Transmission data reduction in bandwidth scarce environment for the earth surface observation from satellites flying in close formation

The remote sensing Synthetic using Aperture Radar can obtain the information of moving targets on the earth surface using MTI (Moving Target Indication) technique, which detects differentiation of the surface image data obtained in successive sessions taken with two antennas mounted on one satellite or two satellites flying in a close formation. In the latter case, the operation of the differentiation is usually performed at the ground station, then each satellite needs to transmit the large amount of data to the ground station separately, which requires broad downlink bandwidths. However, the earth surface image data at the two satellites clearly have a strong correlation since the static background image data other than the moving targets are essentially the same except for some noise factors. Focusing on this fact, we claim that the transmission data can and should be reduced by taking off the redundancy of the two data and sending only the differential parts. In this paper, we discuss the coding theory matter related to this redundancy suppression first, and then discuss source, channel and network coding, and also the Slepian-Wolf theorem for two correlated sources. We next propose a coding and compressing scheme for a certain simplified model. We further generalize it in order to apply this idea to a more realistic model and present a procedure of the proposed scheme.