Signals And Data Fusion In A Two-Band Radar

The paper addresses important problems related to detection of small RCS targets. The probability of detection of such targets can be improved by fusion of non-coherent signals and/or plots from a few independents radars (channels) working in different bands. As an example a two-band radar system working simultaneously in S and C bands is considered. Its antennas are placed back to back. The paper describes the concept, similar to the MIMO method described in [1], of signal processing in this radar. The fusion of signals is called centralized fusion (detection) and the fusion of plots is called decentralized fusion (detection). The signal processing algorithms described in the paper are based on the original concept that for the fusion of signals (i.e., the centralized fusion) the cueing method is used. It means that signals reflected from targets in two bands are summed only if the targets have been detected by the decentralized fusion and are being tracked. Thus the centralized fusion takes place around the predicted target positions. The probability of detection versus SNR curves were calculated for separated and combined actions of the centralized and decentralized detections. Quantitative results of these simulations shows the possibility for substantial improvement of the target detection probability as compared to single-band radars.