from DRM spectrum center frequency
power difference is 0dB
And now:
same as above, but carrier at +10Hz from center
The results show clearly what I said above: Carrier off DRM center is harmful to signal quality. How much depends on the exact frequency offset.
Now you can draw your conclusions.
btw: Test equipment:
Fraunhofer DT230 with channel simulator to mix in the carrier
AOR AR 7030 with Fraunhofer modification
Soundblaster Audigy as recommended by Fraunhofer
PC 3.2GHz with 2GB RAM
RF level is 55dBµV to get max. receiver performance (http://privat.albicker.org/kurzwelle-drm/bilder/aor_snrkurve.png)
I'm sorry, but I have to say that your Signal generator seems to have a very bad phase noise figure! :eek: I hope you haven't paid to much for it?!
and now you can go to Fraunhofer and complain about their noisy equipment ...
To make things complete here's another shot:
Same parameters as above, but 0Hz offset, that means: interfering carrier is exactly in center of DRM signal.
Very low effect on DRM signal quality.
In addition to this screenshot (http://www.drmrx.org/forum/attachment.php?attachmentid=50490&d=1245183804) at exactly +1kHz here is another one at +1.1kHz offset.
This shows that the DRM signal is vulnerable throughout the spectrum.
maxpower
17-06-2009, 10:53
This shows that the DRM signal is vulnerable throughout the spectrum. http://www.drmrx.org/forum/attachment.php?attachmentid=50514&d=1245229533
You try to betray me (us) with this screen :mad:
With this bitrate (23 kBit) at that low SNR (15dB) + extra Text or Multimedia-Services (xxkBit ?) in a 9 khz channel you wouldn't get any decoding anyway.
Apart from that: In my opinion you create direct mixing effects between your signals that make the effects. Intermodulation effects between a OFDM & a AM Signal mixed together in a lab.-environment is not a problem of the DRM System.
You search for setups that doesn't work and try to tell us this would be a general fault of the system. :eek:
"Over and out" from me at this point to you
you create direct mixing effects between your signals that make the effects. Intermodulation effects between a OFDM & a AM Signal mixed together in a lab.-environment is not a problem of the DRM System.
You simply accuse Fraunhofer to sell rubbish as a test system.
Just think about your words :mad:
The results show clearly what I said above: Carrier off DRM center is harmful to signal quality.
You don't seem to realize that your "lab measurements" are mainly testing the performance of the DRM software you're using. Let me assure you that in your case (interfering carrier near DRM center), the DRM signal remains 100% decodeable.
Sure, it's an interesting observation that you can confuse Dream (or some other decoder) by injecting a certain type of interfering signal. Dream is a fine piece of software, but it's not perfect, and certainly not guaranteed to give you optimal reception against all imaginable types of interference.
I have repeated your experiments and seen the same things you report. And then I simply used a notch filter to get rid of the offending carrier and continued listening without any problems. The notch filter doesn't even have to be very precise or narrow. In fact, you can easily notch out a 1-kHz wide chunk of DRM signal and decoding will still work because of error correction.
Please understand that I'm not suggesting that an average radio listener should be required to do such tricks to receive DRM broadcasts. But it would be relatively easy to modify the DRM decoder to make it less susceptible to this type of interference.
To be fair, I think other people in this thread should acknowledge that you have discovered a real problem in Dream and possibly some other DRM decoders. But I'm not so sure the problem is very significant, and more importantly you should not jump to conclusions about DRM's robustness (or lack thereof) based on this.
You don't seem to realize that your "lab measurements" are mainly testing the performance of the DRM software you're using
You are perfectly right, the DRM software is part of the complete system and therefore also relevant for the behaviour the listener will experience.
Occasional cross tests with the Fraunhofer software radio showed that it basically behaves the same way.
To be fair, I think other people in this thread should acknowledge that you have discovered a real problem in Dream and possibly some other DRM decoders. But I'm not so sure the problem is very significant, and more importantly you should not jump to conclusions about DRM's robustness (or lack thereof) based on this.
Thanks, my aim was primarily to
to bring to attention that there is a problem in the signal chain
to help to analyze this to track it down to the roots
to let people who have more knowledge in digital signal processing than I have think about a possible solution
since currently I see reception problems that I (as a long time short wave and AM listener) think an average listener will not accept (based on the experience with our customers).
Thank you again for your input, I hope this thread will be back to a beneficial and respectful discussion :)
Would you just be kind enough to send me the recorded IF signal esp. the one with 'carrier at +10Hz from center'?
(...)
verify your result and come up with a simple solution -notch filter. I'd like to give it a investigation.
Of course I can do that, but please be patient for a few days since I did the tests some time ago and have to redo them to record the IF files.
Please tell me how long they should be (regarding the play time) to be of use for you.
I'll upload the files to a webspace and tell you the address when I have them ready.
One other thing I should know before I start:
The Fraunhofer Software radio gives me the possibility to record
either the IF12 input
or an IQ12 signal generated from the input
Which one is or more use for you?
If you are about testing the whole thing again, maybe you could try it again with a bitrate ~14 kbps and one audio-channel only as your first test was with >20 kbps and two services.
Thanks for your interesting results.
Thanks for your timely response, Bernard.
The play time should be 3 minutes minimum to be useful. Of course if condition permitted, I don't mind longer files because my decoder takes some time to settle down after locking to the signal (usually decoder takes longer time to lock to the signal in difficult conditions) and I don't know if your DRM signal strength is variable during the whole experiment.
