Information Technology of Jamming Cancellation

Introduction. Impact of the jamming leads to the high losses since it decreases effectiveness of radiolocation systems, anti-aircraft missile systems and communication systems. Strategies of forming and setting of the jamming are improving and the power of the jamming increases. In this regard, it is important to improve jamming cancellation systems. The task of the improvement for based on matrix calculations methods of the jamming cancellation is actual considering the breakthrough development of the computational methods which allows realization by digital circuit engineering. These include the most modern machine learning algorithms aimed at solving signal processing tasks. The requirement of the stable operation is important for the jamming cancellation systems under conditions of uncertainty. Other demand is an operation in the real time and a simple hardware implementation. The purpose of the paper is to increase the efficiency of the jamming cancellation in the antenna system (under conditions of uncertainty) based on the new randomized computation methods and their realization by the matrix-processor architecture. Results. The approach based on singular value decomposition and random projection is proposed. It provides effective jamming cancellation in the antenna systems under conditions of uncertainty that is, the sample has small length, there is an own noise of the measuring system, the input-output transformation matrix have undefined numerical rank and there is no prior information about useful signal. Communication. The increase of the efficiency of the jamming cancellation includes the increase of the stability and jamming cancellation coefficient, and the reduction of the computational complexity. The increase of the jamming cancellation coefficient is provided by use of stable discrete ill-posed inverse problems solution methods of the signal recovery based on random projection and singular value decomposition. The decrease of the computational complexity is achieved by the realization of random projection and singular value decomposition as the processor array which makes parallel computations. Keywords: information technology, jamming, machine learning, algorithms, singular value decomposition, random projection, conditions of uncertainty, signal recovery.