A. N. Novikov - Ph.D. (Eng.), Research Scientist, Military Academy of Strategic Missile Forces named after Peter the Great. E-mail: band31@mail.ru

Adaptive systems are used successfully for solution of a number of tasks for a long time. One of the examples of complicated adaptive system is the adaptive antenna array where spatial signal processing is carried out. Spatial processing of broadband and superbroadband signals is connected with use of adaptive multichannel systems and weighed summation of various channels signals. Weight coefficients can be chosen on the basis of various criteria. An important stage of creation of such systems of signal processing is transformation of the covariance matrix of signals and creation of the device which can realize this operation. The greatest complexity here is creation of multichannel systems with the high speed of data transmission. To overcome this complexity it can be used the specialized devices intended for realization of a certain set of operations. Scientifically based recommendations for creation of the device realizing the most difficult stage of a method of spatial signal processing - transformation of the covariance matrix are worked out in this article. The transformation of the covariance matrix of interfering signals is proposed to carry out on the basis of iterative algorithm of the transformation with the use of the bordering method described in the well-known literature. However, unlike the earlier developed algorithm of the transformation of the covariance matrix of interfering signals for the construction of the reciprocal matrix, property of hermiticity is considered. It allows to reduce considerably computing costs and to increase the speed of specialized processor, and to reduce hardware expenses on the construction of transformation device. Thus, the algorithm of the transformation of the covariance matrix of interfering signals on the basis of the bordering method taking into account its hermitian properties is worked out in this article. On the basis of the developed algorithm the function chart of the block of the transformation of the covariance matrix of interfering signals is proposed.