And the signal I need is that single plain 12KHz IF that is supposed to be fed into the line-in port of sound card(ie. no IQ signals).
BTW: if possible, it would be even better if you can also provide test samples of different robust modes and QAMs . That would be immense help for me, thanks in advance!:D
Occasional cross tests with the Fraunhofer software radio showed that it basically behaves the same way.
This is not surprising. I haven't used the Fraunhofer software myself, but I would assume they all use very similar FFT-based algorithms for DRM.
Here's a brief description of what I think is happening. The interfering carrier at the exact center of the DRM signal is orthogonal to DRM subcarriers and will not cause any problems at all. But at some offset (which is not an integer multiple of DRM subcarrier spacing) the energy of the interfering signal will "leak" into the subcarriers. See: http://en.wikipedia.org/wiki/Spectral_leakage
Like I already suggested, the cure is simply to notch out the interfering signal before the FFT operation, and it doesn't even matter if your notch filter removes a small part of the DRM signal. Of course, with a little bit of clever DSP, it should be possible to cancel an interfering carrier almost completely with only minimal damage to the DRM signal around it.
I don't know if your DRM signal strength is variable during the whole experiment
I would use the above setup, that would give max. signal quality of about 32...35dB SNR when undisturbed.
This way you can investigate the effect of the carrier without too much influence of noise or intermodulation.
File length will be around 5min, I hope this should be ok.
if possible, it would be even better if you can also provide test samples of different robust modes and QAMs . That would be immense help for me, thanks in advance!
I am not totally free in the configuration, but I can select between different setups.
Do you prefer Audio or PRBS (for BER Measurements) as MSC content?
Here's a brief description of what I think is happening.
Thanks for the explanation and the link, I think I have to read it several times to understand it - but it's not bad to learn new things from time to time .. :)
it doesn't even matter if your notch filter removes a small part of the DRM signal
I've seen this before when e. g. I reduced IF bandwidth too much or used Passband Shift. And this was what I expected DRM/COFDM to be: robust against loss of part of the spectrum.
All the more I was puzzled when I did my tests since they showed the straight opposite behaviour.
Sadly the AOR doesn't have a notch and I simply don't have the time to build one in this turbulent times.
So let's see what zfyoung will work out ...
I am not totally free in the configuration, but I can select between different setups.
Do you prefer Audio or PRBS (for BER Measurements) as MSC content?
I would prefer Audio as MSC content, for that is a typical broadcast setup. And I can also evaluate the impairment done to audio reception due to the minor dropout.
Wait for your good news.....
Wait for your good news.....
Just sent you a PM ...
The interfering carrier at the exact center of the DRM signal is orthogonal to DRM subcarriers and will not cause any problems at all. But at some offset (which is not an integer multiple of DRM subcarrier spacing) the energy of the interfering signal will "leak" into the subcarriers. See: http://en.wikipedia.org/wiki/Spectral_leakage
Since DRM subcarriers are orthogonal to each other interference effect also should be a minimum if the interfering carrier hits one of the subcarriers exactly.
This is what I did today,
carrier at +468.75Hz (or about there) at 0dB.
Result: SNR at 31.3dB and perfect audio.
... and again: 10Hz off (carrier at +478.75Hz) the picture is quite different.
So obviously it's all a matter of orthogonality ...
As the whole discussion is far off- topic of the old BBC WS thread in reception results (1296 kHz) where it had been started, I moved the posts to this new thread. So please use this thread from now.
Looking at the concerns that have been posted here, giving the impression the whole system can not work, I wonder how many users currently have this kind of problem when receiving DRM transmissions, especially in the target areas of the transmissions.
In all the years of receiving DRM I only noticed it once (about two years ago), and I was just at the edge of the coverage area. Actually there had been a Tx fault and as soon as I informed the station the problem was solved within a few days.
Simone
As the whole discussion is far off- topic of the old BBC WS thread in reception results (1296 kHz) where it had been started, I moved the posts to this new thread. So please use this thread from now.
Thank you, I was thinking of suggesting this - as I know this kind of procedure from other forums I am member in. :)
giving the impression the whole system can not work
nobody claimed that, I think.
But I also think that it wouldn't be a good idea not to use room for improvement.
Co-channel may not be a problem with the few stations working at this time but is a very serious problem in MW reception at night (this is just long-time experience as "DXer", but also proven by professionals, see BBC 855kHz report).
And I personally think that improvement in this direction would be a contribution in reducing the nighttime shrinking of the coverage area in presence of an AM station on the same channel (situation of course will be the same as before when you have two DRM stations co-channel - but this in another discussion)
I think the official DREAM software still needs some improvement in case of strong co-channel interference.
This is the screen shot of MSC constellation after channel equalization where the interference carrier happens at different frequencies.The DC of DRM is on 12kHz, and the character of carrier is as denoted on the title of each JPG file.
As we can see: The gravity center of each cluster of 'signal cloud' is not aligned with the center of grid box on some occasions which happens to show very bad SNR. Actually, they are drifting outwards like on a inflated balloon. As a matter of fact, under normal conditions, no matter how noisy the signal is, the mean signal point for each QAM (ie. the center of each cluster of signal cloud) should remain more or less the same after channel equalization.
So, it seems something is not right in its channel estimation/equalization module?
PS: testfiles courtesy of DRM-OM
vBulletin v3.5.3, Copyright ©2000-2012, Jelsoft Enterprises Ltd.