SDRs Open the Possibility for 18kHz Bandwidth HF Data
providing more robust communications and higher speed data.

With new Software Defined Radio (SDR) transceivers that use 
computer audio as the IF and DSP for filtering and modem, 
wider bandwidths than the traditional 3kHz SSB transceivers 
are possible. 

12kHz or 20kHz BW is certainly within the range of SDRs. 
This opens the possibility for some very fast and/or robust 
HF digital modes that can take advantage of wider bandwidths 
when needed, or could scale down and up in speed or bandwidth 
depending upon propagation conditions and need to coexist and 
share with other spectrum users. 

There are bandwidth limits for ham emissions in many countries.
In some countries there is a 6kHz limit. In others, it is 9kHz 
or 10kHz or 20kHz. Some countries have no bandwidth limit, other 
than the obvious requirement of simply keeping your signal within 
the ham band :) Some countries such as USA don't specify finite 
bandwidth.

Using 18kHz bandwidth, it is possible to have a digital mode 
that could decode at -20dB or -23dB SNR. That is significant, 
because it opens the possibility for a new use of HF bands that
often become unused in parts of the solar cycle when propagation 
is poor. When the band "goes dead", and everyone else is gone, 
one could simply switch to the wide/robust mode and continue 
communications... the equivalent of tremendously increasing 
transmit power on both ends of the QSO. 

Following is a side note for USA hams pushing the data 
speed-vs-bandwidth envelope: 

Contrary to what many USA hams believe, there is presently 
NO FINITE BANDWIDTH LIMIT enumerated for HF data emissions 
by FCC in USA.

There is an FCC rule that bandwidth shouldn't be wider than a
"phone" emission and that it be no wider than necessary for the
purpose. The 300 symbols/second limit for "data" in FCC rule 
DOES NOT SET ANY BANDWIDTH CONSTRAINT, and it applies only to 
the "data" subbands. If the content is images or audio sounds 
such as voice, there is no 300 symbol/second limit in the 
appropriate subband.

Assume that for important purposes, you need a data transmission 
to carry as much data on HF as possible in a very short time. 
Perhaps you are sending it for important EMCOMM purposes...
You also need it to be very robust, and work in negative SNRs.  
So, the wider the bandwidth, the better, since a wider transmission 
can be faster and can be made more robust with spectral and 
time redundancy. That important need satisfies the "no wider than 
necessary" requirement of the FCC rules.

The next question for USA hams becomes: "How wide is an HF "phone" 
emission, if one can't exceed that bandwidth?"

AM (Amplitude Modulated) "phone" emissions are the foundation of
FCC rules. A normal AM signal with human voice modulation can be 
as much as about 18kHz bandwidth. That is also the normal
AM broadcast bandwidth. Some shortwave AM broadcast stations use 10kHz
bandwidth. However, for communications purposes, especially by hams,
the AM audio modulation source is usually low-pass filtered with a
cutoff at around 3.5kHz to 5kHz, and this usually limits the AM signal
to about 7kHz to 10kHz. 

But, there are presently hams who communicate using higher fidelity 
AM with 18kHz bandwidth on HF (especially 80m & 10m) and MF (160m).  
Effectively, the status quo is what sets a defacto bandwidth limit 
on HF data in USA at somewhere between about 7kHz to 18kHz, 
depending upon how you want to interpret it or push the envelope.
Perhaps this open door will be closed to USA hams in the near 
future if narrow finite bandwidth limits are put in place on HF. 

The availability now of wide bandwidth on HF combined with 
availability of SDRs, or transceivers that can be modified for 
wider IFs, certainly provides a window of opportunity for 
experimentation by hams to develop faster and better HF data. 

Bonnie VR2/KQ6XA

Reply via email to