Re: [Discuss-gnuradio] Project enquiry/interest

[Sajeev Manikkoth]

Thanks Marcus, and adding some more details. Current signal detection
mechanisms for years totally rely on signal attributes (frequency, timing,
amplitude) for differentiation of signals. Cognitive detection mechanisms
and new PHY layer techniques which emulate human eye like detection and
differentiation need to be developed. One simple scheme I can think of is:

Human eye can differentiate 2 similar items or let us say identical twins.
And when we find it difficult, we add different identification marks on
those twins to differentiate and identify. In a similar fashion may be a
transmitter can add a unique identification while modulating/transmitting,
and the receivers can look for those. Receivers shall first tune to the
channel frequency, and then to the unique transmission id to latch to the
desired transmitter.

In general my interest was to see interest and projects which develops
cognitive detection mechanisms and associated new PHY and MAC layer
techniques. Hope I am making some sense now...

Best regards,
Sajeev 

-----Original Message-----
From: Marcus Müller [mailto:address@hidden 
Sent: 23 February 2015 19:55
To: Sajeev Manikkoth; address@hidden
Subject: Re: [Discuss-gnuradio] Project enquiry/interest

Hi Sajeev,

On 02/23/2015 10:56 AM, Sajeev Manikkoth wrote:
> Hi Marcus,
>
> Thanks again for the detailed explanation of current access technologies.
As
> discussed current scheme allows shared access of the channels in time,
> frequency, and space.
Yes, that's how I understood this discussion.
>  What I am talking is about a full simultaneous
> parallel use or access of channel.
I really really don't understand what you mean with that -- what other
dimensions than time, frequency, space (incl. polarization) and code can
you imagine, that would distinguish one electromagnetic wave from another?

>  This is kind of necessity as wireless
> bandwidth demands are ever growing and we are hitting spectrum scarcity.
Spectrum scarcity has been a reality ever since Marconi!
> The
> scheme I am discussing is close to CDMA/MIMO. CDMA base stations already
> differentiate handsets using same frequency with signature sequences.
> Implementing a similar approach on the handset side also to differentiate
> base stations or similar approaches can be in place.
CDMA handsets of course already do CDMA -- otherwise, they wouldn't be
able to communicate with the base station (which would be
disadvantageous, I reckon).
LTE handsets (at least from what I remember about LTE Release 8, which
is the original LTE release) can make use of MIMO. Probably they already do.
>
> In its simplest form the requirement is to allow 2 FM stations using same
> frequency in a location area. And the receivers tune to the station names
to
> enjoy different music rather than just to the frequency !
Well, that would then necessarily be some kind of diversification by
coding -- be it CDMA, or be it multiple lower-rate streams embedded in a
broadcast transport stream, which is what DVB does. That doesn't
inherently increase spectrum efficiency -- instead of 100 channels with
bandwidth b, you get 1 channel with bandwidth 100*b, because you can't
cheat channel capacity, and as long as you can't change SNR, the only
thing you can increase to get more information from transmitter to
receiver is to increase bandwidth.
>  Nothing new as a
> concept, limitations to achieve this reasons we have all the existing
> implementations, but 100s of years of engineering fineness. Now this
should
> be possible with soft transceivers using today's digital radio techniques
> combined with soft techniques...
What kind of soft techniques? Soft decision decoders?
I still don't really understand where you think that current technology
falls short and what's to improve, but I think I'm getting closer to
understanding exactly what kind of research is of interest to you;
please do elaborate!

Greetings,
Marcus