Modern communication systems often operate in a complex interference and signaling environment, while there are various ways to reduce the error of the restored signal. Some of the methods relate directly to mathematical processing algorithms in the receiver, however, there are other approaches based on spatial filtering of signals. In particular, in recent years, an approach based on the weighted processing of signals received from different antennas has been actively developed using the correlation matrix of the input signal, which makes it possible to use information from the antennas more efficiently by choosing an antenna with the maximum level of useful signal and lower levels of noise and interference, which is physically the formation of an equivalent radiation patterns of the receiving antenna array with a maximum on the path with the maximum level of the useful signal and minima on others. The application of this approach is of practical interest, especially in systems with active interference, for example, from electronic warfare stations, as it can improve the quality of signal recovery. Separately, it should be noted that in the case of active interference, a method based on the minimum RMS error of restoring the pilot tones should be used to select the eigenvector for weight processing (which is possible in OFDM), since if the maximum eigenvalue is selected, it is unknown whether it will be signaling or interfering if it is high. This paper presents an experimental study of an adaptive algorithm for processing spatiotemporal signals for a MIMO-OFDM communication system with different levels of active interference from an electronic warfare (EW) station. In this case, experiments are carried out both in the descending (Downlink, from the base station to the mobile) and ascending (Uplink, from the mobile station to the base station) channel using adaptation on both the BS and MS sides, and both. It is shown that the application of the algorithm can improve the quality of signal processing and reduce the bit error rate for a wide range of signal–to-noise ratios (SNR – signal-to-noise ratio), even with imperfect channel estimation (by pilot tones).
