Organization of the on-ground data processing for remote sensing purposes using finite rings and fields concept

In most of the modern on-ground data processing systems for remote sensing the digital signal conversions, on the analogy of the analog signal, are defined over a field of complex numbers, C, as the digital signal values (as well as those of the analog one) may be regarded as a subset of the field. Such an approach requires substantial computing and hardware expenses. For a signal defined in the finite range N (N = 0, 1, …, n − 1) most of the algorithms require the number of operations proportional to log N, N, N² and even to greater powers of N. That is why the building up of the data processing system mathematical model with a minute number of computations, as well as the computations organization, is a problem of great importance. It is not always necessary to build up a model using transformations over the field C. These transformations may be also defined over some other (abstract) fields, e.g. over the finite field GF(p) or finite residue ring moduli M, Z_M. This approach gives numerous advantages. They are discussed in the presented paper along with limitations of the approach. A series of the data processing system models are described. The design concepts for the on-ground images processing equipment are also discussed.