Pulsed wideband SNR problem solved by a dynamic carrier?

[metiman]

I am disappointed by the fact that receiver noise seems to be directly proportional to bandwidth. With a raised cosine shaped pulse transmitter the bandwidth is the more or less the reciprocal of the pulse duration. Short pulses mean large bandwidths and that means greatly reduced range, at least with current receiver designs.

What I'm wondering is whether there is any conceivable receiver design that could get around this problem. Note that I am talking about a pulsed carrier wave that carries no information other than the pulse itself. The gaussian filtered pulse would not be phase or amplitude modulated beyond the amplitude modulation created by on-off keying (OOK).

I'm thinking of something along the lines of a reciever designed to track and sync with the transmitter pulse shaping function in a similar way to how a frequency hopping receiver and transmitter are synced to stay on the same frequency. In both cases, if you freeze time at any given frequency, you should have only a very narrow band of frequencies that the reciever has to listen on.

In order to actually modulate this dynamic carrier frequency with information I would use pulse position modulation in the hope that I could achieve either an ultra low bandwidth or even a bandwidth free communication channel. I guess pulse duration modulation might also be used except that it might add sidelobes in the frequency domain.

This scheme is not intended to help with the wideband SNR problem in the general case. Just in the case of pulsed transmission induced bandwidth. So is the idea plausible? Do you think we will come up with a solution to the wideband SNR problem in the next 50-100 years?

 

[skeptic2]

Welcome to Physics Forums.

You seem to be headed in the right direction but you should also consider Direct Sequence Spread Spectrum (DSSS).

As far as a solution to the wideband SNR problem, I remember seeing an IEEE paper around 1980 that showed mathematically that ordinary phone lines couldn't support baud rates higher than 14,400.

 

[Mike_In_Plano]

Classic electronic warfare problem. If you open up the bandwidth, you're subject to the entire spread of frequencies and front end overload. If you narrow the bandwidth, you're subject to ringing from impulse jamming. No matter what you do, he's going to hit you with peaks that exceed your signal level and this will likely cause your AGC to squelch what you wish to hear.

Using DSSS, you need to get the receiver to sync with the transmitter. As long as they are out of sync, you have difficulty. Of course, it's much nicer once your in sync and all the other events are rejected.

With narrow bandwidth comes the need for dynamic squelching (when he rings you) and a smart AGC that won't get spoofed by an occassional peak. In addition, information redundancy / data spreading helps to reconstruct the destroyed portions of the message.