James Miller G3RUH
The root of the "alligator" problem is attributable to the use of a wideband, flat response linear transponder. The strongest signal sets the spacecraft ALC level, which reduces all other signals as well. This effect causes the others to increase their uplink power which depresses the ALC still further.... This kind of linear transponder has the virtue of simplicity of design, and is fine for phase 3 satellites, conceived and implemented as they are with essentially late 70's technology. Probably, phase 4 satellites will embody similar concepts. Now is the time to learn from our mistakes. The problem of one or two over-strong signals is overcome by splitting the transponder spectrum into small sub-bands, and selectively controlling the power in each band. This can be done with banks of filters, as in ARSENNE. (It wasn't - jrm 94/Nov). Yer phase 5 satellite will do it all-digitally via the Discrete Fourier Transform and cheap silicon. Assume a transponder pass-band of 64 kHz. This will be digitally sampled at say 128,000 samples/second. Now let's suppose we want our sub-bands or spectral cells 1 kHz wide. For this, samples are taken in bursts of 1 milli-second duration, making 128 samples per burst, 1000 bursts/second. Now the on-board signal processor applies the fourier transform to these 128 samples, so converting the signal from the time domain to the frequency domain or spectrum. Each of the resulting 64 spectral cells is checked and adjusted (even to zero!) so that inappropriate amplitude variations are evened out. The 64 cells are then inversely transformed back to the time domain (1000 times/sec), and pass on through the transponder as per normal. By this means every 1 kHz of transponder passband is allocated the same amount of power, so every user gets much the same service. This simple technology is available NOW. Doubtless the Dalligator or Digital Alligator will emerge to defeat it! For further reading see "Poisson, Shannon, and the Radio Amateur", J.P.Costas, Proc. IRE vol 47, pp 2058-2068, Dec 1959.
Feedback on these pages to KB5MU. Feedback on the article should be sent to James Miller
Created: 1994 Nov 13 -- Last modified: 2005 Oct 29